The Deptford Phase: An Archeological
JEiALD T. MILAVICH
A DISSERTATION rP.ESFNTD 70 THE CGRAilUATE COUNCIL OF
THE LUIVERSITY OF FLORIDA IN PARTIAL
FUL.F!!LL ENT OF TIHE J EQU!REILSNTS FOR T'IE DEGREE OF
LOCTOR OF PHILOSOPHY
UNIVEPSI'. OF FLORIDA
Jerald T. Milanich
ACKNOWi LED 1CMENTS
The nature of archeological research is scmeachat caricious, Often
data gathered in the field is not related to the problem. inder study.
T'nis unhoped for but not unique occurrence can !ead to a number of
alternatives, ranging from very unpleasant to very pleasant. The
disser action research presented here began as a comparative study of the
post-A.D. 700 aboriginal cultures of norlh-central Ilorida aitd southeast
Georgia. Two months of field researol on Cumberland Isir;d, Georgi, .
how:.ver, produced infor.atiori on he DeptforC culcuri-o hich began nearly
one thousand years earlicr than the period undci study. It is honed
the result-; of this change are as pleasant to th: readc-C as they are
(in retrospect) to nmy'slf.
Field searchh on Cumberland Island was suLprorted by a Natioral
Science Foundation Doctoral DiTssertrtioi Ilmrovc;-net Grant awerded to
Dr. Charles H. Fairbanks, as project supervisor, and to this author, as project
student researcher. I was supported while writing the dissertation
by a Nitional Defense Act Title IV' fellowship administered through the
Graduate School of the University of Florida.
During the two months of field research, any of the part and
full-time residents of the island aided the research expedition in
many ways. The families of Joseph Graves and Coleiann Johnston
pro-vided houei!g for myself and the students used as laborers. Warren
llitl.ock and the Richard 0. Fcrguson family helped to solve the many
logistic problems i:hich arise while conducting arheologi cal research
on an island accessible only by boat or plane. Coleman Perkins,
representing the Table Point Corporation, and Lucy R. Ferguson
graciously granted permission for excavation of sites on their land.
Special thanks are extended to Dr. and Mrs. William R. Bullard
and .!r. and Mrs. Robert Rischarde for many kindnesses shown to
myself and the excavation crew. I am also grateful to Sea Camp
car.p grounds for the liberal use cf their ice, bathrooms, showers, and
The members of my committee, Professors T. A. Nunez, C. H.
Fairbanks, II. R. Bullard, I. R. Maples, and T. H. Patton, were prompt
in their review of this manuscript and offered helpful suggestions
for its revision. Professor Drexel Peterson of Meimphis State University
and Miss Betty Anderson of the University of Georgia provided me with
unpublished data pertinent to my research.
During the preparation of this manuscript I have relied heavily
on the previous research of Dr. Joseph R. Caldwell of the University of
Georgia, Dr. Cordon R. Willey of the Peabody Museum, and the late
Antonia J. Wearing, Jr. At times I have been critical of certain of
their conclusions regarding the Deptford Phase. The importance of
their contributions to our knowledge of the Deptford Phase, however,
should be stressed. Most research is dependent on previous research,
and this dissertation is no exception.
Many of the ideas presented in the text were formulated out of
discussions with fellow students and members of miy committee. Analysis
of the fa:ual remains was c.rrjied out by Mr. Curtiss E. Peterson,
w-orking under Elizabeth Wiing at the Florida State 'Museum. M!r. Peterson
interrupted his own research to interpret and analyze the Cumberland
food bone. His identifications are listed in the Appendix and are
referred to throughout the text.
-. e two individuals to whom I am most grateful are Drs. Maxine
L. Margolis and Charles H. Fairbanks, both of the University of
Florida. Dr. Margolis, who is also Mrs. Milanich, and myself spent
many long hours discussing (often rabidly) the processes of cultural
evolution and their theoretical application to the Coastal Tradition.
I have studied under Dr. Fairbanks for seven years. His intellectual,
professional, and personal influences on me have been extremely great.
I know no way to adequately thank Professor Fairbanks for his guidance
and inspiration except to pursue a professional career toward which he
can point with pride as my teacher and friend.
TABLE OF CONTENTS
ACKNOWTLEDGMENTS . . . . . . . . . . iii
LIST OF TABLES. . . . ... ...... . . .. viii
LIST OF FIGURES . . . . . ... . . . ix
ABSTRACT ........ . . . . . . . .. x
CIIAPTER I THE PROBLEM AND THE APPROACH . . . .. 1
Methodology .. 2
Research Goals .. .. .. .. . ... . 16
CHAPTER II EXCAVATIONS ON C'J'.BERLAND ISLAND, GEORGIA . 25
Stafford North . . . ........ ... . 29
Table Point . . . . . . . .. . 46
Cumberland Island Subsistence Inforc-ation . . . 73
CHAPTER III GEOGRAPHICAL AND NATURAL SEITINCS . . . 79
Geographical Setting .... . . . . . 79
Natural Setting . ............ . ..... 90
Transhlunance, the Coastal Tradition, and
Enviroi;ncn tal Adaptation . . . . . 1
CHAPTER IV LUI1,TUiRAL jAND TEMPORAL SETTINGS . . . .. 116
Thr PrO-iD ford C-as'tal Tradition Cultures . .. 116
N;o-Coa-t-il Tradition Cultures of the
Southeastern Area . ... . . . . . 132
Origils of Deptford Paddle-Malleated Pottery . . 139
Dept ford Phase Evolution .. ........... 1I3
Post-Deptford Coastal Tradition .. . . .... 1.48
Coeiaspt Tradition Ch.ronology ............. 153
Su a. ary . . . . . . . . . . 153
TABLE OF CONTENTS (Continued)
CHAPTER V DEPTFORD PHASE MATERIAL CULTURE . . ... 160
Ceramic Artifacts . . . . . . . . 161
Stone Artifacts . . . . .. . . . . . 173
S ell Artifacts . . . . . . . . . 180
Bone Artifacts . . . . .... .. ...... 132
Yent Co.rplex Artifacts . . . . . . . .. 182
Summary . . . . ...... . .. . .. 185
CHAPTER VI OTHER DEPTFORD PHASE CULTURAL TRAITS ..... 1. 91
Settlement Traits . . . . . . . ... 191
Settlement Patterning . . .... . ... ... 192
Community Patterning. . . .. .... . . . 198
Household Patterning.. .. . . . . . . 201
Social O;an nation ..... . . . . 203
Trade . ... . ..... . ......... . 205
Eartlw.:crk and Eurial Compplexes . . . . . . 206
Ethnic identity of the Deptiord People ...... 209
CitlPI'ER VII SI- 'ARY OF CONCLUSIONS . . ... ..... 22
Tl']e Dcptord lPhase .. ............ .. 212
The neptford Phase Reconstruction and Tiaxonomy:
A Comment . . . . . . . ... 218
APi'EN'DI: VERTEBRATE FAUNAL ANALYSiS OF THE
TABLE 'POJXT HOUSE SITE .. . . ... ... . .22
REIERENCES. . ... . . . . . . . . 228
TIOGRAPiT.CA.L SKETCH . . . .. . . . . . 39
LIST OF TABLES
1 Test I Potsherd and Food Bone Totals . . . ... 34
2 Test II Potsherd and Food Bone Totals . .... . 38
3 Test IV Potsherd and Food Bone Totals . . ... 42
4 Ring Excavation Potsherd and Food Bone Totals . 52
5 House Potsherd and Food Bone Totals . . . ... 71
6 Faun.al List of Species Identified from House
Excavation . . . ... . . . . . 74
7 Faunal List Identified from Cumberland Island
Excavations .... . . . . . . 75
8 Chronology ..... . . . . . . . 154
9 Radiocarbon Date References and Proveniences . . 155
10 Identified Floral and Faunal Species from
Deptford Sites . . ... . . . 195
LIST OF FIGURES
1 Georgia Sea Islands and Cumberland Island . . .. 27
2 Stafford North Excavation Schema. . . . ... 30
3 Stafford North Test III Profile . . . . ... 37
4 Stafford North Test IV . . . . . .... . 40
5 Table Point Excavation Schema . . . ... . . 48
6 Central Portion of 3 by 34 Foot Ring Cut ...... 50
7 Reconstructed Profile of Ring Gap Features . . .. 54
8 Top View of Ring Gap Features . . . . ... 55
9 Table Point House Excavation . . . . . . 63
10 Table Point House Prof ile .. . . . . 64
11 Reconstructed Ilo,.se Pattern . ... . . . . 66
12 'Th- Deptford Region . . . . . . . .. 80
13 Soitheastern Coastal Plain Schenatic Cross-Section . 91
14 Deptford Pottery . . . .... . . . 188
15 Deptford Pottery . . .. . . ........ ..... 189
16 Deotford Tools .. . . . . . . . . 190
Abstract of Dissertation Presented to the
Gradual'- Ciouncil of thie University of Florida in Partial Fu]fillment
of h e. Requirements for the Degree of Doctor of Philosophy
THE DEPTFORD PHASF: AN ARCHEOLOGICAL
Jerald T. Milanich
Chairman: Charles [1. Fairbanks
Major o- D rpa1 l'.lit: Anri'.ropolo. y
Ihe Lep.-ford Phase, an archoologicaliy. known aboriginol cult u-,
existed along ite Gulf and At?:liic coasts of the Southeastern Unite.d
St.'tes frim 600 B.C. to as late as A.D. 700. St.ritigaphic evidcnc
has preiiosliy shc,;n that Lthe Deptford Phase fo'ie-:C.d tihe ceCadi:ic
Arc.aic cultures and preceded the Swift Creek Phase ill Western
Florida and che Wi inngton Fhase of the Georgi. coast.
Beyond chronological placement and descriptions of Fottery types,
little information concerning Deptford cul ture has been produced.
Using reinterpretations of orsz'vJs research and data f.o:, recyl:
cxc-.vatlons oi Cu-,berland Island, Georgia, a reconstrucLI on of
Deptford culture is formulated, and the Deptfcrd Phas- is placed in
the context: af Southeastern prehistory.
The methodological approach employed in this synthesis is popularly
knctwn as processual archeology. Tlis approach stresses the functional
interrelatedness of culture traits and the relationships of the culture
to its natural and cultural environments. The Deptford Phase is
defined on the basis of trait complexes which occur together in
specific natural-environm.c-ntal Liotopes.
These trait complexes include the following: sand- and grit-
tempering in pottery; some sherd- and fiber-tempering; carved paddle-
malleated pottery with some cord marking, punctuating, and brushing as
mniirity surface treatment forms; use of the atlatl and later the
bow and arrow: infrequent use of stone tools, including large and
small triangular points and stemmed points; and poorly developed bone
and shell industries. House types include oval structures with wall
trenches and oval, chickee-like structures. Other Deptford traits -- Yent
Complex, earthworks, subsistence patterning, settlement patterning,
social organization -- are also described.
The Deptford region extended from North Caro!ina south to northern
Florida on the Atlantic coast. On the Gulf coast the phase extended
from Alabana to Charlotte Harbor, Florida. Spcis'ic evolution of these
t-.o .s~'lb-regios seer;s to have occurr-ed sorewhut indcpenidently -f one
another. Site locations .and faunal evidence indicate that the phase
was adapted to both the coas; a bione and h:e river valleys of the
Coastal Plain pine forests. Prob;bl th e Deptford peoples moved
seasonally between these locations.
Subsistence techniques '.:ere cec.tered on hunting, gathering,
and fishing and were circumscribed around band level social organization.
Tic coastal sites were on ecotones ,'ithin the live oak strand, bordering
the salt narsh and lagoon. A variety of fish and marsh animals were
utilized along with deer, turtles, and other hammock fauna. Sea
turtles, seals, and whales or porpoises were also exploited. The river
valley sites were bases for the gathering of the Fall harvest of nuts
and seeds. Possibly this transhumance pattern was reinforced by
social customs, including band exogamy.
Deptford is a unique phase within the Coastal Tradition, a
way of life found on the Southeastern Atlantic coast from 2000 B.C. to
A.D. 1200. On the Gulf coast the tradition ends by A.D. 1. On the
basis of subsistence and material culture the Coastal Tradition can be
archeologicaily differentiated from the Savannah River and the Gulf
Increased contact with other Southeastern phases, including
Tchefuncte and Adena-Hopewell and the Savannah River Tradition, was a
primary influence in the emergence of the Deptford Phase. Heavier
contact in the Gulf sub-region perhaps caused the earlier termination of
the Coastal Tradition in west Florida. The Atlantic sub-region Deptford
Phase was marginal to the areas of rapid Southeastern culture change.
Thus, the Archaic subsistence pattern of the Coastal Tradition existed
until the close of the first nillenium A.D.
THE PROBLEM AND THE APPROACH
Archeology is a scientific methodology by which information on
past cultures is collected and interpreted. As a sub-discipline of
anthropology, the goal of archeology should be a Science of Man and
Culture characterized by predictability in the form of deductive
theorems. The study presented here is a description and analysis of
one specific culture, the Deptford Phase1, known from archeology to
have existed in the Southeastern United States nearly two thousand
years ago. To the uninitiated in Southeastern aboriginal culture
history, and in view of the above idealized statements, such a project
may seem insignificant. In fact, however, the progress of "dirt
archeology" in the Southeast over the last thirty years has been such
that there is an urgent need for such studies -- definitive intcr-
pretations of specific cultural units based on interpretations of
cultural processes rather than solely on descriptions of material culture.
'Phase is used here as "an archeological unit possessing traits
Sufficiently characteristic to distinguish it front all other units
similarly conceived, .l:hether of the same or other cultures or civili-
zations, spatially limited to the order of m;ognltude of a locality or
region and chronologically limited to a relatively brief interval of
time" ('illey and Phillips, i958: 22).
Such an approach, though in practice rare, is certainly not unique.
;'illey and Phillips in their presentation of a theoretical methodology
for American archeology state the following:
In so far as the foregoing definitions and stated
relationships can be formalized as a program for the
integration of New World archeological data on the
descriptive level of organization, they may be
summarized as follows:
1. The primary emphasis should continue to be
placed on the formulation of basic units, component
and phases, in local and regional sequences under
2. Phases should be studied intensively in their
cultural and natural contexts.
3. Their external spatial and temporal dimensions
should be kept within manageable limit.; of magnitude.
J. Their external relatici lhips should be studied
and expressed by means of intcgrative units, horizon
5. Large-scale integrating syntheses should be kept
within the limits of the "are-" as defined herein,
and horizontal correlation of phases in such
schemes should be effected so far as possible by
means s of indndrendent cxtracultural data.
6. On the basis of these integrative studies, phases
should be combined when possible to form the maximum
ub:its, culture and civilization.
7. Constant effort should be made to invest all
units of whatever magnitude with the greatest
possible intelligibility in both the cultural and
social aspects (1958:57).
lBeors discussing the research schema in mere detail, a review of
che theoretical constructs utilized below,; in reference to the Deptford
Phase is neede-'. Ideally this ,.ill allow future researchers to more
fully assess and reinterpret the data. The definitions employed
relate specifically to the problem under sc,-utiny and should be
applied to other archeological studies only after carefully examining
their ability to fully reflect both the culture under study and the
types of information desired from the study.
As has already been stated above, the research project oees its
design largely to hilley and Phillips (1958). In the study this
author has sought to incorporate a theoretical position popularly
known as processuall archeology (Flannery, 1967; Kushner, 1970).
Simply stated, processual archeology seeks to: (1) describe and
explain the adaptations of a culture to its environments, both its
natural habitat and other cultures; (2) describe and explain the
relatedness of cultural systems to one another, to the total culture,
and to the environments; and (3) describe on the basis of empirical
data cultural laws which are ascribable to all prehistoric cultures
o-r specific types of cultures. A basic assumption of processual
archeology is that given sufficient time and conditions a culture will
adapt to its environments. Such an adjustment, however, is not reached
by all cultures due to a variety of factors, including changing
r-;tuiral environment or a physically imposed new: cultural systemss.
While it is important to reconstruct the cultural-natuial
environmental relationships, this is not an end, but one step toward
a culI'ural reconstruction and interpretation. The description on of the
Depcford Phase rcfercd here encompasses subsistence processes and
ecological -idaptations as :well as material! culture and other aspects of
culture. Such an approach, vhiich ideally allows comparative statements
regarding cultures to be generalized, is the first step toward the
creation of culturological theorems in anthropology, in short, toward
a Science of Culture.
Tradition and Horizon
Phases are in a continual state of change. Through time the
cumulative effect of separate changes is a new phase, one sufficiently
different from the old to be discreetly recognizable in the context of
the archeologist's definitions. The evolution of phases results in
culture or phase sequences. It is often possible to trace certain
cultural processes and material manifestations throughout related
evolutionary phases. Such a concept is termed a "tradition". Willey
and Phillips (1958: 34-40) have outlined the origins, development, and
current usage of the term in New World archeology. Perhaps the usage
cited most often is John M. Goggin's:
a distinctive way of life, reflected in various
aspects of the culture; perhaps extending through
some period of time and exhibiting normal internal
cultural change, but nevertheless throughout this
period showing a basic consistent unity. In the whole
history of a tradition certain persistent themes dominate
the life of the people. These give the distinctiveness
to the configurations (1949: 17).
Actual application of the concept, however, has not been consistent
by Eastern United States prehistorians. Goggin (1949, 1953) and
this author (1968) have both written on the Alachua Tradition of
north-central Florida. The Alachua Tradition is a "unilinear"
tradition -nco..passing the evolution of a single line of phases, one
developing into another through time. On the other hand, Caldwell (1958)
apDlies the tradition nomenclature both to similar 'unilinear"
traditions and to ";aacro-iraditions", which are comprised of several
separate lines of phase evolution, e.g., Northern Tradition, Middle
Eastern Tradition. Taxonomically the Alachua Tradition is certainly
:tot equatable to the Northern Tradition (and may in fact be one of
the phase evolutionary lines within that tradition).
This author's usage of the term will be the same as that set forth
by Goggin (1949), as Alachua Tradition. For the more comprehensive
usage of Caldwoll, the term "macro-tradition" is suggested, as Northern
Macro-Tradition. This dichotomy of definition is utilized below in the
defining of the Deptford Phase within a Coastal Tradition. Quite
possibly the Coastal Tradition could be said to be sufficiently close
culturally to other coastal traditions to warrant the definition of
a Southeastern Archaic Coastal Macro-Tradition. Though this is
possible taxonomically, there is not enough empirical data to add
substance to the title.
Prchistorians also differ in the substantive definitions of
specific traditions and macro-traditions. These differences seem largely
to be the result of two categories of definitive criteria. Caldwell
in his Trend and Tradition in the Prehistory of the Eastein United
States (1958) bases his traditions and nacro-traditions largely on
material culture, especially ceramic typologies. Thus, he places the
Deptford Phase in the Southern Appalachian Macro-Tradition because of
the presence of paddle-malleated pottery. l e Wilmington Phase, which
follows the Deptford Phase in the sano region with the same subsistence
technology, is placed in the Northern Hacro-Tradition due to the
presence of cord-marked pottery. The pre-Deptford Archaic cultures are
lumped together, perhaps as a result of their being largely pre-ceramic.
When processual culture traits are utilized as a definitive basis,
as is done here in Chapters Ill and IV, the Peptford and Wilmington
Phases and certain of the pre-Deptford Archaic cultures are included
within a Coastal Tradition. These classificatory schema stress sub-
sistence adaptation rather than ceramic typology. Utilizing Caldwell's
modei, which places Deptford and Swift Creek in the same macro-
tradition, it seems very unlikely that the Deptford Phase coastal
peoples maintained the same "distinctive way of life" as the Swift
Creek Phase peoples of central Georgia. The value of tradition
definitions based on the theoretical methodology of cultural ecology
(Steward, 1954) and cultural materialism is readily apparent.
It should be noted that the study of trait traditions, as opposed
to total cultural traditions, is a valid and frequently employed
methodological tool. The most coupon trait studied as a tradition
in Southeastern prehistory is pottery. Perhaps the most comprehensive
example of such an approach is McMichael's description of Southeastern
Stamped Pottery (1960). Cultural phases within the same tradition may
contain ceramics from different traditions, and one phase may also
contain examples of ceramics from two different traditions. An
example of the latter is the Weeden Island Phase of western Florida
which contains both incised and stamped pottery.
A useful concept in the study of inter-phase cultural traits is
"horizon." Closely related is "horizon style." Horizon has been
defined by Wil.ley and Phillips as "a primary spatial continuity
represented by cultural traits and assemblages whose nature and mode
of occurrei:ce permit the assumption of a broad and rapid spread"
(1958: 33). Thus, as "tradition" is a diachronic integrative device
which is limited spatially, "horizon" is not so restricted and is better
suited to studies of diffused traits. The paddle-malleated pottery
'tradition" is more correctly studied as a horizon.
"Horizon style" differs from "horizon" in that it encompasses a
more limited temporal distribution and refers to a specific mode or
type (MicMichael, 1960: 9). "Horizon styles" are synchronic analytical
tools, which are useful in temporally equating phases since horizon styles
extend across contemporary phases and are the result of rapid diffusion,
suggesting contemporaneity. Thus, the presence of tetrapods, a South-
eastern horizon style, on pots from different phases can provide the
basis for suggesting contemporaneity of phases and culture contact.
Cultv rail-Ceogrphical Units
'illey and Phillips' definition of phase cited above stresses
restricted geogrdphical range. "Restricted," however, s a very arbitrary
measuremient and the concepts employed in discussing spatial zones need
to be clarified. Two types of spatial zones exist: those defined
geographically, by the researcher on the basis of spread of cultures,
:ilid those defiled by the natural environment. The former have been dealt
t.ith brief! by i':illey and Phillips (1958: 18-20). Their definitions,
h,:eveor, are ot fully adequate to deal with the empirical data, as
evidenced by the Deptford Phase. The discussion which follows amends
their position, accordingly.
The bulk of archeological data concerning New W'orld Indian
cuitiures is derived from the excavation of "sites," including ca'npsites,
villages, quarries, burial mounds, and ceremonial centers. Sites,
however, often exhibit occupations by different cultures, which are
distributed through various successive lengths of time. The smallest
manifestation representative of a cultural phase at a specific site is
referred to as a "component (McKern, 1939: 308; Willey and Phillips,
1958: 21-2)." Thus, one site may display several components, each
representative of the inhabitation of that site by a people of a
'Many of the sites discussed in Chapters II to VI are multi-
component ones, and exhibit a Deptford Phase occupation only as one
portion of their history. Comparisons of components and their relative
position in local or single site sequences provide the information for
the reconstruction of multiple phase sequences as well as descriptions
of specific phase development.
"Sites", mentioned above, are the most limited spatial zone and the
most difficult to define. A campsite might be several square yards,
while a ceremonial center may be as large as several square miles.
Suffice it to say that a site represents a unit of habitation, i.e
a community in the ethnological sense.
At a nulti-component site, the horizontal spread of discreet
components may not be exactly coterminous; boundaries of components are
often overlapping and of different size. Both horizontal and vertical
size of components give relative information of the length and
intensity of various phase occupations. Such information is difficult
to derive and more difficult to confidently interpret.
Willey and Phillips (1958: .1S-9) define a "locality" as the
space occupied by a single group. Complete cultural homogeneity is
postulated for the locality at any one time. Several sites may exist
within the locality, and the assumption is that personal interaction
within the locality is greater than interaction with persons outside
of the locality. Geographical barriers help to define localities. The
Deptford sites on Cumberland Island, Georgia, described below, occupy
a locality. Likewise, the sites in the vicinity of Savannah, Georgia,
clustered in the delta of the Savannah River, also represent a locality
or zone of inter-site interaction.
Since localities represent quite homogeneous cultural units, it
follows that such units evolve somewhat differently from other such
units. It will be demonstrated below that such a process did occur
within the overall Deptford Phase evolution at several localities.
The next largest spatial unit discussed by Willey and Phillips
is the "region." They view the "region" as a result of the focus
of archeological investigations on an arbitrary geographical zone.
Examples are the Glades region a(id the Northern St. Johns region, both
in Florida (Cougin, 1949; 1952). Rather than being the result of
investigation of a geographical area, archeological research is usually
centered on a specific phasess, which tends to occupy a spcciFic
geographical, or more correctly, natural zone(s). Regions are not based
on archeological whim, but they are the result of a phis's1 adaptation
to a specific natural inhabitat through which the phase radiates. That
thi aicleologist may or *ay not choose to study a natural area is
This author would d redefine "region" simply as the geographical zone
occupied b:. a phase or tradition. The latter case would occur when
successive phases, forming a tradition through time' evolve in the
same g_-ograuhI cat zone, the result of similar environmental adaptations.
The Alachua Tradition, comprised of two phases, occupied a distinct
region in north-central Florida (Milanich, 1968). In the same way, the
Safety Harbor Thase within the Gulf Tradition occupied a clearly
definable region on the central Florida Gulf coast. Regions are
based on geographical ranges which are in turn the result of exploita-
tion of one or more natural environmental zones. As phases evolve
and adapt to new natural or cultural environments or as phases develop
into new phases with new adaptive techniques, the corresponding regions
also change. There is a one-to-one correspondence between phase and
This revised region concept seems better suited to processual
archeological studies. The study of changing regions gives important
information on the corresponding changing subsistence patterns of
cultural phases. The emphasis is on cultural interaction with the
environment rather than on defining culture in terms of environment.
Though the regions of the Deptford Phase and the Ft. Walton Phase in
nortliwest Florida overlap geographically, the two cultures utilized
the same environment in very different manners. The significance
lies not in the fact of overlapping region. Rather the emphasis should
be on utilizing the same bicue, but with different techniques.
The Deptford Phase evolved into two different phases. 'he region
of the first, the Wilmington Phase, lies along the Atlantic Coast of
Georgia and South Carolina. Thet of the second, the Swift Creek Phase
!Although this statement applies generally to the Southeast, nauy
exceptions exist in the New World, e.g., the prehistoric Pueblo and
Navajo cultures of the Southwest. The Pueblo and Navajo people, however,
did exploit different ecological (and overlapping) niches in the same
region. At the state level of social organization, different phases
may also occupy the same region. In this case, the phases may occupy
the same ecological niches.
within the Gulf Tradition, was along the Flc rida Gulf coast. These
new phases, their regions, and subsistence technology were quite
distinct from one another. In the study of the evolution of the Dept-
ford Phase that follows in Chapter IV, the factors within the Coastal
Tradition responsible for this division are evident. In order to
study them in a manner consistent with the model proposed here, the
term "sub-region" is utilized to refer to the geographical zones
within the Deptford region which later develop into new regions. Thus,
within the region of the Deptford Phase, two sub-regions are
recognizable, an Atlantic sub-region and a Gulf sub-region.
The largest spatial zone proposed by I'illcy and Phillips is the
"area," which approaches the culture area of the cultural anthropologist.
Here "area" will be defined as that spatial zone in .itich intra-phase
interaction is greater than interaction with phases outside the area.
Such a definition implies that diffusion is greater among the phases
within the area than it is with phases outside that area.
During the temporal range of Lhe Deptford Phase the Southeastern
United States interacted as an area. Boundaries of the Southeast area
ranged firrm south-ceni-rri Florida north to the Ohio River and beyond;
from the !';:issippi River and beyond in the west to the Atlantic
Ocean. The Appalachian Mountains of North Carolina and Virginia
probably wore excepted. With the end of the Adena-Hopewell climax this
ar;e became more restricted only to grow even larger with the Miss-
issipian cl.imax. As regions and phases change, so do areas.
in rummnary, the concepts of site, locality, sub-region, region,
and area are used to refer to spatial zones defined by cultural-
geographical parameters which are actually the result of phase adaptation
to selected environments. Groups occupy sites; interacting groups
inhabit localities; phases occupy regions. As phases evolve sub-
regions often become evident, and as new phases emerge new regions also
Natural Environment Units
Larson (1969) in his excellent work on Southeastern coastal
subsistence during the late prehistoric period employs two categories
for detailed analysis of natural environmental zones. These are
"sector" and "section" (Larson, 1969: 5). p
Larson's definitions, however, are too narrow and do not reflect
the adaptive exploitation patterns of the Deptford Phase. According to
his terminology, a society can subsist by exploiting only one "sector".
While this is possible in the case of certain sectors, it is not true
when Larson's Pine Barrens is considered. Likewise, Larson states that
a society can subsist by exploiting only one "section" during a season.
This need not be true, either. It is evident that many societies or
phases were adapted to exploit more than one sector and a variety of
sections. This is especially evident in the case of the Deptford Phase,
whose peoples exploited different environmental zones both seasonally and
Thus, while retaining Larson's approach to the study of culture-
environment interaction, it is necessary to revise his terminology.
This author will employ biomee" and "biotope" or "micro-environment" to
refer to the same types of units used by Larson. A biomee" is a large
climatic-environmental zone best defined by example, e.g., the Pine
Barrens biome or the Coastal biome of the Southeastern Coastal Plain.
Biomes are made up of localized environmental situations, or
"biotopes." Examples of biotopes wi thin the Coastal biome are the
narsh and lagoon, the beach strand; and the live cak strand. The
bmioes and biotopes within the Depcford region are described in detail
in Chapter III.
Cul turologica! Theorems
The definitions and discussions of terms outlined above have
their basis in the axioms and theorems of cultural anthropology. Wherever
applicable in this paper, an effort will be made to deductively apply
these theorems to the Deptford Phase. What is the relationship of
archeology and cultural anthropology if the two do not share the sase
Tilis study of the Dcptford Phase is a description of the specific
evolution cf that phase, the adaptation of that phase to its e'niro:ment
and its efforts to increase its adaptive sp:ciali>.i tans, thus
garnering higher levels of energy for subsistence. (Us:age here of
"'specific evolution" follows that of Sahlins and Service in their ,'ork
on Evolution and Culture (1960).)
Sablins and Service ([150) w th Kaplan and larding also set forth
several very interesting principles concerning the evolution of
cultures nand the changing relationships between a phase and its region
Ithe author's terminology). ilany of the interpretations of phase
sequeinces, especially tlho.e po:.tulating replacem:ent of populations
rather than evolution of' phass, caln be disproved on the basis or these
sane principles. 'h- pr inpciples hare been calie:i by Sah:iins and Service:
(J) the Priinciplie of Adaptation and Stability; and (2) the Law of
The Principle of Adaptation and Stability focuses on two processes
intricately involved in cultural evolution: creative or adaptive
ehangcs, and a tendency toward maintaining a status uuo (stability
or conservation). At first glance these aspects of cultural change
seem. paradoxical. Closer scrutiny, however, sho-s them to be extremely
-co patible concepts. In fact, many cultures may adapt in order to
Adaptation both to the natural environment and to other cultures
occurs within a given phase. The natural environment acts as a stimulus
for subsistence pattern changes which affect, in turn, other aspects
of cuituce. Adaptation to stimuli presented by contact with other
~:l!tures leads to societal and ideological change; which in their
;in bring about economic changes. These changes cause further change
in societal and ideological cultural processes. Adaptation is the
e:isence of specific evolution.
Ad:pt-ation, by virtue of creating sp.:cc.alization, involves a need
Cor stability. Generally, once adapted to m:axim:um exploitation of
a biome or biomncs, a culture's many manifestations act to conserve one
another -- ideology seeks to maintain the technology as is, and vice
vej:sa. However, .'hen n:w stirmli act on a cultu-e, that culture must
act on the sti::uli to mnint.,in its adaptive specializations. thii is
the pri'.ess of "changing a little to avoid changing a lot."
The Pi.ncl.ple of Adaptatioa anid Stalijlity i is illustrated .in the
trnadi.tLon concept. Mi iy tradi.ticns, such as the Coastal Tradition,
span sclern! tlo-.usand years, always '.i,:int.ining tlh s-:ce "basic iway of
lifioe." Froa the acc:"an'l t.iin oif sal ch1 nh.:s a' phases emerge.
Traditions end wiher patterns of adaptive mechanisms are no longer maintained.
This occurs as a result of ei-ther the introduction of now exploitive
patterns (as in the Deptford Gulf sub-region), or as the result of
culture contact intense enough to produce significant societal and
ideological changes. These societal and ideological processes cannot
reinforce the technological processes, thus leading to subsistence
The Law of Cultural Dominance states that in a given environment
biomee), the culture (phase) which most effectively exploits the
energy resources in that biome will spread at the expense of less
effective phases (Sahlinas and Service, 1960). In a region occupied
by tio phases, the less adapted phase cannot displace the better
adapted phase, and, eventually, is itself displaced (the relative
levels of adaptation measured by the amounts of energy derived frcm
exploitation of the environment). hnis leads in nart to a one-to-one
correspondence between phase and region.
assuming that the subsistence technology of the Deptford Phase was
tihe best adapted to its region (at that timl'e), the introduction of
other phases .with subsistence techniques less adapted to that region
would not be successful. Phases intruding into the Deptford region
couJd not permanently displace the Deptford Phases unless a superior
exploitive technology were involved. In the discussion of the Coastal
Tradition blow, it will become evident that there was no major
subsistence charge during the tradition and that the coastal sub-
sistence pattern was utilized exclusively. Thus, it is very unlikely
that the Coastal Tradition phase sequcece -- pre-Deptford, Deptford,
Wili'ington -- is the result of migration of peoples into the Deptford
region. The Principle of Adaptation and Stability and the Law of
Cultural Darn ;i.ra'e both suggest evolutiion rather than migration.
In summary it is apparent (and inevitable) that the Deptford
Phase and the specific evolution of that phase are not unique. They
are only one case within the realm of all Culture, and, consequently,
follow the laws attributed to Culture.
Tie Deptford Phase has the distinction of being the Southeastern
aboriginal phase with the widest geographical distribution, but which
is least known archeologically. A large amount of field research has
directly or indirectly investigated sites with Deptford components,
producing ceramic typologies, za great deal of misunderstanding, and
little other information besides relative daces and a few radio-
carbon dates. Past areal syntheses which have inclulJed the Deptford
Phase have interpreted its position either on the basis of ceramic
typology or on its chronological placemerno lative to other South-
As a consequence of this tendency to interpret on the basis of
ceramic ur temporal analogy, various arcicologists have described the4
Deptford Phase as Archaic (Willey and Phillips, 1958: 119; Goggin,
1949: 22), Formative (Ford, 1966: fig. 1; 1969), and Early Woodland
(Willuy, 1966: 286-37). Bullen (1965: 308) refers to the Deptford
Phase as horticultural, and Caldwell places the phase within his
Southern Appalachian Tradition (1958: 49). Much confusion reigns con-
cerning the relation of McLcod (,Alabama) and CartorsvJlle (Georgia)
ceramics to the Deptford Phasa, with flauchope (1966: dS) and Biillen
(1361: fig. I) ,-uati.ng Carte-svi lie and Deptford. Again these
misinterpretations are the result of trait counting, based on ceramic
comparisons. The large number of incorrect interpretations concerning
the status of the De:-tford Phase and the variety of these interpreta-
tions decree that the process of elimination be used in reconstructing
the Deptford Phase; in other words, it will be necessary to point
out what is not Deptford in order to make clear what is Deptford.
Larson, discussing the late prehistoric cultures on the South-
easterm coastal plain, has presented an eloquent plea for the type of
archeological approach which will be utilized here. He concludes:
It makes for little understanding of aboriginal
culture in the Southeast to discuss the stylistic
unity between the Gulf Coast of northwestern
Florida and the Caddo area when there is no
appreciation of the fact that designs on pots ray
be the only thing that the two areas hold in conmiron..
It is necessary to approach the coastal cultures
in a different manner if 'we are to obtain significant
answers to questions of cultural adaptation in the
region. The traditional approach will not provide
them (1969: 324-25).
The specific types of informatLin on the Deptford Phase which the
author seeks to produce include the following: (1) delineation of
the geographical region of the phase; (2) detailed descriptions of
natural biotopes of the region and cf cultural adaptations to these
biotopes, including subsistence technology, settlement patterns, and
social organization; (3) descriptions of material culture; (4)
chronological placement of the phase; (5) evolution of the phase and
region into new phases and regions; (6) relation of the phase to a
Coastal Tradition including a definition of the tradition; and, (7)
relation of the phase to the Southeastern area.
The knowledgeable reader certainly has all ready noted the author's
reluctance to use (or define) such terms as Archaic, Formative, Early
Woodland, and Middle Woodland. This is intentional. Their meanings
have become too obscured, at times referring to either periods, stages,
or subsistence patterns.1 Unless quoting, they will be avoided, or
their meaning specifically noted. Chapter VII comments on the uses of
taxonomic terminology. The pre-Deptford cultures of the Coastal region
(which may be divisible into discreet phases with additional research)
will be referred to as "pre-Deptford cultures."
The Deptford Phase derives its name from the Deptford site in
Chatham County, Georgia. Deptford is one of a number of sites, both pre-
Deptford and Wilmington and later, in the vicinity of Savannah which were
excavated under the auspices of the Works Projects Administration of
Georgia in the 1930's.
The first excavations at the Deptford site were conducted by Antonio
J. Waring, Jr., and Preston Holder in the Pall of 1957. These were
followed by excavations by Catherine McCann in 1940. Neither of these
excavations have been published in their entirety. A nmaus--ript by
Caldwell and McCann (n.d.) summarizing the Waring and Holder tests and
reporting McCann's excavations is in existence (writecn after 1942), with
copies being passed hand-to-hand among Southeastern prehistorians.
Excavations by Holder carried out in the late 1930's in Glynn County,
Georgia, re:iain largely unpublished though a manuscript has been
Ceptford ceramic typologies and summaries have been published by
Caldwell and Waring (1939; a; b; c), Soars andr. Griffii' (1950), and
ISee Mason (1970) for a discussion of the conru.sion surrounding the
te-' i "Middle W!'oodland."
Caldweli and McCann (1941). Other early work includes a very brief
account of Holder's work on St. Simons Island off the Georgia coast
(1938b). Recently the Waring Papers, edited by Stephen Williams (1968),
have been published and include two formerly unpublished manuscripts.
These are a summary of Deptford ceramics by Waring and Holder (1968)
written about 1940, and a synthesis of the culture sequence at the
mouth of the Savannah River written by Waring (1968) in 1955.
Much of the information derived from the early excavations along
the Georgia coast was diffused by word of mouth at the Southeastern
Archaeological Conference meetings. Consequently, the presence of
Deptford ceramics began to be recognized throughout much of the South-
eastern coastal plain. In South Carolina, James B. Griffin reported
Deptford pottery from the Lake Charleston site near Beaufort (1943).
Griffin (1915) also reported Deptford potsherds in a collection From
the now famous Thom's Creek site south of Columbia, South Carolina, in
the Piied'oont area. However, this author believes that those ceramics
(mostly linear check stamped) are related to the Piedmont cultures
(as Mossy Oak or Cartersville) rather than the Deptfcrd Phase.
To the south i'Wiley ard Woodbury recognized the Deptford Phase
within the aberigi.'ni culture sequence of northwest Florida (1942).
!'i.iley's interest in the west coast of Florida led to more work in that
area culminating in his Archeology of the Florida Gulf Coast (1949). This
volume contained the most complete description of the Deptford Phase
yet published, a listing of sites along the Florida Gulf coast,
ceramic typologies, and a brief discussion and list of culture traits
(353-61, 507). Willey viewed the Deptford peoples as being shellfish
collectors with band-type social organization. He also recognized
and reported the "life potentialities" (possible adaptations to
environment) of the Deptford Region in Florida. Based on the early
work in Florida and elsewhere, Willey and several of his colleagues
anticipated the concept .of a Coastal Tradition (and, perhaps, a
Piedmont Tradition) noting:
The suggestion made by William H. Sears (1948)
that Archaic is merely Woodland without pottery
arose from the same observation that we have
made above, to wit, that the addition of pottery
had little if any effect on the configuration of
the culture (Early Woodland) as a whole. W'e are
simply turning his statement around to read:
Early Woodland is merely Archaic with pottery.
Fairbanks was expressing the same idea in pointing
out that Early Woodland cultures in the Southeast
are in effect merely transit onal from Archaic to
Middle Woodland and no change in basis economy
was involved (1949, p. 59) (Willey and Phillips,
The 1930's and 1940's research into the Deptford Phase provided
the information for the chapter on Georgia and South Carolina
(Caldwell, 1952), which was presented in the first major synthesis of
the Southeastern area, Archeology of Eastern United Sti;'r, edited
by James B. Griffin (1952). Caldwell observed (p. 316) that "Freml
the foregoing (his discussion of Deptford) it will be seen that we
know practically nothing about . the Deptford Period, ard some
definite knowledge must await excavation of one or more 'pure'
Deptford Period sites."
Such a lack of knowledge, however, did not prevent Cajidwell from
including the Dept-ord Phase within his nr.ore recent Southeastern
synthesis (1958). In Trend and Tradition he views Deptford as a
member of the Southern Aupalachian Macro-Tradition, as are the S'wift
Creek and Mossy Oak Phales. On the basis of subsistence technology,
such a grouping is totally invalid. Caldwell, however, does view
the Southern Appalachian Macro-Tradition as the continuation of a
"generalized Archaic subsistence pattern," one which differs From the
Middle Eastern Macro-Tradition, whose economy is based on collection
of seeds and nuts (1958: 35). This interpretation, that both the
Piedmont-Fall Line cultures and the coastal cultures are temporal
extensions of the preceramic phases in those respective regions, agrees
with Willey and Phillips, Sears, and Fairbanks, as cited above.
Caldwell's error is in grouping the preceramic phases of the coast and
the piedmont into one nebulous Archaic Macro-Tradition. Caldwell
(1958: 14) does offer the best description of the Coastal biome
previous to Larson's (1969), and he does postulate a transhumant
economic pattern for the coastal cultures.
Other less general syntheses concerning the Deptford Phase include
Bulien's articles on the Florida Transitional period (another nebulous
Lerm based on ceramic interpretations) (1959; 1969). The latter
report.on Sunday Bluff, Florida, provides important data on the
excavation of a non-coastal Deptford component site.
Perhaps the best discussion of the ceramics of the Deptford
Phase in northeast Florida and their significance as indicators of
culture contact is found in Sears' report on the Willey Browne sites
in Duval County (1957). Sears, in his Tucker Site report in Franklin
County on the west Florida coast, discusses the ceramic and cultural
affinities of the Deptford Phase, citing ties to contemporary
Tchefu~cte, l' ops-.rel and .acrhi Florida cultures (1963). La arus
(.1965b) also provides important information, including a radiocarbon
date, for Doptford in northwest Florida.
The best information on the relation of the Deptford ceramics to
the Southeastern area ceramic horizons is presented by Fairbanks- (1952)
and by McMichael (1960). McMichael's The Anatomy of a Tradition: A
Study of Southeastern Stamped Pottery is the most comprehensive
analysis of the origins and diffusion of paddle malleated pottery in
One interesting aspect of the evolution of the Gulf sub-region
is the introduction of a complex burial and religious ritual.
Descriptions of this intrusion and the subsequent adaptations within
the Deptford Phase are discussed in Chapters IV and VI of this paper.
McMichael (1960) briefly points out the existence of some of the
introduced burial traits. Sears describes the traits and interprets
them as an extension of the Hopewell climax into Florida (1963). IHe
organizes the traits into the Yent (earlier) and Greenville (later)
Complexes. Another site within the Yent Co-iplex is the Oakland Mound
in northwest Florida excavated by Morrell (1960). The work of Kellar,
McMichael, and Kelley at the important Mandeville Shle in Clay County,
Georgia, also provides information on the origins, form, and spread of
the Yent Complex and its relation to the Deptford Phase (n.d.; 1962).
In 1960 the central topic of the 17th annual meeting of the South-
eastern Archeological Conference was the Deptford Phase. The proceedings
of this conference, a transcribed tape of an open session chaired by
i'taring, were published in 1966 (Southeastern Archeological Conference,
1966) and reveal several small but informative bits of information
from various archeologists. Williams' succinct "Epilogue" in the
Waring Paers (1968) sunarizes our knowledge of Deptford in less than
a column and a half of text.
Numerous other references to the Deptford Phase, especially in
regard to ceramics, can be found in the literature. Most are found in
site reports which simply list the number of Deptford sherds excavated
and their provenience within the site. These are not discussed
in this brief history of previous phase research.
Recent field research into the Deptford Phase, carried out in the
Summer, 1970, on Cumberland Island, Georgia, is reported in Chapter II
of this study. Excavations on St. Catherine's Island, Georgia, have
recently been carried out by the University of Georgia under the
direction of Joseph ahd Caldwell. It is expected that the results of
those excavations will be reported in a Masteo's thesis by Betty
Anderson of the University of Georgia. Miss Anderson was gracious
enough to provide this author with information from her report, adding
important data to this reconstruction of tihe Deptford Phase. Radio-
carbon dates gathered from the St. Catherine's research were reported
by Cald.lel at the Southeastern Archeological Conference meeting in
South Carolina in the Fall of 1970. These are incorporated into
It is hoped that the study presented here, which brings together
much of the information known about tlhe Deptford Phase, ill mark the
beginning of additional research into the Deptford culture. As Willey
end Phillips (1958) have pointed out, the first step in archeological
research is the intensive study of specific units. Once this is
completed, areal syntheses and detailed, problem-oriented research of
selected cuitirail processes are possible. This paper is meant as a
Chapter II, which follows, presents recently collected archeological
data on the Deptford Phase. This information, referred to frequently
in subsequent chapters, is the result of research specifically oriented
toward fulfilling the research requisites of Willey and Phillips (1958)
which were outlined above.
EXCAVATIONS ON CUMBERLAND ISLAND, GEORGIA
During July and August, 1970, two Deptford Phase sites on
Cumberland Island, Georgia, were excavated under the direction of this
author.l The research was directed toward securing two types of
information: (1) the evolutionary relationships between the Dept-
ford Phase and other Coastal Tradition phases; and (2) descriptive
and interpretive data about Deptford Phase cultural processes,
including subsistence and settlement patterns.
Since the chronological placement of the Deptford Phase has been
::ell established, the research strategy was centered on extensive
excavations rather than intensive excavations. Test units were dug to
locate occupational features which would provide the types of informa-
tion desired. Once location of these features was ascertained, the
test units were greatly expanded, revealing large horizontal tracts.
This method is well suited to providing synchronic, functional know-
ledge of "everyday" aboriginal life. Intensive excavations, on the
'A preliminary paper summarizing these excavations was presented at the
Southeastern Archeological Conference annual meeting in Colu.mbia, South
Carolina, 1970 (Milanich, 1971b). Final analysis of the data has
changed several of the tentative hypotheses set forth in that paper.
A grant from rhe National Science Foundation funded the Cumberland
other hand, are better suited toward providing diachronic data, such
as changing ceramic sequences and temporal relationships among cultures.
In addition to the excavation of the two Deptford sites, surface
collections were made at other sites on the island. The presumed
location of the historic Timucuan village of Tacatacuru, site of the
Spanish San Pedro de Mocamo mission, was found. Reports on the mission
site have been published elsewhere (Milanich 1971c; 1971d). The
pertinence of the Timucuan material to the Coastal Tradition is
discussed in Chapter VI.
Cumberland Island is the southernmost barrier island along the
Georgia coast (Figure 1). The island is approximately eighteen
miles long and varies in width from less than one to more than three
miles. The first site excavated, the Table Point site, is located on
the northwest tip of Table Point, a Finger of land which juts out into
the tidal marshes on the northerly, inland side of the island. Exact
location is 30 52' 7" North latitude, 810 27' 1" 1 ,st longitude
(United States Department of Commerce, Nautical Map 841-SC, 1969).
Stafford North, the second site, is located 4.2 miles south of
Table Point, also bordering tidal marshes and situated in a live oak
hammock (Figure 1). At the two sites the environmental situation is
like that found throughout the Coastal home of the Deptford Region,
allowing exploitation of sea, beach, ocean, island, land, and marsh
.rd lagoon biotopes [see Chapter Ill, "Geographical and Natural Settings").
Extensive shell middens 0.5 feet to 1.0 feet thick are found at
both sites. Horizontal and vertical stratigraphy were both apparent.
MjJden deposition tended to be in separate circular piles, roughly
20-30 feet in diameter and scattered randomly in lines parallel to the
Figure 1. Seartga Sea Islands and Cum'berland
high tide line. Each pile probably represents refuse from one family
during the occupation of a single house structure. Placement of the
houses was near to the marsh in a linear arrangement rather than in
a circular or other complex community form. This depositional
pattern contrasts with that of the St. Johns Tradition to the south
which is characterized by thick (at times more than twenty feet) shell
heaps which form long linear middens rather than separate piles. These
differences reflect either a smaller population within the Deptford
Phase or, more likely, shorter occupations at specific site locations.
Both sites displayed a confusing mixture of ceramic pastes, includ-
ing fiber-tempering, crushed sherd-tempering, grit-tempering (large
amounts of quartz from sand size to inclusions 2 mm. in diameter),
and contorted sherd-tempering, as well as combinations of the above.
A relatively small portion of shards at the sites (large compared to
reports from other Deptford sites) displayed a red exterior slip.
Surface treatment included check, simple, linear check, geometric,
complicated, and cord-wrapped-paddle stamping (the most frequent forms)
with some incising and brushing also occurring. Check stamping
predominated. The thickness of the cord impressions on the cordmarked
sherds does not resemble either the Wilmington Heavy Cord Marked or
Savannah Fine Cord Marked types. Rather they are like cord markings
dating from the first millennium B.C. and which are found throughout much
of the Southeast, and which range in thickness from 1 to 2 mm. in
Relative frequencies of sherd types varied from shell pile to
shell pile, constituting a horizontal stratigraphy. Other recovered
artifacts included food bones, shell tools, and a very small amount of
bone artifacts. Only two stone artifacts were found, one a chert chip
and the other an agatized coral hammerstone. Pits, postholes, and
other evidence of structures, however, were numerous and extremely
In general on the Georgia coast (based on the work of Caldwell,
Waring, Holder, and others), it is thought that: (1) the pre-Deptford
coastal cultures are associated with fiber-tempered pottery; (2) the
Deptford Phase is associated with paddle stamped ceramics; and,
(3) the Wilmington Phase is associated with sherd-tempered, heavy cord
marked pottery. This implies that relative frequencies of the three
ceramic series within a site can give information on relative ages of
the respective locations. When horizontal stratigraphy is present, as
at Cumberland Island, such a simplistic assumption may give false
information, except when dealing with discreet (in time) units, such
as features or individual shell heaps. Recognition of this fact is
important in interpreting the Cumberland materials where both
horizontal stratigraphy and features constructed of earlier shell deposits
The Stafford North site, composed entirely of scattered shell
piles, stretches about 435 yards north-south along a tidal marsh
(Figure 2). At the extreme southern end of the site is a small stream
which drains an inland swanp and empties into a tidal creek. The level
113H ...... ...
of the stream fluctuates with the tide only near its mouth where it is
below the level of the high tide. Consequently, the stream water
back away from the marsh is fresh. During low tide the water at the
stream's mouth is also fresh. The north end of the site lies about
150 yards north of a similar stream. Undoubtedly these streams drew the
aboriginal inhabitants to the site.
Just south of the site at the Stafford Estate is a large tidal
mud flat. Thus, marsh, lagoon, island, inland stream, and mud flats
were all readily accessible to the site's occupants.
While on the island, it was noted that wherever there were shell
middens, both exposed or buried, little scrub undergrowth was
present. Low grasses invaded the sites creating park-like areas with
live oak tree covering. This flora is quite distinct from the
secondary growth in fields originally cleared for cultivation in the
19th century. The overgrown fields are covered with dense pine
forests and palmetto. These floral differences made location of shell
midden sites easy, especially since the middens were often exposed by
the uplifting action of tree roots. Other midden locations (separate
piles) were located by probing in the park-like grassy clearings with
metal rods. Buried shell middens were located by this latter method.
At the Stafford North site north-south and east-west axes were
staked cut to form the basis of a grid system. Measurements within
this grid were made in feet measured north and east from an arbitrary
point southwest of the site. Originally it was planned to sample the
length of the site, placing excavation units at selected locations off
the axes. However, the first 5 by 10 foot test unit, placed just
north of the southern end of the site, intersected a living floor,
making a more extensive excavation necessary. Four 5 by 10 foot units
(called henceforth "squarcs") were placed there. Three more squares
were placed south of the living floor excavation, two intersecting
shell piles and one between piles as a control (see Figure 2). The
northern three quarters of the site was untested.
nTe S by 10 foot squares were referred to in the field by the
number of their southwest corner grid stake, as Square 360N (north)
S00E (east). For interpretation in this paper, however, the
excavation units are referred to as Test I, Test II, etc., as shown
in Figure 2.
Vertical control was maintained with a transit. Both the grid
system and the datum plane were tied in with a concrete surveying marker
at the east edge of the site. This information is on file at the
Florida State Museum in Gainesville. All sifting of excavated dirt
and shell, unless noted, was done with a mechanical shaker screen
with 3/4 by 3/8 inch mesh expanded metal screening. Digging was carried
cut by natural strata rather than arbitrary levels. At times, however,
thick shell zones were divided into arbitrary, usually six inch, levels
to test for intra-stratum changes in artifact frequencies. None were
apparent, suggesting the separate shell piles were deposited over a
brief period of time as has been expected.
The southernmost square excavated, Test I (Square 113N, 530E), was
approximately 150 feet from the marsh and 200 feet north of the southern
creek. Test I w's placed -o intersect a circular shell pile which
appeared from the surface (and confirmed by excavation) to be 25 feet
in diameter The shell zone was excavated in two arbitrary levels.
Miatrix of the shell was a sandy, dark brown humus. Below the shell,
whose top extended to the ground surface, was a mottled grey-brown
old humus zone which overlay buff to white sterile sand. Artifacts
from the Test I strata are shown in Table 1. Table 1 also shows
numbers and weight of food bones. No stcne, bone, or shell artifacts
were present. Sherds smaller than one-half inch square were counted,
weighed, and discarded.
If the assumption is true that the shell represents debris from
one family during a restricted period of time, the pottery frequencies
from the two levels of shell should be the same. Even with the very
small samples, this seems to be roughly the case. Again assuming a
restricted period of time for the midden deposition, it is evident that
a variety of pastes were employed simultaneously at the site, or
that trade was very common.
The majority ceramic series is Deptford with a mixture of grit
and sand-tempering. Deptford plain and paddle maileated ceramics
account for 75.0 per cent of the total from Test I. The one Deptford
Geometric Stamped sherd displayed a diamond design similar to
specimens illustrated by Waring and Holder (1968: Fig. 43 b, d-f).
Both shord-tempered, cord marked sherds and fiber-tempered
plain shards occurred in equal amounts in Test I, both in the shell and
in the underlying old humus. Root action has no doubt moved some
sherds and shell into the old humus. Such artifacts were present in
root molds at both sites. It is difficult to assess the sherd-
tempered and the sand-Tempered potsherds. The two paste types seem
to be in use at the same time as fiber-tempering, and, like fiber-
itemering, chey are in use throughout the temporal range of both sites.
N) r, 0
i 00 N co
S pedam rS
0 proj-iJs a
-p.o3 [ m
NM 00 NM
[ -C 101
d ai.a -'r.12 C) o co0
Li) C) -
'H Li) 0C
NC 00 N
. tt o
Snr -Nl If
0-i 1- r
The confused Test I ceramic picture becomes somewhat clearer when
compared to the ceramic frequencies from Tests II and IV below. Ihese
comparisons suggest that within the Deptford Phase there is a decrease
in fiber- and sand-tempering and an increase in grit-sand-tempering.
With the increased use of paddle malleating, there may have been less
of a need for cleaning the clay and then adding tempering as was done
previously with hand-modeled pottery. Such a change marks the
transition from the fiber-tempered, pottery-making coastal phases to
the Deptford Phase.
Shell recovered from Test I and all other excavations at both
sites was between 90 and 95 per cent oyster Crassostrea virginica.
Other univalves and bivalves included clam Mercenaria mercenaria,
whelk Busycon perversum, channeled whelk B. canaliculatum, ribbed
mussel Modiolu3 demissa, shark eye Polinices duplicatus, marsh periwinkle
Littcrina irrorata, mud snail Nassorius obsoletus, lettered olive Oliva
sayana, and giant Atlantic cockle Dinocardium robustum. The unchanging
shell frequencies throughout the occupational periods of both sites
suggest few subsistence or climatic changes.
All the identified food bone from the Test I excavation was
white-tailed deer Odocoileus virginianus. The subsistence pattern of
the Deptford Phase is dealt with in detail in Chapters [II and VI.
Test II, Square 270N, 535E, was excavated to test artifact
content in a non-shell area and to see if structure patterns were present
beside the separate shell piles. The dirt was not screened, although
care was taken to examine the spoil during excavation for artifacts.
No artifacts were recovered.
A profile of Test II showed a dark grey modern humus and duff layer
overlying a zone of grey humus and scattered shell. The latter is
probably temporally coeval with the site's occupation. Under this
zone was a brownish-grey humus with very little shell, probably an
old ground surface. As in all other excavation units, the old humus
was underlain by sterile sand. Intense occupations, then, seem to have
been restricted to the shell-piles themselves.
Square 360N, 500E, Test III, was placed to intersect the edge of
a circular shell pile. The west wall profile of the 5 by 10 foot
excavation unit is shown in Figure 3. The slope of the edge of the
shell pile is clearly visible in the profile drawing. It is signifi-
cant that an area (Zone C) beside the shell pile and on the same
level as the shell contained historic 19th century glass and hardware.
Zone C was covered with a layer of grey humus. This humus must have
been deposited within the last 150 years. Probably agriculture in
adjacent fields hastened the erosion of topsoil which was washed toward
the marsh, covering the site. Elevation of the site is 5.0 to 7.0 feet
above mean sea level, while the fields are about 10 feet above mean
Table 2 lists potsherds and food bone recovered from Test III.
Frequencies and types contrast somewhat from Test I, although sample
sizes are extremely small. Most significant is the increase ii: semi-
fiber-tempered ware and the decrease in the amount of Deptford
_~~_ _U~~PL VMm~
sluuu~-l~---rro~sraarw I- - ~Ba~-~o----~---~ICIIIII1
fi -[qllo ^-i r Ni
u IId ",
p"" I Il
UTi~d j. K M
I- r4 r cL O t
* dmai-puas j ,
0-I 0 .-
S .o to
o ,o o U
O N o N C
in (M L" cM Crq
-ito NO 4"
in bt CO
M .M U
to *t to
series pottery. These differences suggest that the Test III shell pile is
earlier than that of Test I. The presence of sherd-tempered plain and
of cord marked ware indicates that early within the Deptford Phase
these techniques of pottery manufacturing were known. The cord marked
sherds are both sand-tempered and sand- and sherd-tempered, and they
are, perhaps, trade sherds. Other cord-marked sherds, however, are
grit-tempered and are identical in paste with the Deptford ceramics.
Cord marking is a minority form of surface treatment with the Deptford
Food bone from Test III was scarce. The two identified species
were deer and halmnerhead shark Sphyrnidae.
The area of most intensive excavation is Test IV, comprised of
Squares 510N, 490E; SOON, 490E; 500N, 495E; and 50N, 500E. Lack of
time prohibited a more extensive investigation of this area. The
features, stratigraphy, and material recovered together suggest
a slightly concave living floor possibly under an open, oval structure.
Stratigraphy within Test IV was very similar to that encountered
in Test III. A very thin modern duff zone unscreenedd) overlay a
mixed shell and humus occupation zone (Zone B), which filled a slightly
concave depression. Shell within this zone was very lensed. Under
portions of Zone B was a thin (0.2 foot) layer of crushed shell
(Zone C), which lined the depression. Figure 4, a top view of Test IV,
shows the outline of the depression, the crushed shell living floor,
three postholes, and the one pit encountered. A profile (mirror image)
West Profile-mirror image
1.9' below sea leve-
- --- depression
- -. -
1 inch:4 feet
Numbers are features.
Fig:ire 4. Stafford I arth Test IV.
through the depression is also shown in Figure 4. Again, the old humus
graded into buff sterile sand.
It is possible that the depression, which is longitudinal, was
naturally formed. Other similar though more linear "ditches" are
evident nearby on the site's surface, the result of water run-off into
the marsh. If the depression were the result of similar erosive action,
it seems that a living floor would not have accumulated in the ditch.
Also, no water-scrting of the underlying humic zone was apparent.
The alternative explanation for the depression and living floor is
that an oval area was cleared, creating a concave surface on which to
work or live. The underlying humic zone was thinner than corresponding
zones encountered in Tests I and III.
Table 3 lists artifacts from the four squares. By far, the
largest number of sherds and food bones came from the crushed shell
living floor. The nature of the artifacts and the shell suggests
heavy traffic over the area. Twenty-two Deptford Check Stamped sherds
from one vessel were cleared in place in the crushed shell zone. They
were reconstructed into about one quarter of a short, squat, open-
mouthed pot with rounded bottom, flaring rim, and a simple rounded lip.
Dimensions are estimated as: height, 45 cm; mouth diameter, 50 cm.
Also recovered from the living floor was one of the two stone artifacts
from the Cumberland excavations, a barrel-shaped, agatized coral
hammerstone with a height of 5 cm, and a diameter of 4.9 cm.
Most noticeable in Table 3 is the relatively small amount of fiber-
tcmpered pottery compared to Tests I and III. Also, Deptford series
types have increased slightly over that of Test i to 78.7 per cent,
and for the first tine Deptford Simple Stamped sherds are present.
Taole J. iest IV i'otsherd d.U IoVU U. C IVU.tS
"5 0 .
top she] ] 2 l 4 2 7/ 37 is
35.2 4.2 2.e, 4.2 1.4 4.2 35.2 1.4 5.6 1.4 2.8 1.4
105* 16g 12g 13g 3g 10g 289g 2g 15g 4g 15g 3g 485g 43g 18g
IC' ...-. 010 -- .S "ri']): U C TO U
top shell 2? ., 1 5 15/25/- 1 7 r 2 c 71/3 37 15
35.2 4.2 2.B 4.2 1.4 4.2 35.2 1.4 5.6 1.4 2.8 1.4
l0S l6g 12E 13g .ig lOg 289g 2g 15g 4g 15g 3g 4S5g 43g 18g
crushed shell 10/1 L 7/2 3 2/1 66/10d 5 3/2 3 2 3e 105/16 25 91
living floor 9.5 0.9 6.6 2.9 1.9 62.8 4.8 2.9 2.9 1.9 2.9
68J 9g 58g 18, 5g 1123g 33, 16g 8g 94g 19g 1451g 50g 105g
shell beside 2 2 9
tl cor 100.0
4g 4g 15g
old humus 8/1 1 15/if 4 1 29/2
27.6 3.4 51.7 13.8 3.4
57g 19g 113g 34 1Gg 239g
TOTAL 43/2 4 9/2 3 5/1 106/13 9 6/2 4/1 4 5 4 207/21 71 106
228g 44g 70g 13g 21F J5g 1525ig 67 22g 15g 12g 125g 22g 2179g 158g 123g
a2 ,ith red paint; i2 with red paint; crushed, sand-tempered; d22/4 from one vessel,
2 with red paint: el grit-tleper: d crude coiiplicated stamped, 1 sherd-teipered brushed,
1 semi-fiber-tecipered brusned; 2 wilh red interiors, 1 with white interior.
Top figure is number; middle figure is percentage based on number; lower figure is
weight in grams. Slash mark indicates rim shards.
St. Johns series pottery, not evident in the other previous tests,
,Ans,-:it ts.7C cert o0 .In'sa "c- H t 'i, .gg t.n cecd
contact with outside cultures. One Deptford Check stamped sherd has
a white (slipped ?) interior, and several sherds have red slipped
exteriors. 'The curvilinear complicated stamped sherd is sherd-
tempered, again suggesting trade. Also listed in Table 3 ate three
brushed or combed sherds, each with a different paste. Brushing, as
cord marking, may be a Deptford technique, perhaps introduced through
The contorted, sherd-tempered plain sherds from Tests I, II, and
III all are very similar to one another, having a smoothed, at times
burnished surface, and a reddish paste and surface color. These
sherds are identical to samples of Thomas Plain, Tchefuncte Plain,
and Larto Red Filmed (Thorne and Sroyles, 1968) in the tNype collections
of the Florida State Museum. They probably are trade sherds from the
Mobile Bay and southern Louisiana area and reflect th;e increased
contact of the Deptford Phase with non-coastal cultures.
Identification of the food bone from the test, most of hhich came
from the living floor, was deer, red-breasted merganser Mergus serrator,
salt iarsh terrapin Malaclemmys terrapine, catfish Bagre :;arinus, and
Caleichchys felis, sheepshead Archosargus sp., eagle ray MyylIobitidae,
raccoon Procyon lotor, and requiem shark Carchiarindidae.
Features 3 and 4 (see Figure 4) both turned out to be nsotholes
rather than pits. Both were 1.0 feet in diameter ac their bottoms,
which were flat, and both extended down from the shell zone 0.9 foot
ipto sterile sand. Shell was used to anchor the posts in the postholes.
The third posthole, shown also in Figure 4, was not given a feature number.
Size and depth were the same as the other two postholes. A humic stain
was present Jn the latter postmold, suggesting that the posthole
rotted in place. None was observed with features 3 and 4, where
vertical lensing of the shell in the posthole mold could indicate that
the posts were removed, allowing shell to slide down into the posthole.
These two postholes were in the depression-living floor area. Perhaps
they were removed when the presumed structure was built. The un-numbered,
humic-stained posthole was on the fringe of the depression and may have
been a support post.
Feature 2 was an oval, bell-shaped pit 1.0 feet deep with a mouth
width of 1.0 feel. At its widest point diameter was 3.6 feet along
the major axis and 2.0 feet along the minor. The pit was probably
used for storage, and later was filled in by dirt, shell, and other
refuse. Contents included 1 Deptford Check Stamped sherd and catfish
bones of the same species listed above. Feature 7 was a sand lense
rather than a pit.
Summary of Stafford North Excavations
As indicated by ceramic frequencies based on small sample
sizes, Tests III, I, and IV seem to represent three different occupa-
tions at the Stafford North site. All three occupations, however,
were Deptford and seem to form a temporal sequence. The earliest
occupation, Test III, was early Deptford with the largest relative
amounts of pre-Deptford fiber- and/or seni-fibel tempered pottery
present. Present also w re sand-itempered cord marked pottery and sherd-
temperad ware, both cord marked and plain. Tea latter were probably
trade sherds from the Tchefuncte Phase. Cord marked pottery was
probably in the process of being accepted by the Deptford peoples as
a technique of surface treatment.
Test I, the next occupation, exhibited a decrease in fiber-
tenpered pottery and an increase in sand- and grit-tempered Deptford
types. Some of the sherd-tempered plain sherds have a very contorted
paste. A distinct Deptford Phase ceramic series has emerged from the
more mixed pottery of the early Deptford Period. This series is
characterized by both stamped and plain ware. Fiber-tempering still
was used somewhat, either indicating trade with peoples who have not
yet accepted sand- or grit-tempering and stamping, or fiber-tempering
was a majority Deptford technique. The latter explanation seems most
Test IV represents a still later Deptford occupation. Deptford
series ceramics increased with a further decrease in fiber-tempering.
St. Johns sherds appear, some with Deptford stamped decorations,
indicating increased contact with non-Deptford cultures. Deptford
Simple Stamped sherds were added to the Deptford series, and red
slipping of vessels seems to have increased.
The Test IV excavations revealed a concave, possible intentionally
cleared, oval depression ca. 22 by 14 feet. A living floor filled the
depression and was covered by additional midden refuse. Over this
depression a light, open structure, perhaps a thatched chickee-like
covering, was built. Extending down from the living floor were three
postholes and one bell-shaped pit. A somewhat similar structure with-
out a concave floor was uncovered at the Table Point site adjoining a
house structure. Thus, the open Stafford North structure may have
been attached to another structure. Future research is needed at the
identified food bone from the site, due to the small sample,
included only 10 different animal species with a minimum of 30
individuals. Table 7 at the conclusion of this chapter lists the
total species and number of individuals for both sites. The species
listed are those expected from exploitation of a coastal environment.
The Table Point site, which received more extensive excavation
than the Stafford North site, is composed of many extremely
scattered separate shell piles. Also at the site is an aboriginally
constructed shell and dirt ring with a large oval midden in the center.
Deposition of humus on the site, especially in the eastern portion, has
buried shell maiddens and a portion of the shell and dirt ring under
a 0.5 to 1.0 foot deep layer of sterile soil. This deposition is a
result both of natural accumulation of organic matter and soil run-off
from nearby, now defunct, agricultural fields, the latter a process
similar to that described for the Stafford North site.
The densest portion of the site stretches 150 yards north to
south along the high tide line of the marsh edge. Width, measured from
the high tide line to the eastern edge of the shell ring, is about
110 )ards. Shell piles can be found scattered over most of the
northwest quadrant of Table Point. No fresh water supply could be
located at the site, though a swamp in the center of the Point may
at one time have drained into the imarsh.
Excavation techniques employed at the Table Point site were the
same as those used at the Stafford North site. Excavation units were
staked out as either 10 by .0, 5 by 10, or 5 by 5 foot "squares."
At times it was necessary to place excavation units off the north--
south oriented grid system. Location of such units was tied into the
grid. Both the grid system and the datum plane were recorded in
relation to a permanent United States Geologic Survey concrete
marker. This information is on file at the Florida State uhlscum.
Elevation of the site ranged from 4.5 feet above mean sea level
near the marsh to 8.0 feet at the eastern side of the ring. Because
the shell piles were more scattered than at the Stafford North site,
more underbrush was present within the confines of the site. IWhere
1.0 feet of humus or more had collected over the shell, scrub bushes
grew. .Most of the piles, though, remained covered with only a sod
Figure 5 shows the excavated portions of the site and the outline
of the shell-dirt ring. Exposed sections of the ring are indicated by
solid lines, while dotted lines represent the buried portions located
by probing wt.th a metal rod and a coring tool.
Portions of the shell in the western half of the ring were
removed during the 19th and 20th centuries for use as road paving
material. The outline of the ring, however, is very visible.
S1 Tide 11e
High rid e
10 -- 30 40 50 L
0 10 20 30 40 50 ---*3TJ ia ,-
Solid lines ait exposed
shell; dAtted are sub-
Figure 5. Table Poliit Excavation Schema.
Wherever the ring was cut into by excavations, its edges were well-
defined, especially along buried sections.
Excavations showed present thickness of the ring to range from
0.7 to 2.0 feet. Diameter of the ring is 220 feet with the width of
the shell and dirt border varying from 20 to 25 feet. Two gaps
30 and 20 feet wide are present on the southwest and north sides
of the ring. I'ithin the ring is a longitudinal shell midden 150 feet
long (east to west) and 60 feet wide.
Discussion of the ring excavations will be divided into
descriptions of the ring fill, the southwestern gap, the area enclosed
by the ring, and the central midden.
Two 5 by 10 foot squares were excavated into the fill of the
ring border; one 5 by 10 foot square and one 10 by 10 foot square
were positioned to intersect the edge of the shell ring in the south-
west gap; and a 3 by 35 foot trench was dug across the north side of the
ring. Figure 6 shows the central portion of the western profile of
this trench which is labeled as "3 x 34 feet" in Figure 5. Excavations
failed to locate a borrow ditch or a ditch built intentionally on
either the interior or exterior sides of the ring.
The fill of the ring was a mixture of shell, mostly oyster, and
grey humus. Density .f the shell in the hemic matrix was much less
than in the shell middens at the site. The shell was not lensed in
the ring as it was in portions of the middens, ann in appearance the
fill was quite different from the shoel miiddens. Mo living floors
were evident and no charcoal lenses were present. The shells and pieces
of food bone and sherds tended to be moie broken than in the middens.
Together these facts suggest that the ring was constructed by
using old midden shell and humus as building material rather than
being in-place-deposited shell midden as from a circular village.
Table 4 lists the sherd counts and bone associated with the ring.
The highest frequencies of cord marked pottery, both sand- and
sherd-tempered, came from the ring fill (see Table 4, row 2). The
frequency was also high in the old humus under the ring (row 3),
suggesting that the ring was built during a period when cord marked
pottery was increasing in popularity at the site. The actual frequency
of cord marked pottery during that time was probably higher than
indicated in Table 4, but the presence of redeposited sherds from
earlier periods decreases the relative frequency. St. Johns series
sherds and semi-fiber-tempered sherds are also present in the pre-ring
Humus deposited over the ring (row 1) exhibited St. Johns, cord
marked, and fiber-tempered sherds, a result, perhaps, of the recent
disturbing of portions of the ring during the 19th and 20th centuries.
Of the seven contorted paste, sherd-tempered sherds in the post-ring
humus, two displayed red-slipped exteriors. The sherds are indis-
tinguishable from the Lower Mississippi type Larto Red Filmed
(Thorne and Broyles, 1968) and must indicate trade with that area.
Three Busycon perversum shell picks were also recovered from the
post-ring humus. All displayed single hating holes.
Food bone species identified from the ring fill were tiger shark
Galeocerdo cuvier. de-r, catfish, salt marsh terrapin, raccoon, pocket
gopher Geemys cf. cumberlandius, and channel bass Scianops eccllata.
The food bones were deposited in the ring along with the midden shell
used in the ring construction. They probably date from varying periods
within the Deptford period.
Table 4. Ring Excavation Potsherd and nod Bone Totals.
OC ..HS '-' S CC C
HOE Ca 0 043
4 u C 0- C I -a 4- C S 04
.- ... C- .5 O 0. 0 0.n
I post-ring 30/1 7 5 15/2 25/1
humus 32.6 7.6 5.4 16.3 27.1
130g 90g 29g 59g 150g
2 shell ring 10 2/1 10/1 16/1
fill 19.2 3.8 19.2 30.8
56g 35g 87g 153g
3 old surface 13 5/1 7/1
under ring 33.3 12.8 17.9
81g 26g 61g
I occupation 16/2 3 19 9 16 538/2d
humus in 11.3 2.1 13.5 6.4 11.3 41.1
gap 61g 23g 137g 49g 99g 523g
5 occupation 6 1 4 6/3
humus 16.2 2.7 10.8 16.2
inside ring 2Sg 3g 12g 140g
6 shell pile 2 1 1/1 18/3
in 8.7 1.3 4.3 78.2
500N, 49IE 9g llg 2g 153g
7 central 6 4 3 3
midden 35.3 23.5 17.6 17.6
shell 39g 24g 14g 77g
8 old hunus 3 10 3 3
under central 7.5 25.0 7.5 7.5
mtdden 32g 65g 12g 27g
77/3 5/1 37 28 57/5 136/11
17.5 1.1 8.4 6.3 12.9 30.8
365g 5Sg 312g 167g 311g 1284g
Top figure is nunuber; middle figure is percentage based on number;
lower figure is weight in gr'mns. Slash maaf indicates rim sherds.
2 5 3 92/4
2.2 5.4 3.3
15g 9g llg 493g 102g 21g
1g 15g 14g 18g 4g 70g 462g 9g S8g
1 3 5/2 5/ 9/ 9
2 .2 5.4 3 12. 12.
Ig g 85gl 45 43 g 279gg
3 11 2 43 6f 52/3 19 37
2.1 2.1 1.4 2.8 2 .0 11.
28g 20g g 14g 1g 4g 7g 146g 11g 9Sg
1 3 5/2 /2 13h 3/5 24
2.6 7.7 12.8.1 12.5.
9g7gg 17g 82g 303g g 49g 23g
3 3/1 2 40 B/If 145/5 66 83
2.1 21 1.4 2. 510
g 20g 3g g 17g 3g I2g 22g
1 1 2 3/28 13h 37/5 24 25
172.7 2.7 5. 8.1 3
9g 9g 7g 17g 82g 30Sg 43g 25g
i 23/4 6 10
4g 178g Cg ?2g
1 17 8 3.9
17g 171g 25g 72g
3 12/1 2 4/2j 40/3 3 17
7.5 30.0 5.0 10.0
12g 187g 7g 29g 371g g 12g
6 9/1 14/1 14 14/2 21/2 23/4 411/30 181 219
1.4 2.0 3.2 3.2 3.2 4.8 5.2
62g 56g 219g 41g g g 138g 197g 3275g 358g 522g
a2 with red pint; b2 sher'.-te'nered smooth (1 uith red paint); c2 'th
red paint; sherd-temrer- tetrapods (1 ith red pint); el with
cross-simple ;tamping; l4/ Bectford Linear Check, 2 eroedJ 1 St. Johns
Check Stamped 1 St. JohnCs rushed, 1 unidjnti ficd 2eroedd, 1/1 with
check stampin,; 1 wCith si.,ple stiaping: 14/1 sand-temper-d, 2 uniden-
tified eroded, 1 shcrd-temp.red bushed, 1 suher-te.,pred plain.
Two 5 by 10 foot squares within the gap and a 5 by 10 and a 10 by
10 foot square intersecting both the gap and the ring provide the
information on that portion of the site. Within the gap, stratigraphy
showed a brown humus, much like the humus collected on the ring, over
a very dark brown-black humus which was both contemporary with the
ring (on the same level) and extended a brief distance under the ring.
Perhaps erosion of the ring caused some of the shell-dirt fill to
fall into the gap covering some of this zone. An occupation floor
characterized by a 0.2 foot thick layer of scattered crushed shell and
artifacts and one possible hearth area with charcoal in it was
recognizable within the brown-black humic zone. About 0.5 foot of
brown-black humus had accumulated on top of the living floor. Figure 7
shows a schematic profile diswing through the gap and ring. The
profile line is marked on Figure 8 as A A'.
The level of the living floor was the same as the bottom of the
shell ring fill, suggesting that the floor was deposited at the sane
time that the ring was constructed. The living floor was more
reminiscent of an area which had undergone heavy traffic than an area
on which household activities had taken place.
Additional evidence for activity on this occupation floor is
showii in Figure 3. Several features extended down from the floor, all
grouped together in Square 465N, 480E (5 by 10 feet) which was located
inside the gap against the edge of the ring.
Feature 1 was a circular area of charcoal and burnt sand about
1.5 feet in diameter and less than 0.2 foot thick. The top of the
charcoal, a possible hearth, was at the same level as the top: of the gap
Figure 3. Top View of Ring Gap Features.
Feature 1 partially overlapped Feature 2, which was a nearly
circular depres-icn 5.0 feet in diameter. The depression was first
defined at the top of the sterile buff sand zone underlying the
occupations strata, 0.3 foot below the bottom of the occupation zone.
The dark nature of the occupation floor made that zone indistinguishable
from the dark fill of the depression. Originally the depression
probably extended down from the occupation floor. The bottom of the
depression sloped downward. Depth in the deep side was 0.7 foot.
Against the northwest side of the depression in the deepest portion
of the bottom was a nearly circular posthole. The posthole, easily
visible against the buff sand matrix, was 2.0 feet in diameter and
0.3 foot deep.
The best explanation for Feature 2 is that it 'as a shallow pit
dug with a sloping bottom in which a large post was set. The sloping
nature of the bottom of the pit allowed the post to be slid in
horizontally and raised by buttressing the bottom against the wall of
the pit. Once in a vertical position, the post was anchored by filling
in the depression. The bottom of the post was 1.3 feet below the top
of the occupation floor. Several lenses of crushed shell, mapped
within the occupation floor, probably were deposited around the in-place
post, suggesting that the occupation floor accumulated after the post
was raised. The Feature 1 fire was built adjacent to the post during
the deposition of the occupation floor. Additional excavations in the
gap failed to revel any other postholes or other features.
Thi1 practice of raising posts in large pits or trenches was in
evidence at the hocse excavation at the 'able Point site also. The
technique was a convenient way for a few people to raise what must have
been heavy posts.
Exact use of the post within the gap is unknown. Its association
with the ring and gap suggests that it was either part of a structure,
an explanation not verified by the archeological evidence, or that it
had ceremonial significance, perhaps as a marker.
One fiber-tempered sherd and one Deptford Cross-Simple Stamped
sherd came from the fill of Feature 2 along with two bone fragments
identified as raccoon. Food bone from the occupation floor was
identified as deer, salt marsh terrapin, opossum, and catfish.
Table 4, row 4 lists ceramic artifacts from the occupation floor
and from the midden debris accumulated on top of the floor. This
zone (Zone C in Figure 7) is believed to represent a closed context and
again illustrates the large variety of ceramic types and pastes present
in the Deptford Phase. Sherds from the Deptford series constitute
57.5 per cent of the total (61.9 per cent if sand-tempering is
considered a Deptford paste form). Two very rounded Deptford Check
Stamped tetrapods or pot lugs, one with a red slip, were recovered
from this zone. Both are grit- and sand-tempered. St. Johns Check
Stamped and brushed sherds and simple stamped, fiber-tempered sherds
are probably copies of Deptford surface treatment techniques. It would
seem that different Deptford groups chose different paste forms for
their pottery. Deptford grit-and sand-tempered pottery is not the sole
ceramic indicator of Deptford culture.
Area enclosed by Ring
Two 5 by 10 foot squares were placed in the area enclosed by the
ring (Figure 5). Stratigraphy in both was identical: a brown, probably
post-ring, humus overlying a dark brown occupational humus, which was,
perhaps, at least partially contemporary with the ring and the gap
occupation floor. These both were above sterile buff sand. The
occupation zone contained some scattered shell and artifacts, but
nothing was found to indicate the intensity of occupation found in
Row S in Table 4 shows the ceramic types from the occupation
zone. The sherd types include fiber- and semi-fiber-tempered sherds
with Deptford decorations. Most likely this occupation zone was built
up over an extremely long period of time, and frequencies are not
important. Food bone was identified as soft-shelled turtle Trionyx ferox,
cetacean (whale or porpoise), catfish, and requiem shark. The interior
portion of the ring did not exhibit intense occupation, and the
little material recovered can be explained as debris deposited over
an extremely long period by various Deptford bands at the site.
Possible explanations for the function of the ring are confused
by the presence of the central shell midden, which in shape is quite
different from other middens at the Cumberland sites. Also, the
conclusions from the excavation of the squares within the ring suggest
that any occupation enclosed by the ring was restricted to the central
midden. Two tests, one a 5 by 5 foot square and the other a S by 10 foot
square, were placed in the central midden. These excavations showed
that the midden was in-place-deposited midden rather than redeposited
material as was the -ring fill.
Stratigraphy was simple: a dense shell midden, often lensed, within
a matrix of dark brown humus lying over a mottled, grey and brown old
humus zone. Under these layers was the sterile buff sand zone. No
living floors or evidence of structures weere encountered.
The old humus zone underlying the westernmost test was almost
sterile, while that under the eastern test contained artifacts and
some scattered shell. This suggests that the western portion of the
midden was deposited first with contemporary accumulation of some
scattered refuse to the east of the original midden. This original
midden may have been circular. Through time refuse was deposited more
and more to the east, moving the midden's edge in that direction.
Eventually dense midden covered the old humus. Sherd frequencies
between the two tests are not different, suggesting that this
horizontal enlargement of the midden occurred over a relatively brief
period of time.
Table 4 rows 7 and 8 list artifacts for the shell occupation
zone and the pre-midden old humus, combined for both squares. The
12 Deptford Geometric Stamped sherds in row 8 all seem to be from
the same pot and all came from the eastern test. The stamp motif is
composed of very precise diamonds rather than checks. Deer, salt
marsh terrapin, catfish, raccoon, musk turtle Sternotherus sp., and
red drum Pogonias cromis were identified from the food bone remains.
Feature 3 was assigned to a possible pit in the western test.
Closer examination, however, showed the feature to be a humic lens in
the shell midden matrix. Other than containing deer and catfish bone,
the lens was sterile.
Summary of ring excavations
Determination of the function of the Table Point shell and dirt
ring is difficult. Lack of a ditch either on the outside or inside
of the ring would seem to negate its being a European-like defensive
palisade, as does the presence of two large gaps in the ring. Webb
and Snow (1945: 31) describe the median dimensions of Adena sacred
circles (based on a sample of 74 from Ohio and Kentucky) as 212 feet
in diameter with a single gap of 30 feet. These measurements compare
quite well with the Table Point ring size.
The excavations on the northeast side of the ring which uncovered
a Deptford Phase house structure showed that the ring was built on
top of the humus associated with the outside of the house. A radio-
carbon date of A.D. 55 (standard deviation of 95 years) for the house
indicates that the ring was constructed after ca. A.D. 100 or later.
The presence of both cord marked pottery and sherds similar to
lower Mississippi valley types both suggest widespread contact with
phases throughout the Southeastern area. If the ring's function was
ceremonial, it may be the result of contact (either directly or
indirectly) with Adena-Hopewellian peoples. The ring is definitely
later than the shell rings found on the Georgia and South Carolina
coasts, which are associated with pre-Deptford fiber- and sand-tempered
pottery. This is evidenced by relative dating with the radiocarbon-
dated house and the presence of Deptford, cord marked, and St. Johns
pottery in the humus under the ring.
Rings, constructed of dirt with interior ditches, are found at
several sites in the Ckeechoebec Basin in south Florida (Milanich,
196Sb). At the Ft. Center site on the west side of Lake Okeechoobee
one of three intersecting rings was dated by pottery and other
artifacts from ca. 500 B.C. to A.D. 1 (Milanich, 1968b). Hopewellian
elements are present at Ft. Center. Thus, the stimulus for the
Table Point ring could have come from south Florida as well as from
the Ohio River area.
The central midden within the ring, which seems to have been
contemporary with the ring, suggests that the shell and dirt pile
was a defensive enclosure rather than a ceremonial circle. Perhaps
the ring anchored a vertical post palisade. Elsewhere at the-side,
posts are anchored in trenches or pits, and shell and dirt are used
as fill to support the posts. More excavations are needed to ascertain
the relationships of the ring and the central midden and gaps. The
large size of the central midden relative to other middens at the
site is also puzzling, but it tends to reinforce the hypothesis that
the ring was a defensive enclosure, perhaps with entrance gaps.
Possibly the ring was not completed before abandonment of construction.
ne 5 5 5 foot test square was placed over a buried shell midden
40 feet northwest of the house excavation. The middn '.as located
by probing. Few artifacts were recovered from the excavation of the
0.5 foot thick shell occupation zone. They were 1 grit-tempered cord
marked sherd and 3 sem i-fiber-tempered plain sherds. Food bone was
deer anrd requiem shark. 7t was concluded that the shell midden was
deposited beside an area of more intense occupation. Lack of time
prevented excavating more tests over a wider area of the site.
The most informative result of the Cumberland Island research
was the excavation of a Deptford Phase house. The structure is the
first archeologically excavated Deptford house and provides a unique
source of information on Deptford household processes.
Figure 9 shows the outline of the house excavations and the
uncovered portions of the associated features. Excavation of the
extreme western wall of the house was hampered by the presence of
several large live oak trees and a stand of saw palmetto. Tree roots
tended to grow either over or under midden shell and shell fill
associated with the house, rather than growing through the shell
(taking the route of least resistance). This prevented major disruptions
to the shell-filled features. Close to the trees, however, the
natural thickening of the roots, some nearly a foot in diameter,
uplifted the shell fill of the wall trench, making interpretation
The house proper was less than 10 feet from the edge of the shell
and dirt ring, enabling comparative stratigraphy to be made as
discussed above. Stratigraphy of the house was simple. Figure 10
shows a profile through the center of the house, across the 485E grid
line labeled A A' in Figure 9. The orientation of the house was not
determined until after the baulk across the house had been left, thus
the profile does not exactly bisect the house.
In Figure 10 the sterile buff sand zone, Zone F, underlies the
house occupation zones. The top of Zone F was 1.1 feet above the
NOS" ---- --
water table at high tide. Above this zone was Zone D, which
represents both the old humus under the house and humus (non-shell)
accumulated during the house occupation. The lower portion of this
zone was in the process of leaching out. Under the shell midden areas,
Zone C, the old humus was much darker, the result of leaching of
organic materials downward from the shell. This zone is labeled
Zone E. Zone E, which was entirely old humus, and the lower two
thirds of Zone D was almost sterile. The artifacts associated with
the house were in the upper one third of Zone D, indicating that this
zone accumulated during the occupation of the house.
Zone C is the shell occupation zone which is confined to the
presumed cooking area. Zone C' in the "sleeping area" contained some
scattered shell, much less than Zone C. A thin layer of sterile grey
modern humus and sod capped the house strata, sealing the zones from
modern disturbances. Once it was discovered that this zone was
sterile, it was scrapped off and discarded with the sod.
"Work area" stratigraphy
Tangent to the northeast side of the house was a buried, either
oval or circular area of shell midden. Over this shell was the same
Zone B modern huiic layer as over the house. A very distinct brown,
mottled old humus layer underlay the shell.
House and work area features
The walls of the house were constructed of posts set side by side
in a wall trench and anchored with a mixture of shell and dirt fill.
Figure 11 shows a reconstructed floor plan of the house walls, the
associated house features, and the work area. The wall trench had
0 10 20 30 4
I-- ----------------- --10---20 -i-30----
S tca osts 1set in trench
npen ivig - ..
and work-p --_
Figure 11. Reconstructed )iouse Pattern.
a width at its tip of 3.0 to 3.8 feet and narrowed to about 1.0 feet
at its base which was flat. The average depth was 1.2 feet. Deepest
and widest portions of the wall trench were on the east side. Perhaps
most of the weight of any roof beams was supported by this side.
Often circular depressions 0.5 to 0.8 foot in diameter extended
0.1 to 0.3 foot below the trench bottom, probably where individual
posts were set in place. Lack of humic stains in the trench
suggests that the wall posts were removed before rotting in place.
Several profiles across the trench showed vertical bedding of shell
which might have occurred -hen shells slipped downward into the holes
left when the pots were removed.
The wall trench was dug with the interior wall steeper than the
exterior side. Occupation debris (shell, sherds, and food bone) which
had accumulated on the cooking area floor was swept against the vertical
wall posts, creating a characteristic low hump of refuse along the
interior cooking area walls. On the exterior edge of the trench humus
accumulated in the depression left by the settling of fill in the
Length of the oval house along its major axis was 32 feet and
width was 22 feet. These measurements are very similar proportionally
to the oval living floor area (22 by 14 feet) at the Stafford North
site. Two openings were left in the house wall. The largest on the
southerly end of the house was probably to allow cooking fire smoke to
escape. The other on the eastern side must have been an entranceway,
all-wing admittance to the two partitioned roons.
Inside the house the partition, placed in a shallow slot trench,
separated the presumed cooking and sleeping areas. This partition
trench, originally filled with a shell and dirt mixture around vertical
posts, was at some point in time redug with dark brown, shell-less
humus used as fill (Figure 10). No evidence was found to indicate
that posts were set in the slot trench, but this seems to be the best
explanation for its presence. The fire pit was round when first dug by
the Deptford house inhabitants. Later it was extended and became more
rectangular. This extension would have brought the pit closer to
the original partition (the two almost intersected) and may have been
the reason for the replacement of the partition. The second partition,
which overlay the first along part of its length, curves away from
the fire pit extension. Only one potsherd, a cord marked rim sherd,
was recovered from the partition fill.
The fire pit, believed to have been about 4.5 feet in diameter
originally, was extended to a length of 9.0 feet, measured at its top.
Depth was 2.2 feet. The wall closest to the partition was steeper
than the opposite wall, suggesting that the pit was dug (by scooping
out dirt) from the south side (less steep), probably after the
partition was built.
Alternating lenses of shell and ash in the fire pit (shell was
95 per cent oyster with the remainder being species mentioned above)
indicate roasting of small quantities of shellfish at one time, two
to three dozen being the best estimate. Burnt shells were recovered
in the shell midden zone around the fire pit where the roasted
oysters and other shellfish were eaten. Many pieces of food bone were
recovered from the fire pit and the refuse in the cooking area. These
are discussed below in this chapter. Several Busycon pervrsum picks were
excavated from the fire pit and from against the interior house walls.
The least broken food bones were also against the house walls.
Those on the floor around the fire pit were very fragmented due to
the heavy traffic in that area.
A radiocarbon data of A.D. 55 95 years (sample UGA 129
analyzed at the University of Georgia) was obtained from a Busycon
perversum pick taken from the fill of the fire pit. The shell did
not show evidence of burning.
Adjacent to the fire pit on the west side was a small posthole
(see Figures 9 and 11) which might have served as a spit support or
similar device. The largest amounts of food bone cane from the
west and south sides of the fire pit.
A posthole central to the cooking area and close to the fire
pit probably helped to support the roof, or it was in some way
connected with a spit device. The large size of the posthole stain,
i.4 feet in diameter, suggests a roof support. As was the case with
the posthole in the ring gap, the central post was set in the deep
end of a large sloping depression. Fill of the depression included
1 St. Johns Plain sherd, 2 grit-tempered plain sherds (Deptford paste),
and several very fragmented pieces of unidentified food bone.
The annular shell area on the eastern side of the house was
about 25 feet acToss and 0.6 foot thick. Excavations showed that the
north and cast sides of the shell ended abruptly. Humus accumulated
beside the shell was sterile. The general pattern of shell deposition
and the finding of flat-lying potsherds suggest that the shell was
deposited in place rather than being dumped from inside of the house.
Little of this shell ias burnt.
Another post set in a longitudinal trench was aligned with the
north edge of the shell "work area." The postmold, 1.2 feet in diameter,
extended down from the deep end of a trench 7.1 feet long and 1.5
feet wide. The trench was very shallow, only 0.5 foot deep in the
lower end. The postmold extended 0.2 foot below the bottom of the
trench with a humic stain present. Sterile shell fill was used to
anchor the post.
Possibly this oval or circular work area was roofed, supported by
several large posts, as the one excavated, and by part of the house
wall. A large amount of artifacts, especially deer antler, was
recovered from the shell occupation zone. About one fifth of a
Deptford Simple Stamped pot (36 sherds) was found, the sherd lying
together in the shell. Shape and size of the pot seems to be much
the same as the pot recovered at the Stafford North site. The shell
area may have been an outside, warm weather sleeping and working area.
House and woik area artifacts
Table 5 lists ceramic and bone artifacts from the various
features associated with the house. In addition, one chert chip was
recovered from the fill of the house wall trench, and a broken,
polished bone pin tip cane from the top of the old humus underlying
the house. Four Busycon rerversum picks were found around the fire
pit, 7 on the inside of the wall trench, 1 within the fire pit, and
1 in the occupational humus zone outside of the house. An almost
square coquina "hone" was associated with the picks in the cooking
area shell. Two Busycon columella used as awls were recovered from
the sleeping area occupation floor. No picks were located in the
presumed sleeping area. Stone tools were very scarce among the
Cumberland Island Deptford peoples, and most of the tools must have been
of wood and shell withb some hone used.
Table 5. House Potsherds
t to 0 .
+ I C 4-H 0 4-
kitchen shell 17 3 5/1 4 51/1
19.5 3.4 5.7 4.6 58.7
84g 27g 69g 13g 439g
sleeping area 9 3 13a
23.7 7.9 34.2
65g 5g 97g
other house 4 1 1 15
floor 16.6 4.2 4.2 62.5
27g 8g 6g 195g
old humus 4/1 2 5
in house 28.6 14.3 35.7
25g Sg 62g
occupation 12 5/2 41/4
floor out- 19.7 8.2 67.2
side house 69g 33g 388g
ol4 humus 7 1
outside house 63.6 9.1
wall trench 37/1 2 57/1e
fill 32.7 1.3 50.4
325g 20g 733g
fire pit fill 11 7
sork area 12 3 2
shell 1.4 3.8 2.6
90g 27g 13g
and Food Bone Totals.
S. 40 t.H E 0 4
0 2b 38 13 7
2 5 57/2 14 92
2.3 5.7 99.9
9g 40g 681g 25g 165g
1 10 26 38 13 7
2.6 26.3 5.3 100.0
27g 30g 26g 250g 26g Sg
2 l1 24 6 4
8.3 4.2 100.0
10g 6g 252g 12g 220g
1 2 11/1 2 14
7.1 14.3 100.0
9g 10g 114g 3g 27g
1 2d 61/1 17 29
1.6 3.3 100.0
7g 10g 507g 36g 34g
1/1 2/1 11/2 2 4
9.1 18.2 100.0
21g 6g 66g 3g 3g
2 3 6 6 3 3/1f 113/3 39 129
1.8 2.6 5.3 2.6 2.6 99.8
35g 97g 15g 21g 62g 1312g 81g 521g
18 3 34
130g 3g 53g
36/49 5 15 3/2h 2i 78/6 16 32
46.1 6.4 19.2 3.8 2.6 99.9
436g 72g 58g 45g 8g 749g 20g 147g
work area 2 1 3 = 7
old hum's 66.7 33.3 100.0
19q 2g 21g 24g
TOTAM, 1H/2 2 7 12/3 9 191/6 41/4 9/1 1 36 3/2 11/1 10/1 417/20 112 352
25.7 0.4 1.6 2.7 2.0 42.7 9.2 2.0 0.2 8.1 0.7 2.5 2.2
823g 20g 40g 11g 18g 1973g 49Og 190g 27g 128g 45g 74g 112g 4082g 209g 984g
Top figu.re i3 nu.Lcr; middle figure is percentage based on number; al sherd-tempered; b1 Deptford Linear Check Stanped, 1 sherd-tempered; Cl
lower nu::.Cer is :;eight in gra.s. Slash -ark indicates rim shards. rectilinear stamped; dl punctatd sand-tempered cross-incied; Sal from
one pot; el podal support; 2 St. Johns Cjeck Stamped (1 pot bottom) 1
grit-terpered with corn cob markings (?); n2 incised rims; 12 Dcptford.
Perhaps the most significant aspect of Table 5 is the close-
ness of total ceramic frequencies to total frequencies from Test IV
at the Stafford North site (the supposed oval structure, Table 3).
This suggests close temporal proveniences for the two occupations.
The only major difference is the presence at the Table Point site of
11.2 per cent of Deptford Simple and Cross-Simple Stamped types as
compared to only 2.4 per cent at the Stafford North Site. This is
due to the large number of sherds found from one pot in the Table
Point work area. Comparisons of the frequencies from the two sites
show : total Deptford series (Table Point to Stafford North) 79.8 to 78.7;
St. Johns series 8.1 to 6.7; contorted paste, sherd-tempered 1.6 to 1.4;
fiber- and semi-fiber-temipered 3.2 to 1.9; cord marked 2.0 to 2.4;
sand-tempered plain 2.7 to 4.3; "other" 2,2 to 1.9.
Unique sherds include one sand-tempered sherd with corn cob like
markings from the sleeping area, and two incised rips on fiber-tempered
paste. The incised rims, which are triangular in cross section, arn
similar to other specimens of the type Orange Incised fro!, Sunday
Bluff (Bullen, 1969: 28, Pl. II j). Bullen notes that the particular
motif is unique at Sunday Bluff, but is similar to other specimens
from the Sinmmer Haven site (Bullen and Bullon, 1961) in St. Johns
\ Deptford Check Stamped tetrapod and a portion of the bottom of
a very thick, small cylindrical vessel of St. Johns paste with check
stamping on the bottom were also recovered. The reconstructed
diameter of the St. Johns Check Stamped vessel is about 6.0 cm wIith
a wall thickness of 2.2 cm.
The ceramic artifacts associated with the Table Point house
structure are much the same as those found elsewhere at the Table
Point and Stafford North sites. Deptford Check Stamped, Linear Check
Stamped sherds with grit and sand paste are present along with sherds
exhibiting St. Johns paste, fiber-tempering, semi-fibcr-tempering,
shecd-tempering, and sand-tempering. The contorted paste sherds also
Cumberland Island Subsistence Information
Table 6 lists the species of animals other than shellfish
identified from the excavation of the Table Point house structure
including number of individuals. Table 7 lists species and number
of individuals from both sites. Identification of the food bone was
done by Curtiss E. Peterson, graduate zooarcheology student at the
University of Florida working at the Florida State Museum.
The Table 7 species list is quite similar to the species reported
by Waring and Larson (1968: 26S, 275) for their excavations at the
Sapelo Island shell ring. This is to be expected if there was little
change in the Coastal biome and exploitation of the same biotopes
occurred throughout the Coastal Tradition.
The excavation of the house structure and recovery of food bone
from this closed context provide an excellent chance for an analysis
of one household's diet. The sample of food bone, however, was small
and it may not be indicative of the entire range of animals utilized
by the house's inhabitants. Tables listing Peterson's detailed
analysis of the house food bone are presented in the Appendix.
Table 6. Faunal List of Species Identified from House Excavation.
cf. Monachus tropicalis
West Indian seal
Whale or Porpoise
Salt marsh terrapin
Table 7. Faunal List Identified from Cumberland Island Excavations.
cf. Monachus tropicalis
Geomys cf. cumberlandius
West Indian seal
Whale or Porpoise
Salt marsh terrapin
Red breasted merganser
Although the faunal sample was small, important information was
provided on food preparation techniques. The animal species identified
are those expected from the types of biotopes thought to have been
exploited by the coastal Deptford peoples (see Chapter III).
Almost the entire left hindquarter of a deer was recovered from
the east side of the house against the interior of the house wall.
The bones included the femur, tibia, a metatarsal, and a distal
phalanx from the same animal. Evidence of butchering marks or charring
from cooking was not present, suggesting that perhaps the haunch was
boiled in a vessel or roasted in such a manner that no exposed bone
was burnt. The ferlur and tibia were uncovered in normal anatomical
position, perhaps a result of the bones being cast aside after the
meat was stripped off and cooked or dried separately. The deer
seems to have been a young adult. Portions of bone from a left front
quarter were found just outside the east side of the house. Peterson,
citing the two quarters and a lack of vertebrae or ribs at either
site, postulated quartering of deer in the field with only the
haunches being brought back to camp.
Butchering techniques used to clean turtles can be derived from
analysis of the turtle bone. Petersor noted that the only pieces of
turtle shell identifiable as to anterior or posterior position were
anterior and from the ventral side of the turtles. This suggests that
the turtles were opened from the anterior-ventral end to remove the
identified vertebrate species, other than seal, are those normally
found on coastal sites, both during the Deptford period and at the
present time. Seal bones have not previously been identified at
Southeastern coastal sites, though the species Monacus tropicalis
was present up to the 20th century along that coast. Disappearance
of the seal was due to human predation rather than a natural climatic
change (Allen, 1871). It is due to its known former presence on
the Southeastern Atlantic seaboard that the species tropicalis was
assigned to tile Cumberland Island specimen.
Exploitation of the iarsh and tidal streams accounted for the
largest number of animal species (see Table 7). From both sites bones
from raccoon (20 individuals), salt marsh terrapin (22), and various
fish species verify heavy utilization of the marshes and streams. At
least 37 individual deer were identified, making deer the largest
source of neat relative to other species. The fragmentary condition
of the fish bone, which made identification nearly impossible,
lowered the number of individuals in the various fish species. Probably
fish were more important than is shown in the table. Several minute
fish vertebrae were recovered from the fire pit fill. These could not
be identified, but may be anchovy.
The pocket gopher Geoyvs cumberlandius has been found only on
Cumberland Island. It is highly improbable that two pocket gopher
species existed together on the island, thus the species cunberlandius
was assigned to the specimen from the Stafford North site. Whether
the gophers were utilized for food or died naturally in the ground is
unknown. No articulated Geomys bones were found.
Identification of see turtles Cheloniidae, which nest on the
beaches during the months of May to July, suggests summer occupation
of the site. ,Te bone sample was not adequate to provide specific
inforiiation on occupation during other seasons. Animal species and
hunting and gathering techniques associated with the Deptford Phase
peoples are discussed in more detail in Chapters III and VI. The
natural environment of the Deptford region, including Cumberland Island,
is also described in Chapter III. Information derived from the
Cumberland excavations is used to provide data for the discussion
which follows concerning the Deptford economic pattern.
GEOGRAPHICAL AND NATURAL SETTINGS
One basis for the definition of the Deptford Phase offered here
is the relationships between natural environment and Deptford culture,
i.e., the occurrence in certain biotopes of specific groups of trait
complexes and the usage of those biotopes by the Deptford people. In
the past, descriptions of the Deptford Phase have been circumscribed
around elements of material culture, especially ceramics. Other
phases possessing similar traits were seen as variants of Deptford.
This, however, is not the case since the Deptford Phase, by definition,
is coirposed of traits which occur together and in specific natural
settings. Consequently, the geographical distribution of the
Deptford Phase described here does not include some portions of the
Southeast where Deptford pottery types occur.
The bounda--ies of the Deptfcrd region correspond to the following
present-day geographical Pnd political features (see Figure 12): from
Cape Fear, North Carolina, almost due south along the Atlantic coast to
the mouth of the St. Joh-is River, continuing south across peninsular
Florida to Charlotte Harbor on the Florida Gulf coast; from Charlotte
0 100 200
Figure 12. The Deptford Region.
Harbor north and then west along the Gulf Coast to the Florida-
Alabama border. The inland boundary parallels the coastal line, forming
a zone about fifty miles wide, except where sites extend farther
inland along river valleys. This V-shaped region is the result of
both cultural and natural factors which are described in this and
the following chapters.
The north-south distance from Cape Fear to Charlotte Harbor is
approximately 600 miles. Coastal distance from Charlotte Harbor to
the Florida-Alabama state line is about 550 miles. The inland site
farthest from the coast, Mandeville in Clay County, Georgia, is just
over 100 miles from the Gulf of Mexico. Another inland site, Sunday
Bluff in Marion County, Florida, is 75 miles from the Gulf. These
measurements show the spatial extent of the Deptford Phase, especially
the long coastal line.
The plotting of Deptford site locations within this region
reflects the geographical division of the Deptford region into Atlantic
and Gulf sub-regions (Figure 12). These sub-regions later evolve quite
differently in regards to specific Deptford evolution. Geographical
locations of sites with Deptford components are discussed below by
sub-region and by coastal or non-coastal provenience.
It is generally thought among archeologists that the greatest
number of Deptford sites and, hence the greatest Deptford cultural
concentration, is found along the Atlantic Coast. This assumption is
a reflection of the first major Deptford excavations being carried out
in the Atlantic sub-region, and the existence of a great deal more
published an. unpublished information regarding that sub-region.
In fact, however, northwest Florida alone has surpassed the
Atlantic sub-region in the number of Deptford sites. his is due
largely to the surveys carried out by William C. Lazarus. Tn addition
to the known sires, many of the Gulf coastal sites are probably
submerged under the Gulf of Mexico and will remain undiscovered.
Atlantic Sub-Region (Coastal)
At the northern extent of the Atlantic sub-region, Deptford
materials have been reported from sites near Cape Fear, North Carolina,
by Caldwell (1952: 316), Waring and Holder 11968: 143-44), and Haag
(Caldwell, 1958: 49). Excavation of Deptford sites in this locality
is needed for comparative purposes.
Farther south in South Carolina, Waring and Holder (1968: 143)
have noted that "pure sites occur frequently" along the entire coast.
Again, field research is needed. Two sites, 38-Ho-l and -6, have
been located near Myrtle Beach (Miller, 1950), and Griffin (1943) has
described Deptford materials from the Lake Charleston site near
Beaufort. The presence of Deptford ceramiics at Beaufort has also been
reported by M' rma dicke H. Floyd (Caldwell, 1952: 316). Just north of
Beaufort is the Potato Island site (Caldwell and Waring, 1939a), also
exhibiting a Deptford Phase component.
The W.P.A. surveys in the vicinity of Savannah produced a
relatively large number of sites with Deptford components: Bilbo,
Deptford, Refuge, Trene, Dotson, New Ya macraw, Oak Grove, and a site
or. Pipeiakhrs Creek (Caldwe31, 1952: 315; Caldwell and Waring, 1939a).
Just across the Savannah River in South Carolina is the Gullahgyden
site (Caldwell and Waring, 1939a). Sites on Wilmington Island (located
southeast of the present city of Savannah in a salt water marsh)
include Meldrim, W'althour, and Oemler. The latter three also contain
large Wilmington components. Site reports on the excavation of
several of these sites have recently been published in the Waring
Papers (Williams, 1968), and other unpublished and uncatalogued
reports are on file at the University of Georgia.
Recently Caldwell has located at least two Deptford sites on
St. Catherines Island: Seaside and Wamassee (Betty Anderson, University
of Georgia, personal communication). Holder encountered Deptford
materials farther do-wn the coast at the Evelyn Plantation mound complex
in Glynn County and on St. Simons Island. Chapter II of this paper
reports on the excavation of two Deptford sites on Cumberland Island,
the Stafford North and Table Point sites, both in Camden County. Other
unlocated sites must exist on the central and southern Georgia coast,
but much of the area remains unsurveyed.
Amelia Island, the northernmost Florida barrier island, provides
little evidence for Deptford occupations. Bullen and Griffin's
survey of the island (1952) revealed a total of 48 archeological sites.
Only three of these sites contained Deptford sherds, and these
contained a total of 13 sherds out of the 4009 sherds collected. The
rinmber of Orange and St. Johns ceramics, nearly one quarter of the
total, suggest that A.-elia Island was principally a St. Johns Tradition
The coastal zone from Amelia Island to St. Augustine does
exhibit some sites with Deptford components, indicating that the area
was a f,-inge area culture with various cultures moving in and out of
the zone. Near the mouth of the St. Johns River, Sears (1957) excavated
several sites with Deptford components. The three sites, Du-58, 59, 61,
are known collectively as the Willey Browne sites, and they provide
important data on Deptford-St. Johns relationships. Deptford sherds
have been found at the Fountain of Youth site, Sj-31, in St. Augustine
by John Goggin (1952: 105). These, however, may represent trade
Atlantic Sub-Region (Non-Coastal)
The evidence for the presence of Deptford sites in the interior
of Georgia and South Carolina is scanty. Either the sites do not
exist, or, more likely, they are small camp sites and insufficient
surveys have been undertaken. A number of sites in the Fall-Line-
Piedmont biome of northern Georgia located by Wauchope (1966) contain
what he refers to as Deptford ceramics. Whether or not these are
actually Deptford sherds from the coast, perhaps traded in to the
Piedmont, or whether they are variations of simple and check stamped
Piedmont ceramics remains to be ascertained.
The majority of sherds from these sites, as Mossy Oak and Carters-
ville, and the subsistence technology associated with the cultural
materials both suggest culture phases quite distinct from the Deptford
Phase. Fairbanks (1956) has similarly reported Deptford shards from
Macon, Georgia, from the Stallings Island site (1942), and from site
91 61 near Gainesville, Georgia (1954). Again, whether or not these
sherds are Deptford trade sherds or Piedmont variants is not certain.
Recent excavations by Drexel Petcrson at Groton Plentation located
on the South Carolina side of the Savannah River seventy miles from
the coast have revealed Deptford components (personal communication).
The only other Deptford site definitely known for the Atlanitic sub-
region is Floyd's Island in the OKeefenokee Swamp (McMichael, 1960:
Other information regarding the Atlantic sites is not specific.
Caldwell states (1953: 49) that many Deptford sites extend up the
Georgia river valleys, and Kelley notes that there is a large number
of Deptford sites in the vicinity of Brunswick, Georgia (McMichael,
1960: 106). In general the work of C. B. boore regarding the Georgia
and South Carolina coasts (1897; 1899) is not detailed enough to
determine if any of the sites reported by him are Deptford or not.
Most of the sites are coastal burial mounds, though, rather than
Gulf Sub-Region (Coastal)
A ceramic series very similar to the DeptFord series has been
described for the Alabama coast frcm Mobile Bay Lo the Mississippi-
Alabama state line. Known as the McLeod (DeJarnette, 1952: 275-76) or
M.Leod--Deptford Series (Sears, .966: 14) the ceramics are local variants
of the overall Southeastern stamped pottery horizons. The HiLeod-
Ceptlord nomenclature was selected to point out the similarities to
Deptford ceramics, but was interpreted by many prehistorians to
represent cultural ties. However, since it has been shown that no
cultural relationships exist between the two phases, the McLeod-
Deptford terminology has Lcen dropped.
Thut western boundary cf the Deptford region is placed at the
Alabama-Floiida border. Certailnly, cas with the Deptford-St. Johns
fringe area, a similar irea exists in western Florida with many sites
being mixed culturally. Sites exhibiting pure Deptford components
are first recognized just east of Pensacola in Okaloosa Coutity,
A number of sites have been surveyed and recorded in Okaloosa
County. These include the Ft. Walton site, Ok-6, which displays a
long Coastal-Gulf Traditions sequence (Fairbanks, 1965; Lazarus,
1965a: 113), and Ok-10, 23, 30, 36, 46, 63, 67, and 72, all surveyed
by Lazarus. Data from Lazarus' surveys is on file in the University
of Florida state archeological site file.
Walton County is the location of three sites, W''1-29, 42, and 69
(Lazarus, 1965a; 1965b). 11-29, the important Alligator Lake site,
has produced the earliest Deptford Phase radiocarbon date, early 7th
century B.C. At the site contact is indicated between the Deptford
Phase and contemporary phases of the Pickwick Basin in northern Alabama
and eastern Tennessee and the Bayou La Batre ceramics of the Mobile
Bay area. This intense culture contact laid the base for the later
development of the Gulf sub-region into the Gulf Tradition (see
Eastward along the Florida Gulf coast, several sites are located
in Franklin and Iakulla Counties: Fr-1, 2, and 9 (Willey, 1949: 353);
and "Wa-6, 7, and 34 (IVilley, 1949: 292-93, 353; Allen, 19514. The
Tucker site, Fr-4, has been excavated both by Scars (1963) and Phelps
(1966a). Their results have been published and are informative
concerning the Depi-ford Phase inl northwest Florida, especially
relationships with the Coastal Tradition.
Information on sites from Apalachee Bay (Jefferson County) to Tampa
Bay is scarce, a result of relative sea level changes, lack of surveys,
and, in the Tampa Bay area, modern construction. Phelps (1966b:24)
has investigated the Williams site, Ta-32, in Taylor County hiiich
displays a Dcptford component.
Current research by the University of Florida at a mound complex
site in Dixie County, Garden Patch site Di-4, has revealed Deptford
middens. Just south of the Garden Patch site at the Crystal River
site, also a mound complex, Deptford materials are present (Bullen, 1953b;
Sears, 1962). Deptford middens have also been noted at the mouth of
the Crystal River. Though sites are not recorded for the north
peninsular Florida coast, Goggin (S.F.A.C., 1966: 12) has noted that
along the coast from the Withlacoochec River to the Florida "bend"
sites are found "IWhenever a bit of high land touches the Gulf .
Over one half of these sites are Deptford and/or Swift Creek."
South of Tampa Bay Deptford materials occur even loss frequently,
having been found at the Yellow Bluffs-Whitaker Mound, So-4 (in mound
fill) (Milanich, 1971a), and at the Englewood site, So-I (Bullen,
personal communication). The latter site, just north of Charlotte
Harbor, marks the southernmost extension of the Deptford Phase. The
Charlotte Harbor area is a fringe area between the Deptford Phase and
the little known south Florida phasess.
Gulf Sub-Region (Non-Coastal)
Deptford sites are found away from the coast along many of the
streams and rivers which empty into the Gulf. Bullen's surveys have
locate. several small sites, probably band-size camp sites, in
Jackson County, Florida. These are on the Chottahoochee River, Ja-18,
42, and 62, and the Apalachicola River, Ja-5 (Bullen, 1950; 1958).
Farther north, just off the Chattahoochee River in Clay County, Georgia,
is the Mandeville mound complex site, 9-Cla-I (Kellar, et al., 1962:
n.d.). Close by is site 9-Cla-2, also a Deptford site (Nonas and
Baillou, 1963). The Mandeville site is similar to the Crystal River,
Horseshoe Beach, and perhaps Evelyn sites in that all display a Dept-
ford component associated with and/or followed by a growth in site
size and increased mound building, probably indicative of increased
social complexity and horticulture. A single mound displaying Deptford
materials is the Oakland Mound, Je-53, in Jefferson County (Morrell,
In peninsular Florida Deptford ceramics occur in small numbers
at nearly all sites, again suggesting small, temporary campsites.
Post-Deptford cultures recognized these former campsites as potential
village sites, resulting in the Deptford components being overlain by
Swift Creek, St. Johns, Weeden Island, Alachua, or other components.
Such a situation occurs at the Manatee Springs site in Levy County
(Bullen, 1953a), and at the Melton site in Alachua County, A-7 (Smith,
n.d.). In the latter instance a small Deptford midden underlies a
later Swift Creek-St. Johns burial mound.
The Sunday Bluff site, Mr-13, located along the Oklawaha River
in Marion County has been excavated and reported by Bullen (1969).
Both Sunday Bluff and the nearby Colby site, Mr-57, exhibit a mixed
St. Johns-Deptford component. The two sites indicate intermittent
occupations by band-size groups, as evidenced by horizontal stratigraphy
and intra-site, relative midden sizes. Central Florida seems to be a
fringe area between the east coast St. Johns Tradition and the west
coast Deptford phase.