FLORIDA ANTHROPOLOGICAL SOCIETY PUBLICATIONS
FIBER-TEMPERED POTTERY IN SOUTHEASTERN UNITED STATES AND
NORTHERN COLOMBIA: ITS ORIGINS, CONTEXT, AND SIGNIFICANCE
Ripley P. Bullen and James B. Stoltman
The Florida Anthropologist) Volu m e 2 5 > N u m ber 2Â¡ Part 2
19 7 2
THE FLORIDA ANTHROPOLOGIST is published quarterly in March, June,
September, and December by the Florida Anthropological Society, Inc. at
3301 College Avenue, Ft. Lauderdale, Florida 33310. Subscription is by mem
bership in the Society for individuals interested in the aims of the Society. An
nual dues are $4.00; student membership $2.00. Requests for membership and
general inquires should be addressed to the secretary; subscriptions, dues,
changes of address, and back issue orders to the treasurer; newsletter items
t the president; and manuscripts for publication and books for review to the
editor. Second class postage paid at Ft. Lauderdale, Florida.
Preface by James B. Stoltman i
The Cultural Context of Early Fiber-tempered Pottery in Northern
Colombia by G. Reichel-Dolmatoff 1
The Orange Period of Peninsular Florida by Ripley P. Bullen ... 9
Fiber-tempered Pottery in the Tennessee Valley by John W. Griffin 34
The Late Archaic in the Savannah River Region by
James B. Stoltman 37
References Cited 63
OFFICERS OF THE SOCIETY
President William M. Goza
P. O. Box 246, Clearwater, Fla. 33515
1st Vice President George Magruder
440 Tenth Ave., Indialantic, Fla. 32901
2nd Vice President John W. Griffin
46 St. George St. St. Augustine, Fla. 32084
Secretary Cliff E. Mattox
P. O. Box 531, Cocoa Beach, Fla. 32931
Treasurer Leon Reyniers, P. O. Box 8451
3301 College Ave. Ft. Lauderdale, Fla. 33310
Editor-Resident Agent Ripley P. Bullen
Florida State Museum, University of Florida
Gainesville, Florida 32601
Three years: Yulee Lazarus
Ft. Walton Beach, Florida
Two years: Wilma B. Williams
One year: Thomas Gouchnour
At large, for one year:
Charles A. Hoffman, Jr.
Benjamin I. Waller,
These papers represent revised versions of four contributions to a seven paper sym
posium held at the 68th Annual Meeting of the American Anthropological Association in
New Orleans, Louisiana, November 20-23, 1969. The intent was to bring together schol
ars who had first-hand knowledge of fiber-tempered pottery-producing cultures in the
Southeastern and Caribbean areas to discuss the "Context, Origins, and Significance" of
these distinctive ceramics. Participants were asked not to concentrate attention too
much upon fiber-tempered pottery itself, which is well described in the literature, but
to focus more upon the total context--temporal, spatial, ecological, and cultural--with
in which this pottery is found in their respective regions: north coastal Colombia (Rei-
chel-Dolmatoff), Florida (Bullen), northern Alabama (Griffin), and the Georgia-South
Carolina coast (Stoltman).
I shall attempt in this preface to anticipate some of the significant and broader
issues raised by the various occurrences of fiber-tempered pottery in the greater Cir-
cum-Caribbean area. The following papers present a great deal of data concerning
these broader issues, and it is hoped that by broaching these subjects here, the reader
will be able to follow for himself the threads of continuity that run through these papers
even in the absence of a summary discussion.
From the time of its earliest recognition in the middle of the nineteenth century
(Wyman 1875), fiber-tempered pottery has held a special interest for archaeologists in
southeastern United States. It was at the shell heaps of the St. Johns valley in north
east Florida, where fiber-tempered pottery was abundant, that C. B. Moore (1892-94),
building upon the pioneering work of Jeffries Wyman, defined the first prehistoric ce
ramic sequence in North America on the basis of careful excavation. At the base of
this sequence was fiber-tempered pottery in an archaeological context that indicated the
manufacturers were hunter-gatherers who had relied heavily upon aquatic resources,
especially shellfish, for their subsistence.
During the latter half of the nineteenth and the early decades of the twentieth cen
tury, the prevailing way that archaeologists conceptualized the past was in terms of
what Stuart Piggot (1960:88-90) has termed the "technological model". Under this model
various ages in prehistory were distinguished by diagnostic technological criteria, the
Neolithic being characterized by polished stone tools and other traits including, espe
cially, pottery. Due in large measure to the writings of V. Gordon Childe beginning in
the late 1920's, the technological model was replaced by an economic-subsistence model
in which the main criteria for the Neolithic shifted from polished stone, pottery and other
such technological traits to food production. Nevertheless, the tendency persisted, even
in Childe's own writings (1951:75-86), to regard pottery as a characteristic Neolithic
trait--an invention of food-producers.
Under such a view, whenever pottery was found in archaeological contexts that did
not suggest food-production, as for example in the Ertebplle shell middens of Denmark,
"neolithic" sites of Siberia, or Early Woodland sites of New York state, it was attrib
uted to diffusion from true Neolithic food-producing cultures. It is not surprising, then,
to see Southeastern fiber-tempered pottery once interpreted as the product of stimulus
diffusion from the northeastern United States, which in turn had received this trait from
Asiatic food-producing cultures via the Siberia Neolithic (Sears and Griffin 1950:2;
Florida Anthropological Society Publication Number 6
Griffin 1952:357; 1960:809).
The advent of radiocarbon dating has drastically altered the picture of the spread of
pottery in North America. For one thing, the Early Woodland pottery of the Northeast
no longer appears as old as -was once thought. A solid-carbon date from the Hunter site
in New York of between 2500 and 2400 B. C. has been supplanted by a newer date run on
gaseous carbon of about 850 B. C. (Bullen 1961:106: Ritchie 1962:584), which leaves no
pottery in the northeastern woodlands dated greater than 1000 B. C. Meanwhile, fiber-
tempered pottery in Florida and Georgia has been consistently dated between 2000 and
1000 B. C. (Bullen 1961). In addition, more recent radiocarbon dates suggest that
fiber-tempered pottery in the Southeast dates as early as 2500 B. C. (Stoltman 1966).
In view of these data Southeastern fiber-tempered pottery can no longer be attributed,
however indirectly, to stimulus diffusion from food-producers to the north.
If one looks to Mesoamerica, an independent hearth of food production and quite
probably of pottery, there is again no currently known culture that could have inspired,
even indirectly, the Southeastern fiber-tempered pottery. Indeed, at the present time
there are no ceramic complexes known in Mesoamerica that have radiocarbon dates as
old as the oldest dated Southeastern fiber-tempered pottery (Brush 1965; Stoltman 1966).
We are thus forced to postulate either an independent origin of fiber-tempered pottery
within the Southeast or a derivation from an as yet undiscovered source somewhere in
Mesoamerica or else look further afield in Latin America for a presumed ancestor.
That Southeastern fiber-tempered pottery may not be a totally indigenous product of
late Archaic cultures in this portion of North America has been suggested by the recent
discoveries of Gerardo Reichel-Dolmatoff on the Caribbean coast of Colombia. In 1961
and 1963 his excavations in the ring-shaped shell midden of Puerto Hormiga recovered
two types of fiber-tempered pottery, along with two types of sand-tempered pottery,
throughout the midden (Reichel-Dolmatoff 1961, 1965). In terms of paste characteris
tics, including manufacture by modeling instead of with coils, the Puerto Hormiga plain,
fiber-tempered sherds are virtually identical to fiber-tempered sherds from southeast
ern United States. What is more, a battery of five radiocarbon dates from Puerto Hor
miga indicates that these Colombian ceramics are comparable in age though somewhat
older than the Southeastern materials. In fact, the Puerto Hormiga dates, which range
from 2552 250 B. C. (LS-1123) to 3090 70 B. C. (ST-153), are currently the oldest
ceramic dates in the New World (Ford 1969: Chart 1).
In view of these data, along with the only slightly younger pottery from the Valdivia
shell middens of coastal Ecuador (Meggers, Evans and Estrada 1965), it now seems
clear that pottery, at least in the New World, can no longer be interpreted as originat
ing only among food-producers. Rather, it is reasonable to infer that sedentariness is
the more basic pre-condition, though not a cause, for pottery manufacture and that food
production is but one of a number of alternative subsistence patterns permitting the at
tainment of a sedentary existence of the sort that could support the manufacture of
fragile, relatively unportable ceramic containers.
Since the New World1 s three early shell midden ceramic complexes (Puerto Hor
miga, Valdivia, and southeastern United States) cannot be derived from any currently
known food-producing, "neolithic" cultures, the interesting question presents itself as
to whether these are totally independent developments or whether there are historical
relationships among them. James A. Ford (1966, 1969) felt strongly that direct his
torical relationships did connect these early ceramic complexes. He devoted the last
years of his life to synthesize the then available data and to search for new evidence in
support of his view that there were a number of culture contacts between Latin America
and the southeastern United States during the Formative Period.
On the time level of these three early ceramic complexes, Ford's survey of the lit
erature, consultation with colleagues, and his own field work in Veracruz failed to un
cover any evidence of human habitation in the areas between them. Particularly relevant
to our main problem is the total absence of sites of the appropriate age, much less of the
appropriate sorts, anywhere along the Caribbean and Gulf coasts between Puerto Hormiga
and southeastern United States.
The issue here is much more significant than fiber-tempered pottery alone. What is
involved is the fundamental question of the mechanism, nature, and significance of cul
tural influences from Latin America on the aboriginal cultures of southeastern United
States. Fiber-tempered pottery is significant, for our purposes, only insofar as it is
relevant to this larger issue. Alone, it must be admitted that fiber-tempered pottery
hardly constitutes adequate evidence for culture contact between these widely-separated
areas; however, considering the overall context within which it is found, the possibility
of culture contact cannot be dismissed out of hand.
There are three features of the overall context of fiber-tempered pottery at Puerto
Hormiga and in southeastern United States that increase the probabilities that genuine
historical ties connect these widely-separated areas:
1. In both regions this pottery is associated with shell middens that reflect a sim
ilar economic-subsistence base~-hunting and gathering with a heavy dependence
upon aquatic resources, especially shellfish.
2. The radiocarbon dates indicate an overlap in age although the Colombian mate
rial begins earlier while the Southeastern pottery endures longer.
3. The two regions are connected by prevailing winds and oceanic currents whose
generally south to north movement is consonant with the time gradient in the
radiocarbon data pertaining to the putatively homologous cultures (Stammel 1965:
Given these shared general conditions, plus at least a few specific resemblances in
content such as the fiber-tempered pottery itself, ring-shaped shell middens (which are
rare or absent in both areas at any other times), and marine shell "hammers" or "picks"
(conch or other large marine shells with the end of the columella rounded or beveled and
the body of the shell carefully broken then perforated to facilitate hafting) that are also
found both at Puerto Hormiga and in the Southeast, and one is led to conclude that the
probability of contact is indeed high enough to warrant serious further investigation.
That this culture contact came in the form of colonizing ventures as Ford (1966:782)
has suggested seems unlikely in light of the current evidence. Both Bullen and Stoltman
emphasize below that in Florida and along the Georgia-Carolina coast fiber-tempered
pottery appears as a non-disruptive additive to a cultural inventory whose local roots are
far deeper than the pottery itself. If actual movements of populations were involved be
tween Latin America and the Southeast, one would expect to find an assemblage of demon
strably intrusive traits indicative of a site-unit intrusion (Willey et al. 1956) somewhere
in the Southeast. No such sites are currently known, although one might theoretically
hold out for the possibility that such sites may eventually be found beneath the waters of
the Gulf of Mexico off Florida's west coast where drowned, fiber-tempered pottery sites
are known to exist (Phelps 1966a; Bullen 1 969a). In the absence of such evidence, how-
ever, fiber-tempered pottery has more the appearance of a trait-unit intrusion in the
Southeast if, in fact, it is intrusive at all and not an independent invention.
James B. Stoltman, Chairman
and organizer of the Symposium
THE CULTURAL CONTEXT OF EARLY FIBER-TEMPERED POTTERY
IN NORTHERN COLOMBIA
Instituto Colombiano de Antropologa
In recent years a sequence of Early Formative phases has been established for the
Caribbean lowlands of Colombia by work at a number of shellmounds which contained the
cultural remains of semi-sedentary, food-collecting groups who inhabited this region
from about 3000 B. C. to 1000 B. C. (Reichel-Dolmatoff 1961; 1965). During the first
half millennium of this sequence, from approximately 3100 to 2500 B. C. a well-devel
oped complex of fiber-tempered pottery made its appearance. It is the purpose of the
present paper to specify the peculiarities of the cultural context within which this par
ticular pottery is found.
Most of the alluvial plains of the Caribbean lowlands of Colombia are characterized
by a very hot and dry tropical climate. The arid to semi-arid savannah country is in
terrupted here and there by groves of shrubby decidious vegetation, while the flood
plains of the rivers are dotted with innumerable lagoons and oxbow lakes. Along the lit
toral there are extensive estuaries, with a fringe of mangrove swamps, lying behind the
wide, sandy beaches, sometimes with deep inlets and a maze of channels connecting the
sea with the lagoons. Rains are concentrated in the months of May-June and September-
October, while during the remaining months a very dry climate prevails.
The food potential of this area is considerable in terms of permanent protein and
vegetable resources. The sea, rivers, and lagoons teem with fishes; the beaches, estu
aries, and inland lakes offer a wide variety of molluscs and Crustacea and, in addition,
an abundance of reptiles such as crocodiles, turtles, iguanas, and smaller lizzards.
Water-fowl, a variety of rodents, deer, monkeys, armadillo, manatee, and small mam
mals are common. From historical sources, we know that as recently as 150 years ago
large tracts of land lying close to the seashore were heavily wooded. There is, too, an
abundance of edible insects or larvae, of honey-bees, and of wild-growing fruits, espe
cially palm trees which bear highly nutritious seeds. Most of these food resources are
not of a seasonal character but are perennial. As concerns farming potential, during the
eight dry months of the year humidity is insufficient for plant growth except under arti
ficial irrigation, and seed horticulture is therefore severely limited, but the country
seems to be exceptionally well suited for root-crops. It is within this type of general
physical environment that most of the sites, all of them shellmounds, are located in the
vicinity of old shorelines or of estuaries, but always close to savannah country and dry
The type-site for fiber-tempered pottery is Puerto Hormiga, a group of shellmounds
situated on the Canal del Dique, a silted-up outlet of the Magdalena River, south of Carta-
Florida Anthropological Society Publication No. 6.
EARLY POTTERY, COLOMBIA
gena. The site is located on the river bank, six kilometres from the seashore and a
short distance from extensive mangrove swamps. The surrounding country is flat and
crossed by many creeks and channels, a group of large lagoons extend toward the south.
About five kilometres distance are the eroded remains of several old river terraces.
The principal mound is a ring-shaped accumulation of clam shells, with a diameter
of 80 metres, which encloses a flat circular area. Width of the ring varies little from
an average of 20 metres and its height above ground level is hardly more than one metre.
The distribution of some slight eminences seems to indicate that originally there were
several small individual middens which eventually fused into a circle by the overlapping
of their bases. There are also indications that what eventually became a circle orig
inally had a bracket-shaped ground-plan with diagonally opposed openings. This pattern,
of slow circular formation, can also be observed at other shellmounds and persists in
time up to at least 1000 B. C.
In the vertical distribution of pottery types throughout the mound, there are several
aspects that must be emphasized. In the first place, fiber-tempered pottery is predom
inant and constitutes the majority of all sherds (70 per cent); it is not just another ware-
type among others, but it is the predominant ware. In the second place, two quite differ
ent types of fiber-tempered wares can be distinguished. The most frequent type contains
long, narrow casts of fibers with a round cross-section. It is never decorated and the
technology of manufacture is rudimentary (Fig. 1, 1-3, 5-6). The other type is sparsely
tempered with crushed flat leaves, is occasionally decorated, and its vessels are techno
logically far better made (Fig. 1, 4). In the third place, both fiber-tempered wares are
consistently associated with decorated sand-tempered wares that are manufactured with
technological skill (Figs. 2 and 3).
This association is present in the basal levels of the mound and has been very care
fully checked. If we project the numerical distribution backward in time, it follows that
the beginning and early developments of these different types goes back far into the fourth
millennium B. C. The Puerto Hormiga site, although having a maximum age of 3100 B. C. ,
does not, therefore, represent an initial stage of pottery development.
Like other early shellmounds in this area, Puerto Hormiga is a habitation site, as
could be observed from several living floors covered with trampled shells and containing
hearths and ash layers. Almost all artifacts or other cultural evidence relate to the food
quest, and the same is true also of all later phases up to at least 1000 B. C. The main
constituents of the mound are shells, most of which (7 5 per cent) belong to the genus
Pitar, frequent on the neighbouring coast. The remainder belong to the genus Ostrea.
Shells show no traces of having been broken or cracked open and were probably boiled.
Occasionally we observe units of 30 or 50 shells with a strong yellow, brown, or reddish
tint that might be due to the addition of some condiment during the culinary process. Ba
sal sherds found near the hearths do not show any exterior blackening. It may be assum
ed that the pottery was not used on an open fire but that cooking stones were employed.
As a matter of fact, heat-cracked stones were found near the hearths. Besides, the size
and shape of the pottery point to the preparation of liquid foods and not to its use in toast
ing or storage.
Other food remains are rather scarce and consist of fish bones, fragments of turtle
shells, bones of small mammals, and fragments of crab shells. Fish bones represent
small to medium-sized fresh-water specimens with only a few bones of sea fish. Turtle
shells belong to two fresh-water species, and the few crocodile bones to small juvenile
RE ICHEL-DOLMA TOFF
specimens. Mammalian remains are limited to small rodents. Turtle shells are
burned on the outside suggesting the entire animal was roasted within its shell. All
tubular bones are split longitudinally.
A variety of stone tool types were also found at the site. The most character
istic are small, flat oval stones each with a natural cavity on one surface. They were
probably used for cracking nuts or seeds. Many are broken in half from having re
ceived too heavy a blow with the massive hammerstones found nearby. Their concav
ities correspond in size almost exactly to the shape of certain very hard palm seeds
that are found in the neighborhood. Several small grinding slabs of sandstone or other
rough-surfaced material have a depression in the middle. Abrasions and striations indi
cate the circular movement of grinding produced by an ovoid pestle, several of which
were found in association with the slabs.
A number of small scraping tools, made of chert, show irregular percussion flaking
aimed at obtaining a short cutting or scraping edge. There is no secondary retouch.
These artifacts may have been used in scaling fish, butchering small game, or scraping
roots, but do not seem to have been manufactured for wood working. The raw material
was obtained from the neighbouring river terraces, but was worked locally as chipping de
bris showing bulbs of percussion is quite frequent in the shell midden. A few heavy, bi
facial choppers indicate rough shaping by percussion.
A number of large, naturally rounded pebbles show traces of having been used topeck
or to grind with a circular motion. A large number of irregularly shaped, natural pebbles
and many sandstone nodules were brought to the mound from the river terraces, but show
no traces of having been used. Small round pebbles, with hard, smooth surfaces, may
have been used as smoothing tools in the manufacture of pottery. No ground or polished
stone artifacts were observed.
Three objects of personal adornment are: a small convex shell plaque with chipped
rim and a circular drill hole in the center, a number of perforated fish vertebrae with
carefully ground and polished periphery used probably as necklace beads, and several
shells filled with a red mineral pigment. Imprints of plaited mats were found on the under
side of a clay layer which had been specially prepared to serve as a base for a hearth.
A number of fairly large Melongena shells had been modified by knocking off part of
the rim of the outer flange, exposing the lower part of the axis. The pointed end of the ax
is was then ground or hammered by use into a rounded shape. Many of the shells thus mod
ified exhibit in addition a round, pecked hole in the conic periphery (Fig. 4). The possible
use of these objects is unknown. There is one oval-shaped celt, made of the thick outer
flange of a Strombus shell, with a well-polished cutting edge.
Although the total artifact assemblage is relatively simple, the food resources were
quite adequate to sustain a band-based society on a semi-sedentary level. The generalmi-
cro-environment of Puerto Hormiga offers three contiguous resource areas: 1} lagoon-
estuary system, 2) savannah system, and 3) dry forest system. The nature of the habita
tion refuse points to highly selective food-procuring practices with the lagoon-estuary sys
tem as the main resource area. But even there, not all available resources were used, if
we can judge from the absence of food remains such as those of large sea fish, sea turtle,
marine shells, and a number of mammals, birds, and reptiles which are frequent in that
At the same time the Puerto Hormiga people had achieved a considerable degree of
EARLY POTTERY, COLOMBIA
cultural stability. During the entire occupation of the site, and throughout the next 1500
years, very little change in material culture occurred, except for certain improvements
in the manufacture of pottery. There were local differences in orientation and use, but
the basic food-collecting level remained the same. In the case of Puerto Hormiga, the
combination of shellfish, small game, palm seeds, and some additional vegetable foods
as suggested by the grinding slabs, seems to have provided a stable basis for life. Even
on a later, horticultural level, there is evidence that people had to forage into far more
distant areas to supplement their vegetable food production, and that even small settle
ments rarely occupied the same spot for long periods of time.
So far, only small-scale excavations have been carried out at Puerto Hormiga, and
it is difficult to ascertain if the living floors correspond to seasonal occupations or are of
a more permanent character. It is possible that absence of food remains from savannah
and forest is indicative of a seasonal occupation of the mound, but the equally marked ab
sence of certain marine and lagoon forms available nearby, speaks against this. Hence
the nature of the refuse contents must be explained rather in terms of selective partici
pation in different resource areas.
The occurrence of fiber-tempered pottery in the Caribbean lowlands of Colombia is
not limited to shore-dwelling shellfish gatherers. Bucarelia, located about 150 kms. in
land on the banks of the Magdalena River, contains fiber-tempered pottery which is close
ly related to that of Puerto Hormiga. At Bucarelia the economic base was neither shell
fish gathering nor small game hunting but river fishing with large bone harpoons and rep
tile hunting in the neighbouring oxbow lakes. Only the more compact and resistant, grass-
tempered pottery was manufactured, along with several sand-tempered wares. The pro
blems of adaptation and of the use of resources were solved here is a very different man
ner from that encountered on the coast.
The great variety of local adaptations, and of marked selectiveness in the use of dif
ferent food resources even within very small habitats, might cast some doubt upon the va
lidity of environmental comparisons. If we may judge from the Colombian cases, it seems
that people who were closely related and existed on the same time-level (or represent a
chronological sequence) might have created very different ways of life by dealing various
ly with the economic opportunities offered to them by their microenvironment. What
seems to remain fairly constant are the technique and artistic development of pottery dec
oration. They appear to exist independently of the particular orientation of the food quest
and to have greater continuity through time and space than the rest of the material culture.
The long sequence from 3100 to 1000 B. C. shows quite clearly that, while local adap
tation varies greatly, the basic trends of pottery techniques, shapes, and decorations,
were continuous and interrelated. It is striking to observe, then, the wide range of pot
tery techniques and decorations on the sand-tempered wares which accompanied the fiber-
tempered wares at Puerto Hormiga. Such elements as shape, details of incised decora
tion, stamping, red pigment fill, and the distribution and combination of motifs on the ves
sels show continuity over 2000 years, while other elements such as elaborate plastic ap
pendages or the use of certain tempering materials are limited to a certain phase only.
From this observation, I would suggest that fiber-temper may well be an independent in
vention which, although often indicative of an early technological level, is not necessarily
related to the other pottery developments which accompany it.
In conclusion, the picture arising from the Colombian data is that of small communi
ties which, while concentrating on one permanent protein base, were highly selective in
RE ICHEL-DOLMA TOFF
their use of other available resources. Non-ceramic artifacts associated with fiber-
tempered wares depend for their typology on the particular pattern of ecological adapta
tion. Pottery, on the other hand, which is found associated with the fiber-tempered
wares is quite elaborate and varied. It shows a wide range of well-defined decorative
techniques that suggest a long previous tradition. Many centuries, if not millennia,
must have passed before 3000 B. C. during which the individual techniques or combina
tions might have developed and diffused.
EARLY POTTERY, COLOMBIA
Fig. 1. Plain and incised sherds, Puerto Hormiga.
, 5-6, fiber-tempered, first type; 4, fiber-tempered, second type; 7-8, sand-
pered; 9, drag-and-jab decoration.
Fig. 2. Zoned and cross-hatched, sand-tempered pottery, Puerto Hormiga.
EARLY POTTERY, COLOMBIA
Fig. 4. Modified Melongena shell (hammers), Puerto Hormiga.
THE ORANGE PERIOD OF PENINSULAR FLORIDA
Ripley P. Bullen
Curator of Anthropology Florida State Museum
University of Florida Gainesville, Florida
The Orange period of Florida is by definition that cultural period during which pot
tery was tempered by the addition of fibrous materials, possibly shredded palmetto fiber,
to clean clay in the manufacturing process. This pottery was first noted by Wyman who
1875:20) asserted, "Palmetto fibre was largely used, mixed with the clay, in the making
of the vessels. Moore (1892-94) demonstrated that such pottery lay over extensive pre
ceramic remains and under those producing fine-grained, check-stamped pottery. Using
Moore1 s collection, Holmes (1894) continued this work by differentiating between fiber-
tempered, the soft chalky or early St. Johns, and the later fine-grained harder St. Johns
ceramics which are frequently check-stamped.
James B. Griffin was the first to recognize the great importance of this ceramic pe
riod, to define its most important pottery type, Orange Incised, and to discuss the impli
cations of the designs found on it. Griffin (1945b:218) pointed out that "it appeared likely
that the style and design of this pottery formed one of the component elements of the Tche-
functe ceramic complex of Louisiana and was hence at least as old as and probably older
than the pottery from that aspect. At that time he even mentioned the presence of St.
Johns Incised as a "descendant" of Orange Incised, noting the restriction in design varia
tion with the new paste.
Ferguson (1951:48), in discussing her 1944 work at South Indian Fields, mentioned
eight decorative similarities between Orange Incised and Stallings Incised but noted none
with Tchefuncte or the Pickwick Basin. Goggin's 1952 summary included no post-1948
field work. In 1954 I divided the Orange period into 4-5 subperiods. This paper will mod
ify or redefine these periods slightly. In 1959 I defined a Florida Transitional period to
cover the time between the close of the Orange period and the advent of the Deptford cul
ture in Florida. Much new information has accumulated since then. Recently in the Sun
day Bluff report (Bullen 1969a), I discussed ceramic diffusion across the Gulf Coastal
Plain during the Florida Transitional period, i. e. the first millennium before Christ.
Here, I propose to present briefly the data behind these generalizations--particularly
the dynamics involved--as evidenced by ceramic changes during and shortly after the Or
ange period. Ordering will be by cultural chronology, not by dates of field work. After
that, I will make a few remarks regarding the origin or origins of Orange ceramics and
close with a brief recapitulation of Gulf Coastal Plain dynamics.
First, I wish to make clear the framework within which we are working. The Orange
period started around or shortly after 2000 B. C. and ended before 1000 B. C. (Bullen 1 961).
The Florida Transitional period began about 1200 B. C. and ended around 500 B. C. with a
fully developed Deptford culture and the establishment of some form of regionalism, cer*
Florida Anthropological Society Publication No. 6.
ORANGE PERIOD, FLORIDA
emonial as well as ceramic, in Florida. These two periods are the temporal and cul
tural equivalents of the Colonial Formative, 3000-1200 B. C. and the Theocratic Forma
tive, 1200-400 B. C. periods recently proposed by Ford (1969:5).
ORANGE PERIOD SITES
In 1951, John W. Griffin and I, both then with the Florida Park Service, with some
help from William E. Edwards, then with the Florida Geological Survey, made a large
stratigraphic test at the extremely large Bluffton site (Fig. 5, 14) on the St. Johns River
in east Florida (Bullen 1955). We found for the first time in Florida a plain fiber-tem
pered midden, 2 feet thick, on the top of nearly 30 feet of preceramic freshwater shell
midden deposits. Sherds did not contain appreciable quantities of sand or other non-fi-
brous temper. Walls were relatively thin, about 6-7 mm. with simple flat-rounded lips.
Shapes included shallow, flat-based, straight-sided, circular bowls and shallow, flat-
based, rectangular containers about 4 by 8 by 4 inches in size.
In the middle of the narrow ends of the rectangular containers, and in one instance on
a curving sherd, lips or upper rims had been pulled outward and a little upwards to form
lugs, ears, or, possibly, pouring spouts (Fig. 6, f-h). In some cases, the latter were
blocked off by added clay to form triangular rim lugs with depressions in their tops. This
shape is, I believe, unique to Florida. It vaguely suggests some steatite vessels but not
those usually found in eastern United States, nor do I know of any instance in which stea
tite vessels or sherds have been found in early plain fiber-tempered or temporally equi
valent deposits. Very likely these rectangular containers are pseudomorphs for previ
ously present wooden or basketry trays. Such containers would have been handy for chil
dren' s use while collecting the small Vivparuas sp. shells which form a dominant part of
the St. Johns River shell heaps.
This plain fiber-tempered ceramic zone was overlain by one which contained decora
ted as well as plain pottery. There was no apparent temporal break, the only noticeable
difference being the addition of incised and punctated decoration and an increase in the rel
ative amounts of circular as opposed to rectangular vessels. Paste and rims were the
same and, as for the lower zone, flat bases were not mat-impressed. This decorative
zone was also about 2 feet thick, but its upper part had been disturbed by the construction
of a house and the planting of orange trees. In the disturbed portion were found a little
St. Johns Plain and Check Stamped pottery, nails, and glass, as well as many fiber-tem
Decoration in the superior fiber-tempered zone was limited to two pottery types: Or
ange Incised and Tick Island Incised. Decoration on the former (Fig. 6, c-d) consisted
of concentric diamonds incised with a single-pointed stylus-like tool when the clay was
still tacky, leaving rounded, 1.5 mm. wide, lines (Bullen 1955: Fig. 2, E-G). Other
straight line motifs, if present, were not on large enough sherds to be recognized. The
Tick Island Incised sherds (Fig. 6, e) were classic in that they exhibited incised spirals
with backgrounds filled with punctations and two incised lines parallel to lips (Bullen 1955:
Fig. A-B). Lines are 1. 5-3.0 mm. wide and, as above, rounded in cross section.
Shapes of Tick Island Incised sherds, except in one case, definitely suggested round
mouthed vessels while those exhibiting concentric diamonds were from both rectangular
and circular straight-sided containers. As these two decorated types were found--as well
as sherds from the eared, rectangular, plain containers--in zones immediately following
undecorated levels, it may be assumed they are typical of the earliest decorated fiber-
tempered pottery of Florida. Ford (1966:786; 1969:171), based on the Bluffton data, has
indicated that Tick Island Incised is earlier than Orange Incised. In Sqs. A1-A5 at Bluff-
ton, Tick Island Incised was the lower in two squares, the concentric diamond variant of
Orange Incised the lower in one square; in one square they had equal depths, while in the
fifth a trench for a drainage pipe obliterated the evidence.
In subdividing the Orange period, Orange 1 refers to the plain fiber-tempered sub
period with round, flat-bottomed, vessels and flat-bottomed rectangular containers having
ear lugs at the ends. Orange 2 deposits contain the same undecorated shapes plus Tick Is
land Incised and that variant of Orange Incised which exhibits concentric incised diamonds.
In passing, it is probably wise to mention that these decorative motifs were not found at
South Indian Fields (Ferguson 1951) nor at the Cotten site (Griffin and Smith 1954), the
only previously excavated and reported sites, nor were decorative motifs present at other
sites found in our tests at Bluffton.
At Bluffton, Busycon (carica) gouges (Fig. 6, k) and Strombus (gigas) celts (Fig. 6, i)
were found in the Orange 1 and 2 levels as well as in the lower preceramic deposits (Bul
len 1955:4). These tools suggest woodworking, perhaps the making of dugout canoes or,
possibly, the digging of edible roots. One broad, stemmed projectile point (Fig. 6, j)
came from Orange 2 deposits.
In an attempt to secure for this symposium a date for the beginning of the Orange 1
subperiod, I sacrificed a shell celt from the 13th level of Sq. A5 or practically the base
of the plain fiber-tempered zone (Bullen 1955:4). This tool was dated at Radiocarbon, Ltd.
(their sample RL-32) at 3,660 110 years B. P. or about 1710 B. C. This date seems too
recent, possibly because the specimen had been in the Museum for 18 years as well as
having been treated with alvar and acetone. It is, however, the earliest fiber-tempered
date for Florida.
It was noted at Bluffton that the number of mammal bones per cubic foot decreased
with the introduction of pottery. This may suggest better utilization of such food by boil
ing or stewing as opposed to roasting and gnawing.
Before leaving Bluffton, I should also mention that we found in our major test 51 frag
ments of marine shells, including tools, in both the fiber-tempered and the underlying
preceramic deposits. Our deepest tool was a Busycon gouge found in Sq. A3 at a depth of
15 feet. A steatite bannerstone in Sq. A1 was 10. 5 feet below the surface or 7 feet below
the lowest sherd in that square. These objects testify to trade with Alabama or Georgia
and with the Atlantic coast 5000 or more years ago. A Strombus gigas celt, found 2 feet
below the lowest sherd, should indicate trade with southeast Florida as such shells are
not known north of Palm Beach. Blufftonians were not entirely ignorant of other areas.
The Bluffton site is now a bay of the St. Johns River a quarter of a mile long, perhaps a
hundred yards wide, and certainly 6 feet deep.
The next site to be discussed is the Hill Cottage midden at the Palmer site in Osprey,
just south of Sarasota, on the Gulf coast of Florida (Fig. 5, 6) There, for the second
time, a plain fiber-tempered zone was found in the upper part of a large Archaic shell
midden (Bullen and Bullen MS). In this case, however, the shells were those of marine
ORANGE PERIOD, FLORIDA
Fig. 5- Map of Florida locating mentioned fiber-tempered pottery sites.
animals as the site is situated on the east shore of Little Sarasota Bay, a rich source of
The hollow horseshoe shape of this midden is most intriguing. Basically, it consists
of two parallel middens, each 400 feet long, which have fused at their southern ends. The
distance across both middens is also 400 feet, while their maximum height is 16 feet. We
made many tests, and the central area between them and between their northern ends prov
ed archaeologically sterile. The answer to this phenomena is simple. The east and west
parts accumulated on the sides of a freshwater spring and the inhabitants were careful not
to block the outlet.
We made one major and three minor tests in this midden, three in the western part
and one, Test D, in the eastern part. Test A, the largest, faced the bay and was carried
down to a depth of 12 feet where water was encountered. In the highest 6 inches were
sand-tempered plain and Norwood Plain (Phelps 1965:66) sherds. The latter were for
merly referred to as "semi-fiber-tempered" plain. These same types plus Orange Incised
were in the next 6 inches, while Orange Incised and a sherd of soft St. Johns Plain were in
the third 6-inch level. Between depths of 1.5 and 2 feet, we found only Orange Incised
while, below, down to a depth of 5. 5 feet, Orange Plain was the only pottery type.
Some of the Orange Incised and Orange Plain sherds from below a depth of 1.5 feet
were not curved and hence indicated flat-bottomed, straight-sided, rectangular containers.
Due to the fragmentary condition of the sherds, incised designs could not be ascertained.
All that can be said is that the lines are straight and either close together or paired and
separated by about an inch. One sherd suggested a concentric diamond design. No punc-
tations or rim lugs were found. In comparison with Orange Incised sherds at Bluffton,
those at Palmer are slightly sandy to the feel. They contain extremely fine quartz grains
but these are not discernible to the naked eye. Orange Plain sherds from the lower levels
do not contain this fine sand.
By courtesy of the Exploration Department of Humble Oil and Refining Company, we
have five radiocarbon dates from the above test as follows:
RADIOCARBON DATES FROM TEST A, HILL COTTAGE MIDDEN,
B. C. Dates
Sample G-598, from a depth of 4 feet, came from about the middle of the Orange Plain
zone. It should represent at this site a time before the introduction of incised decoration,
but it should not be the earliest date for plain fiber-tempered pottery at the Hill Cottage
midden. Extrapolation suggests around 1800 B. C. for the beginning of pottery making at
the Palmer site. The date of 2090 B. C. (2100 using a 1950 base) presented by me earlier
(Bullen 1961) for the beginning of pottery at this site, using Sample 599, was based on a
correlation between of Tests A and B. It may be correct but it is not demonstrably so.
Other tests in the Hill Cottage midden, surprisingly, produced only a few sherds.
Test B, made in the highest part of the midden about 30 feet southeast of Test A, was
carried down to a depth of 6 feet and produced an Orange Incised sherd at a depth of 3. 5
feet. Test C, on the other side of the cottage, was carried down to a depth of 7 feet, but
produced only a sand-tempered sherd at a depth of 9 inches. Test D, in the eastern part,
was excavated to a depth of 5 feet, disclosing an Orange Plain and a sand-tempered plain
sherd in the top 1. 5 feet and 6 sherds, 4 Orange Plain and 2 Orange Incised, at 3 feet be
low the surface.
It should be pointed out that, as at Bluffton, there seemed to be no break at the times
of the introduction of, or changes in, ceramics at the Hill Cottage midden. Non-ceramic
specimens from Test A between depths of 1.5 and 5. 5 feet, i. e. in the Orange 1 and 2
zones, included a large bone point, perforated scallop shells, Busycon picks or hammers,
Strombus hammers, Fasciolaria hammers, columella chisels and hammers, Busycon
ORANGE PERIOD, FLORIDA
pounders, a Strombus gigas celt, a notched net sinker (?) of stone, a sandstone rubbing
stone, and red and yellow clayey nodules. The same tools (Fig. 7) were found in the lower
The colored nodules were probably used for paint. The great numbers of hammering
tools reflects the estuarine locale of the Palmer site where oysters and Venus clams were
available in great numbers. Hammering tools were useful in opening these and other
shells. These tools differ considerably from those found at Bluffton where the freshwater
shellfish are not protected by very thick shells.
Other Gulf Coast Sites
Before leaving the Gulf coast of peninsular Florida, I must mention a very large Ar
chaic marine shell midden located a little south of Englewood (Fig. 5, 7) and about 20 air
miles south of the Palmer site. When I arrived, this midden had been bulldozed to form a
retaining wall against which muck pumped up from Lemon Bay was to collect. I was un
able to find any sherds on or in the disturbed or lower undisturbed levels of this midden.
However, collectors, who had been there earlier during the bulldozer operations, showed
me Orange Plain, Orange Incised, and St. Johns Incised sherds as well as some from
I mention this to show that the Hill Cottage midden was not unique on the Gulf coast of
peninsular Florida. Fiber-tempered pottery is also known from a shell midden on Cayo
Pelieu (Fig. 5, 8) further south (Goggin 1954), while L. Ross Morrell has recently found
Orange and Norwood Plain pottery on Marco Island (Fig. 5, 9) much further south yet.
To the north along the Gulf coast, Orange ceramics are known for Perico Island (Willey
1949), in the Withlocoochee river valley, and at Shired Island north of the mouth of the
Suwannee River (Fig. 5, 5, 3, and 2; Florida State Museum collections). I mention these
sites to demonstrate that the Orange period was not limited to the St. Johns River valley
in east Florida.
Old Town, Fernandina
Returning to the St. Johns, I wish to mention briefly another plain fiber-tempered
pottery site. During John Griffin's and my survey of Amelia Island, a small stratigraphic
test was made in an oyster shell midden immediately north of the Plaza Lot, in Old Town,
Fernandina (Fig. 5, 11). The midden, where tested, was 30 inches thick and contained
Orange Plain sherds throughout (Bullen and Griffin 1952:51) plus a Busycon pick.
Summer Haven, Cotten, and South Indian Field Sites
Summer Haven, Cotten, and South Indian Field (Fig. 5, 12, 16 and 18), are all typical
of what I will designate as the Orange 3 subperiod. South Indian Field is an inland site,
surrounded by swampy land, and located near the headwaters of the St. Johns River. Cot
ten is situated on the west bank of the Inland Waterway a little north of Daytona Beach.
Summer Haven is located on the east side of the Inland Waterway, some 15 miles south
of St. Augustine, immediately south of Matanzas Inlet. The last two sites, at least, re
flect expansion from the St. Johns River valley in Orange 3 times.
Summer Haven was a large site substantially destroyed many years ago by the re
moval of shells for road building. We worked in an eastern, undisturbed portion of the
site, which was to be affected by a widening of Route A1A (Bullen and Bullen 1961).
Here the midden was 3. 5 feet thick and contained Orange Plain and Incised pottery except
for a Norwood Plain and a St. Johns Plain sherd in the highest 6 inches.
While some of the sherds were curved, indicating fairly large, straight sided, round
mouthed, vessels with flat bottoms; others were not curved and hence from square or rec
tangular, flat-bottomed containers. Both were about 4 inches deep. A few miniature ves
sels were also flat-bottomed, but had round mouths and in-curving rims. Paste was like
that at Bluffton, but about one per cent of the sherds, all from middle zones, contained
fine limestone inclusions (Bullen and Bullen 1961:4).
Walls were 4 to 13 mm. in thickness, noticeably thicker than for Bluffton, bases
13 mm. thick, and junctions between bases and walls very thick, almost 25 mm. Many
walls were wedge-shaped in that they gradually increased in thickness with distance from
their bases. Lips were simple and slightly rounded or wide and flat. Only one rim sherd
exhibited an ear. Of 354 rims, sufficiently complete for analytical treatment, 181, or 51
per cent, were decorated. Of 130 wide and flat lips, 105 were decorated. In only one case
was the lip but not the vessel wall decorated.
Decorative motifs and designs had a wide range of variation (Fig. 8). They were
composed of various combinations of straight lines, punctations, and ticks. No curved
lines, punctated backgrounds, or concentric geometric figures were present. Designs
included those made of 6 to 9 parallel slanting lines, sometimes bordered by ticks, cross-
hatching, pendant hatched triangles, occasionally narrow hatched bands, and, amazingly,
large running frets (Bullen and Bullen 1961). Rim decoration, to some extent, correlated
in design with decoration on vessel walls (Fig. 8, j, 1). Incising was done with a rounded
or flat-rounded tool and varied in width from 1.5 to 3. 0 mm. Only a few exhibited V-
shaped incisions. Decoration on lips was always better done and, at least usually, with
a different and narrower tool.
An apparently unique feature of Summer Haven ceramics was the presence of a few
rims with rounded tops. These were decorated with rows of punctations while their walls
bore incisions (Fig. 8, i). A similar oddity was what appeared to be lug handles located
a short distance below lips in about the location where similar lugs frequently occur on
steatite vessels. These lugs were decorated with rows of punctations, while above and
below the lugs incised lines were present (Fig. 8, g-h).
It should be pointed out that in every attribute mentioned above, Summer Haven ceram
ics differ from those previously described for the Bluffton and Palmer sites. While there
are some ceramic features at Summer Haven that differ from those reported for the Cot-
ten and South Indian Field sites (Griffin and Smith 1954; Ferguson 1951), all three sites
are remarkably close ceramically and must belong to the same subperiod--Orange 3.
However, the Cotten site midden was substantially thicker than that at either of the other
two sites, and it was undoubtedly inhabited after abandonment of that part of the Summer
Haven site tested by us.
At all three sites, sherds of steatite vessels were found at substantial depths but not
in basal levels. At Summer Haven, steatite sherds were in Levels 3 and 5 of Square B.
A marine shell from Level 4, between the steatite bearing levels, was radiocarbon dated
at the University of Michigan. The date (Sample M-1014) was 3330 200 B. P. or about
1380 B. C. It should approximately date the apogee in Florida fiber-tempered ceramic
development. The presence in the same or adjacent levels of fine limestone inclusions in
pottery and of steatite sherds are suggestive of extra-areal connections. It should also be
ORANGE PERIOD, FLORIDA
noted, in this connection, that fiber-tempered vessels from Summer Haven are more si
milar to steatite vessels than were the rectangular containers from Bluffton.
Careful analyses of sherds from Summer Haven indicate a tendency towards simpli
fication in decorative designs with time (Bullen and Bullen 1961). This would seem likely
as incised designs during the next period, Orange 4, are much reduced in variety and
complexity when compared with those found at the Summer Haven, South Indian Field, and
Other specimens from Summer Haven include bone pins, awls, points, a fish hook,
and shell gouges, celts, hammers, and picks (Fig. 8, m-n). Most of these, if not all,
are holdovers from the previous preceramic period. All, except the hammers, are si
milar to those found at Bluffton but differ from those uncovered at the Palmer site. The
increase in hammers at Summer Haven, as opposed to Bluffton, reflects the need to open
oysters and marine clams.
Similar tools, particularly shell celts or adzes and shell gouges, were also found in
Orange 3 deposits at the Cotten and South Indian Field sites. Some of the most diagnostic
artifacts of this period are bone pins, sometimes quite large, which are embellished with
extremely well executed incised decoration. Motifs are the same as those found on fiber-
tempered vessels of the Orange 3 subperiod. Frequently such pins have expanded heads
(Griffin and Smith 1954: PI. II, 2-3). Other artifacts illustrating similar decoration are
several turtle carapace marginal bones with beautifully incised complicated interlocking
T designs (Griffin and Smith 1954: PI. II, 6-7). These fragments bear drill lashing holes
and presumably represent complicated artifacts. Alternatively, they might be from the
edges or sides of decorated turtle shell rattles. Supposedly, atlatls and other wooden
artifacts were similarly decorated.
Four stemmed points came from Level 4 at South Indian Field (Ferguson 1951:41-42;
PI. 4, I). With straight blade edges, rounded but drooping barbs, and rounded tangs,
these points are a little smaller than those illustrated from Zone 9 at Site J-5 (Fig. 11)
which, however, they resemble except for the rounding of blade and stem corners. Small
to medium sized stemmed points seem the only projectile point type of the Orange period
in Florida. The larger stemmed points (Fig. 11, p-q) were undoubtedly hafted knives.
Before concluding discussion of the Orange 3 subperiod, I should comment on the
food remains found at Summer Haven (Bullen and Bullen 1961:12-13). There was some
alternation in the deposition of coquina and oyster shells in the midden. This may re
flect seasonal occupation or ecological changes in the immediate environment. Shark,
turtle, fish, and deer bones were present in large quantities; bird, raccoon, alligator,
rabbit, dog, and snake bones in reasonable quantities; and crab claws, bear, opossum,
otter, wildcat, and wolf bones in minor amounts. In both quantity and numbers of spe
cies this is a much broader list than those from either Bluffton or the Palmer site. It
indicates a better adaptation to the environment and a greater utilization of the biota.
Hunters must have ranged over a considerable area in their food quest. This greater
territoriality coincides with the introduction of steatite vessels to mark a broadening
Bird bones were particularly interesting as they included those of the extinct great
auk, indicated winter occupation of the site, suggested the climate was a little cooler than
now, and implied the inhabitants must have had some form of water transport. Dr. Pierce
Brodkorb, who identified the bird bones, commented that all four of the water birds--loon,
gannet, great auk, and common murrewould be expected perhaps a mile off shore and
not inside Matanzas Inlet where the site is located.
This multicomponent site, containing the first "pure" Orange 4 subperiod deposits so
far reported, consists of a series of small shell middens located along a small tributary
of the St. Johns River (Fig. 5, L3)* our tests, the compact shell midden proper varied
from 15 to 24 inches in thickness; was underlaid by soft sand containing scattered prece-
ramie Archaic points, knives, and scrapers; and was overlain by soft sand or mixed sand
and shell deposits containing St. Johns, Deptford, and other, more recent, pottery (Fig. 9
a.-d, f-g) and projectile point types.
The shell midden proper, wherever penetrated by us, was compact and contained Or
ange Plain and Incised sherds (Fig. 9, h-k) to the exclusion of other pottery except for
some soft St. Johns Plain sherds, usually in higher levels (Bullen 1969a). Orange Plain
and Incised sherds from Sunday Bluff pertain chiefly, but not exclusively, to flat-bottomed
containers with fairly thin walls, 6-9 mm. in thickness. They do not contain appreciable
amounts of sand or other non-fibrous temper. Some rim sherds are incurved, but most
are from straight-sided vessels. Wedge-shaped rims and thickened, flat, decorated or
undecorated lips, typical of the Orange 3 subperiod, are absent.
Incised designs are extremely simple (Fig. 9, h-j). Alternatively hatched areas are
present, but no complicated patterns such as concentric rectangles, zoned hatchuring,
running frets, ticks as design elements or even composite straight line designs were found
Lines paralleling lips and setting them off from design elements on vessel walls--usually
a prominent trait on Orange Incised vessels--are infrequent. This situation should be ex
pected in a final fiber-tempered horizon because it is only these simple designs--not the
complicated ones found at Summer Haven, South Indian Fields, and the Cotten site--that
are carried over onto the St. Johns Incised and Pasco Incised containers of the succeed
ing Florida Transitional period.
The only exception is a sherd with a wide, incised, flat lip and a paneled decoration
on its side. This sherd (Fig. 9, m), typical of the earlier Summer Haven ceramics, was
found at the very base of Test E in the thickest part of the compact midden proper. Anothe
variant was a punctated fiber-tempered sherd exhibiting three parallel rows of rectangu
larly shaped punctations, sloping alternately to the right and to the left (Fig. 9, 1). This
sherd also came from the base of the shell midden but in a different test. Its design is
somewhat reminiscent of Stalling1 s Island pottery motifs.
That the midden proper at Sunday Bluff represents the terminal fiber-tempered phase
of northeast Florida is indicated by the similarities in shapes and decoration to those of
the succeeding St. Johns Plain and Incised vessels (Fig. 9, f-g), and by the presence in
the midden proper of some St. Johns Plain sherds. This situation--plain chalky ware in
otherwise fiber-tempered levels--has been reported for both the South Indian Field and
Cotten sites (Ferguson 1951:34-35, 38; Griffin and Smith 1954:33, 42, Table IV).
At Sunday Bluff, we were particularly careful to scrape off the surface of the compact
midden proper before penetrating it and are rather positive that the St. Johns Plain sherds
shown in our tables (Bullen 1969a: Tables 1-8) were actually found as indicated. At the
nearby Colby site, Dr. Thomas Gouchnour of Jacksonville also found St. Johns Plain
sherds in fiber-tempered levels of his main stratigraphic test. Indeed, he also found
ORANGE PERIOD, FLORIDA
chalky-fiber-tempered pottery and St. Johns Plain sherds containing a little fiber temper
ing in appropriate levels with a vertical distribution averaging deeper than that of St. Johns
Plain sherds without such temper (Bullen 1969a:39). I accept these data as satisfactory
evidence that the chalky St. Johns ceramics represent an indigenous northeast Florida de
velopment, as previously suggested by Goggin (1952), and that their manufacture started
before the close of the Orange period or during Orange 4 times.
Zone 9 at Site J-5
One more late site needs to be discussed. Whether it belongs in an Orange 5 subpe
riod or in the Florida Transitional may be debated. The presence of St. Johns Incised
sherds suggests the latter period. I am referring to Zone 9 at Site J-5 on the bank of the
Chattahoochee River, just above the Forks, in northwest Florida (Fig. 5, 1).
Zone 9 was a midden deposit, about a foot thick and separated from a higher Deptford
period midden by 3 feet of sterile, water laid, deposits. Specimens (Figs. 10-11) from
this closed midden included 185 Orange Plain, 1 Orange Incised, 14 St. Johns Plain, 1 St.
Johns Simple Stamped (?), and 3 St. Johns Incised sherds; 7 steatite fragments from 3
different bowls; river mussel and snail shells; and 47 chert tools, chiefly medium-sized
stemmed points with drooping barbs; and hafted scrappers with similar basal details
Some Orange Plain pottery represented heavy, flat-bottomed vessels with thick walls
that gradually tapered upwards to a simple lip, 6 mm. wide (Fig. 10, n). Thin, straight
sided rim sherds seemed more abundant (Fig. 10, k, m). One thick sherd exhibited an
"Adena heel" and another a woven splint basketry imprint (Fig. 10, 1, n). Some fiber-
tempered sherds contained a small amount of micaceous material indicating local manu
facture as clays in the Chattahoochee-Apalachicola valleys contain mica fragments. A
few contained some grains of quartz sand, but most, like those of peninsular Florida, only
had casts of fibrous material as evidence of temper.
The chalky paste of the St. Johns series pottery did not contain micaceous materials
and, presumably, was imported from peninsular Florida. The St. Johns Simple Stamped
sherd (Fig. 10, i) may be "linear check stamped" as opposed to "simple stamped." Lines
on this sherd do not seem to be incised, but they might have been made with a drag-and-jab
technique. It varies in thickness from 4 mm. at one end to 13 mm. at the other while the
shape of its inner wall shows it came from a flat-bottomed vessel. Decoration on St.
Johns Incised vessels (Fig. 10, j-k) are simple and made up of straight lines.
Fiber-tempered sherds from J-18 and other sites in the Woodruff Reservoir area
were similar to Orange Plain, i.e. they did not contain either quartz sand or micaceous
materials (Bullen 1958b). No decorated fiber-tempered pottery was found in that basin
other than the one sherd in Zone 9 at Site J-5.
Charcoal from Zone 9 at Site J-5 was radiocarbon dated at Michigan University. The
resultant date, Sample M-394, was 3150 250 B. P. or about 1200 B. C. This is the lat
est date we have for fiber-tempered pottery and the earliest date for St. Johns Incised.
These data clearly indicate that the St. Johns series was being made when fiber-tempered
containers were still being used. However, I would use the presence of St. Johns Incised
(or Pasco Incised, its limestone-tempered equivalent) as the signal that the Florida Tran
sitional or, in Ford's terminology, the Theocratic Formative had started.
FLORIDA TRANSITIONAL PERIOD SITES
That the Orange period was entirely ended by 1000 B. C. is indicated by various data,
most significantly by those from the Zabski site on Merritt Island in extreme eastern Flor
ida. Previous to the work at the Zabski site, I used the 970 B. C. date (Sample M-267,
2920 200 B. P. ) from the Refuge site in Georgia (Bullen 1961) for the end of fiber-tem
pered times. According to Waring, the sample for this date came from a little above the
point in the Refuge midden where sand-tempering replaced fiber-tempering (Williams 1968:
198-99). The closeness of this date to that from the Zabski site, to be presented shortly,
More recently, Phelps (1966a: 19) has secured a date from the Florida State Univer
sity' s Radiocarbon Laboratory on Norwood series sherds from the Tucker site in West
Florida of 2962 120 B. P. or about 1012 B. C. This ceramic series contains both sand
and fiber as tempering materials (Phelps 1965). Such pottery--usually plain, but some
times simple stamped--is known for peninsular Florida where, so far, it has always been
found either in the top of or above Orange period deposits as at the Palmer site (Bullen
and Bullen MS), Summer Haven (Bullen and Bullen 1961), and Sunday Bluff (Bullen 1969)
sites or in association with Florida Transitional pottery types as at Bayport on the Gulf
coast (Bullen and Bullen 1953; Coates 1955). If these proveniences are correct, the above
date, around 1012 B. C. should apply to the close of the Orange period.
The Zabski site (Fig. 5, 17) was originally a closed, one-period midden buried in a
high bluff beside the Banana River (Atkins and MacMahan 1967). Recently the bluff has
been substantially eroded by the river during times of storms. A sample of charcoal from
the remaining uneroded part of the site was radiocarbon dated at Florida State University
and the assay (Sample FSU No. 190) yielded a date of 2910 80 B. P. or about 960 B. C.
At the Zabski site not a single fiber-tempered sherd, nor any of the Norwood series, has
been found in the midden, on the bluff, or in the eroded deposits in the river beach.
Ceramic specimens included St. Johns Plain, Incised, Pinched, Triangular Punctated,
indented, and side lugged (with lugs pushed from the inside outwards); Pasco and Perico
Plain, Perico Linear Punctated (limestone-tempered); and sand-tempered plain sherds
plus fragments of steatite vessels (Atkins and MacMahan 1967: Figs. 4-8). St. Johns ce
ramics at Zabski are fairly thick, soft, and occasionally contain bits of crushed sherds.
The same variety of St. Johns Plain, but no fiber-tempered pottery, was found at the
Cato site, a shell midden presently covered by beach sand and below mean tide level on
the Atlantic shore near Yero Beach, Florida. A Cato site radiocarbon date of 2795 50
B. P. (FSU Sample No. 173) or about 825 B. C. gives a good check on the above Zabski
site date (Bullen, Bullen and Clausen 1968).
Zabski is the first major closed Florida Transitional site investigated, and the range
of pastes, ceramic decoration, and their implications are most interesting. Decorative
motifs on St. Johns Incised are the simple straight line designs mentioned earlier without
any frets, ticks, or decorated lips. Some of the design elements are identical to those
found on Tchefuncte Incised (Ford and Quimby 1945: Pl. 3, d-j). One sherd (Atkins and
MacMahan 1967: Fig. 5, k) has a handle lug, identical to those found on steatite vessels,
on its side about 2 inches below the lip. The St. Johns Pinched and fingernail "punctated"
sherds are practically identifiable as Tammany Pinched of Louisiana.
Triangular punctations are arranged in panels. This arrangement, previously un-
ORANGE PERIOD, FLORIDA
known for Florida, occurs on Stallings Island pottery. The limestone-tempered Perico
and Pasco sherds show close connections with the Gulf coast of peninsular Florida where
such ceramics are well known and probably originated.
Presence of these various traits at the Zabski site indicates the movement of many
cultural influences across Florida and along the Gulf Coastal Plain in the immediately
post-Orange period. They probably started earlier during the concluding phases of the
One trait missing in the Zabski site collection, but well known for the Florida Tran
sitional and sometimes found on fiber-tempered pottery, is that of basketry or woven mat
imprints on the bottoms of flat-based vessels. Possibly this trait, known earlier, may
have been discontinued by the time of occupation of the Zabski site.
I have completed above the presentation of pertinent data regarding the Orange period
of peninsular Florida with primary emphasis on chronology and interareal connections. I
have said little about the geographical extent of Orange culture except to indicate it was
present along all the Gulf coast of peninsular Florida. Undoubtedly, Orange period sites
will be found out in the Gulf of Mexico. There the water is very shallow and two inundated
Archaic sites are known some distance out from the present shore line (Carl Clausen and
Wilfred T. Neill, personal communications).
To complete the picture, fiber-tempered pottery (Fig. 12) is also found at Bolen Bluff
(Fig. 5, 10) and other sites in the interior of the peninsula (Bullen 1958a), albeit spar
ingly. On the east coast, the Orange culture area extends from Fernandina or the Georgia
line at least to Vero Beach, some 300 miles, and for the end of the period, Norwood Plain
sherds have been found in fair quantities at the base of the Peace Camp site (Fig. 5, 19)
on the edge of the Everglades west of Hollywood, only 20 miles from Miami, by the Bre
vard County Chapter of the Florida Anthropological Society.
In spite of this large areal distribution, the heartland of Orange culture--in terms of
number and size of sites, and complexity of decoration--is in the middle portion of the St.
Johns River valley. This river, like many low gradient streams, flows in a meandering
manner and, in its middle course, connects several large lakes. Also in the middle val
ley are to be found many "dead streams, cut-off oxbows, and filled-in channels. This
is and was an ideal region for the growth of Viviparaus shellfish. The CaCO^ removed
from the water by the animals during growth is continually being replaced by limestone
(carbonate) impregnated water flowing from innumerable springs.
Along the western edge of the middle St. Johns River flood plain, southwest of Tick
Island, are a series of fairly large Viviparaus middens, each roughly 100 yards long, 100
feet wide, and 8 to 1 8 feet in height. Totally devoid of pottery, except for that left on the
surface by casual camping visitors during the Florida Transitional period (Bullen and
Bryant 1965), these mounds are the despair of archaeologists as excavation produces
little or no useful cultural inventory. They have been dated to about 3000 B. C. (Bullen
and Bryant 1965:23) and occupation probably started around or before 4000 B. C. As the
St. Johns River moved to, or nearer to, its present channel, these middens were aban
doned and new ones started at new locations, although there must have been a long per
iod of overlap in occupation.
The new middens along or near the present channel of the river grew to be extremely
large in size. Tick Island (Fig. 5, 15), one of the most famous, covered about 10 acres.
They are composed of relatively more mussel and large snail (Pomacea sp. ) shells and
contain more food bones than do the earlier middens. They also occasionally contain ma
rine shell tools and other exotic items indicating a wider area of contact.
In the highest levels of these middens we find flat-bottomed fiber-tempered clay con
tainers both plain and decorated with straight line incising. In the only large midden tes
ted with modern stratigraphic techniques, Bluffton, we found a plain period below a dec
orated one in which decoration was limited to Tick Island Incised and a concentric dia
mond variety of Orange Incised. While various forms of Orange Incised cover most of
peninsular Florida, Tick Island Incised is limited to Tick Island, Bluffton, and a few
other nearby sites as Goggin (1952) pointed out years ago.
This brings us to the question of the origin of fiber-tempered pottery as found in Flor
ida (Bullen I960). Certainly the Orange period, as such, is only one stage in a local con
tinuum as I have gone to some lengths to demonstrate. The only new item is the introduc
tion of pottery. Except under the concept of stimulus diffusion, the introduction of pottery
making from outside the area might be expected to be accompanied by some other new trait.
New non-ceramic traits are not to be found. Projectile points, while smaller and perhaps
relatively broader than those of the Late Preceramic Archaic period, are stemmed and
tend to have drooping barbs. They resemble Culbreath points believed to be the popular
point at the very close of the late pre-Orange phase of the Archaic (Warren, Thompson,
and Bullen 1967). The typology of shell tools tends to be identical before, during, and
after the Orange period.
Perhaps this is the place to mention possible Florida pre-ceramic cooking vessels.
Recently, William J. Webster of Jacksonville, Florida, has brought to my attention the
fact that fire-damaged Busycon vessels (large hollowed-out conch shells) are known for
many of the preceramic shell middens of the St. Johns River valley, while shell tools from
the same middens do not show any fire damage. Webster (1970:1-5) contends these are
cooking vessels and his evidence is very persuasive. If this is correct, we have here a
precursor for clay cooking vessels. These middens contain extensive fire hearth areas
and stones are not available to keep vessels upright. The substitution of a flat-bottomed
clay vessel for Busycon containers on these hearths would be a major technological im
provement and may represent a local development.
Any hypothesis regarding the origin of Orange ceramics--whether it represents in
dependent invention or diffusion, and if so from where--depends on a consideration of four
factors: temper, shape, decoration, and dates of manufacture. Carbon-14 dates clearly
indicate the impossibility of stimulus diffusion from the north (where steatite bowls have
a slight local temporal priority over pottery), as there are no dates for steatite bowls as
ancient as those for Southeastern fiber-tempered pottery.
To the south, one has to go as far as South America to find ceramic cultures with
dates as old as those of the Orange and Stallings Island cultures of Florida and Georgia.
Even the extremely early, 2500 B. C. dates for Rabbit Mount (Stoltman 1966) in South
Carolina leave priority with the Valdivia and Puerto Hormiga cultures of South America.
In passing we should note that the earliest known pottery in Mexico appears in the Purron
phase of the Tehuacan sequence between 2300 and 1 500 B. C. (Ford 1969:18) or at 2440
140 B. C. at Acapulco (Brush 1965) .
ORANGE PERIOD, FLORIDA
Fiber-tempered pottery is found in at least six places in the world: the Norton cul
ture of Alaska, the shell heaps of the Malay Peninsula (Samrong Sen culture), early Jomon
pottery of Japan, the Caribbean coast of Colombia (Puerto Hormiga), the Florida-Georgia
region of southeastern United States, and the Hebrides Islands off the west coast of Scot
land (specimens seen at Edinburgh in I960). In all cases it is the earliest or among the
earliest pottery in each region; that from the Hebrides is, interestingly, dated at 1000
BC. Without resorting to an outmoded Kulturkreis hypothesis, I cannot see diffusion
connecting all these instances. However, I must admit that, according to available C-14
dates, diffusion from Japan to Colombia to South Carolina to Georgia to Florida and to
the Hebrides is possible.
I am also impressed by other facts that refer primarily to decoration. In Current
Anthropology, October 1968, pp. 319-23, under Research Projects, Figure 5 illustrates
linear punctated and dentate stamped pottery including rocker stamping for five sites in
Northern Sudan. It is stated that the pottery is thick and "most sherds are chaff-tempered
(i. e. fiber-tempered), although some sand- and micaceous-tempering" are also present.
Marks, Shiner, and Hays (1968:321-23) state "The pottery suggests a date for this group
as early as any ceramic site near Khartoum" and that occupation occurred "during the 4th
millennium B. C. or between 3000 and 4000 B. C.
Zoned Punctated pottery is on exhibit at Toledo, Spain, for the Eneolithic period,
circa 2300 B. C. Incised zoned cross-hatched pottery is exhibited for the early dynastic
periods of Egypt in the Cairo Museum (cat. nos. 10904-5). Exhibits at the Museo Ar
queolgico Nacional in Madrid show Neolithic pottery decorated with incised straight lines.
In one case they are simple slanting lines extending downward from lips, in another cross-
hatched panels with horizontal parallel lines separating them from lips. Both are good Or
ange Incised decoration. In Archaeological Discoveries in the Holy Land, Perrot (1967:5)
illustrates stone vessels with incised hatched triangles on inner rims with a date of around
3500 B. C. Kenyon (1967:25) in describing the first pottery at Jericho around 4500 B. C.
says that the finer bowls were decorated with painted straight line chevrons.
The first pottery of Northern Khuziston, east of Mesopotamia, found in the 6000 to
5600 B. C. Mohammad Jaffar phase, is described by Hole, Flannery and Neely (1971:282-3)
as "soft, friable, and straw-tempered. Second, there was a variety of deep bowl whose
surface was decorated with geometric motifs in fugitive red ochre paint; zig-zags, chev
rons, or pendant triangles and lozenges. A few of these pots had small troughlike spouts
at the lip. Except for paint instead of incision, this description--including the trough
like pits--fits very well the early Orange period pottery illustrated in Fig. 6.
The point I am trying to make is that both fiber-tempering and simple rectilinear in
cised or painted decoration are typical of the earliest pottery in many places. This is not
an argument for the psychic unity of man. It does suggest that there is no monopoly of
brains and that intelligent people given reasonably similar backgrounds and closely similar
problems are apt to arrive at somewhat similar solutions. Similarities in temper or gen
eral similarities in decoration do not seem to me to be sufficient by themselves to prove
diffusion over large distances without way stations as intermediate donors.
There are certainly some similarities between decoration found at Valdivia on the
Pacific Coast of Ecuador, particularly that found on the relatively late Ayangue Incised
vessels (Meggers, Evans, and Estrada 1965: Pis. 131-34), and that found in the Orange
period of Florida, but vessel shapes are extremely different; moreover, Valdivia pot
tery is so sophisticated and contains so many rather different ways of handling decora-
tion that I find it hard to see any direct connection. The similarities between Stallings
Island pottery from Georgia and Valdivia Multiple Drag- and Jab-Punctate (Ford 1966:
785) seem, to me, to be closer than those just mentioned for Florida. That this type is
earlier than Ayangue Incised at Valdivia should give satisfaction to any diffusionist as will
the radiocarbon date for the Machalilla Phase, 2200 to 1500 B. C. (Evans and Meggers
1965: chart facing p. 8).
Reichel-Dolmatoff, some years ago, very kindly sent me samples of Puerto Hormiga
pottery. Sherds from there tempered with long thin fibers (Reichel-Dolmatoff: 1965:60)
could easily be lost in any peninsular Florida fiber-tempered collection except that they
contain a few grains of quartz and the diameters of the fibers, i.e. the holes left by them,
are noticeably larger than is the case for Orange ceramics. The quartz grains are fairly
large but sparse, and are probably natural inclusions in the clay, not intentionally added
temper. The other type from Puerto Hormiga, exhibiting casts of leaves or grass and
containing small brown bits of vegetable matter, is foreign to Florida. The sand-tem
pered sherds of Puerto Hormiga, of course, have no counterpart in early Orange period
ceramics of peninsular Florida.
Direct diffusion from Puerto Hormiga to Florida seems unlikely because of differ
ences in vessel shapes (circular mouths, round bottoms at Puerto Hormiga, rectangular
flat-bottoms in Florida) and decoration (Reichel-Dolmatoff 1961). Current radiocarbon
dates indicate the plain fiber-tempered pottery of Georgia precedes that of Florida. Dif
fusion of the idea of fiber-tempered bowls from Puerto Hormiga or a similar site to Geor
gia and from there to Florida is a distinct possibility that would fit in better chronologi
cally and avoid the shape and decoration strictures; but one would still wonder why, of the
temper types available, it was fiber and not mineral tempering that diffused.
This theory would seem to require the presence in Colombia of a pre-Puerto Hor
miga site with nearly 100 per cent fiber-tempered pottery and a much simpler decora
tive assemblage. If decoration at such a site were limited to rows of punctations and
cross-hatched lines like certain illustrations in the Puerto Hormiga report (Reichel-
Dolmatoff 1965: Lams. Ill, 1-2; V, 7), it would substantially increase the likelihood of
diffusion from Colombia to the southeastern United States.
Ford (1966: Fig. 3, a-e) has pointed out the similarity between Tick Island Incised
and some of the Barlovento incised and punctated sherds of the Caribbean coast of Co
lombia. While there are general similarities, it is a question whether the degree is suf
ficient for a specific connection. Barlovento designs have pendant garlands or incised
ribbons ending in one external coil of paired, single turn, involutes. Background areas
are filled with angled, ovate punctations or curved parallel lines. Of considerable impor
tance, to my mind, the zoning lines and the main feature are composed of very broad
shallow lines. These occur on 26 of the 31 sherds Reichel-Dolmatoff (1965: Lams.
III-V) illustrates. The other 5 sherds clearly represent a different, presumably minor,
design motif. Workmanship is excellent and sherds exhibit a "smooth" feeling.
There is nothing in Orange decoration that remotely resembles these wide shallow
lines or the flowing impression of garlands. Punctations on Tick Island Incised also dif
fer from those found at Barlovento in being round, deep, and made with a tool held ver
tical to the wall of the vessel. Also, each involute is made of a double, instead of a sin
gle line, and spirals outward by three rotations instead of one. It starts from the center,
which the ones from Barlovento do not do (Ford 1966: Fig. 3, a). Lines of the spiral are
wider than is the case for Orange Incised, but they do not approach those from Barloven-
ORANGE PERIOD, FLORIDA
to. They are plowed in when the paste is damp. Executionis much poorer, and involutes
are not paired.
I, obviously, fail to see a close connection in either concept or execution between
Tick Island Incised and Barlovento incised and punctated pottery. On the other hand I can
point to no other possible place of origin. If Tick Island Incised is a copy of Barlovento
pottery, it is a very poor copy. One should also note that while fiber-tempered pottery is
present at Puerto Hormiga, Barlovento ceramics are sand-tempered. While Tick Island
Incised is relatively early in the Florida sequence, it is areally limited to the middle St.
Johns River valley. These data do not help the case for a Barlovento origin for Tick Is
land Incised designs.
While I seem to have taken an anti-diffusionist attitude in this matter, that is not a
set position on my part. At present my conclusion is one of "not proven. More work is
needed along the Gulf coasts of South and Central America as well as Mexico, Texas, and
Louisiana. Such work could well prove many connections between Colombia and Florida.
Investigators should also realize that the waters of the Gulf have encroached upon the land
a substantial distance since 2000 B. C. 4 meters for southwest Florida (McDowell, Step
hens, and Stewart 1969:745), and the sites we are looking for may be some distance out
into the Gulf. A drop of 1 5 feet would in many places move the Florida coast line out
over 7 miles. Undoubtedly dugout canoes were available as one in Florida has been
radiocarbon dated to 1090 B. C. (Bullen and Brooks 1967).
Returning to culture changes in Florida, we have the introduction or independent de
velopment of fiber-tempered pottery, sometime between 2000 and 1700 B. C. to or by Ar
chaic people who had been exploiting the freshwater shellfish of the middle St. Johns River
valley and the shellfish-producing waters of the Gulf for an extremely long time. The fol
lowing Orange period has been divided into four subperiods based on differences in cer
amic shapes and decorations. During Orange 3 times, around 1350-1400 B. C. the first
definite evidence of "interstate" trade appears with the importation of steatite bowls from
the north. Incised designs, but not the vessels themselves, typical of Orange 4 (also pre
sent in Orange 3) are reported for the Stallings Island site in Georgia (Claflin 1931: PI. 15).
The opposite--typical Stallings Island punctated sherds found in Florida--are limited to
three sherds from widely separated places (Goggin 1952:75).
By 1000 B. C. the manufacture of fiber-tempered pottery had ceased, but before that
happened three new ceramic series were developed: 1) St. Johns, a chalky ware usually
devoid of intentionally added temper but sometimes containing crushed sherds; 2) Pasco,
heavily tempered with crushed limestone; and 3) a semi-fiber-tempered ware containing
sand and referred to in this paper as Norwood (Phelps 1965). Vessels of the first two
pastes were made with flat-bottoms in the Orange style and decorated with simplified
straight line designs as found in the Orange 4 subperiod. The sand-tempering trait of the
Norwood series was probably introduced from Georgia or west Florida, but the other
pastes represent local Floridian innovations.
It seems likely that shortly before this time, i. e. about 1200 B. C. perhaps earlier,
fiber-tempered pottery was carried into West Florida (if it hadn't been there before) and
along the Gulf Coastal Plain to Louisiana, while Poverty Point clay balls were traded into
Florida where they have been found as far east as Tick Island. One radiocarbon date for
Bryant's Landing Shell Midden in Alabama (Wimberly 1960:211) suggests the presence
there at 1540 B. C. of plain fiber-tempered pottery associated with Poverty Point clay ob
jects. This zone underlaid one containing Bayou La Batre pottery. For some reason,
the traders or migrants to the west carried only undecorated vessels~-at least decorated
sherds have not been found-~although Orange Incised designs are found on the flat rims of
steatite vessels at Poverty Point. Similarly decorated steatite vessel rim sherds are il
lustrated by Rouse (1951: 223, PI. 3, M-N) for South Indian Field in extreme eastern
In the Florida State Museum1 s research collections are 32 fiber-tempered sherds
(cat. no. 103301) from Poverty Point. The amount of fiber tempering, as evidenced by
molds, is variable. In most instances the quantity by peninsular Florida and eastern
Georgia (Stalling Island standards) is not great. One sherd is chalky like very early St.
Johns Plain, several have a "soapy" feel like Tchefuncte Plain, but most have a fine,
sandy feel. Examination with a 5-power hand glass fails to disclose any intentionally added
sand or grit temper. Only rarely is a sizable quartz grain present. Much more frequent
are red lumps, believed to be crushed sherds. The sandy feel is caused by a large amount
of tiny quartz "bits" which must have been natural inclusions in the clay. Sometimes the
paste is "lumpy" like that of Tchefuncte Plain.
Rim sherds exhibit simple rounded lips. Some are from large mouthed vessels with
straight sides that have parallel sides or taper upwards from 10 to 6 mm. in thickness.
One incurving rim varies from 7 to 3 mm. in thickness. Most sherds approach a thick
ness of 1 cm. Two flat basal sherds, both with unmarked bottoms are present; one is
1.5 cm. the other 5 cm. thick. Coiling breaks are not present, but evidence of hand
modeling is to be found on several rim sherds. These sherds, except for the fibrous in
clusions, closely resemble those of the Tchefuncte series. Their inclusion of extremely
fine quartz grains sets them off from the Orange series, but I feel they more closely re
semble fiber-tempered pottery from Florida than that from Stallings Island in Georgia.
Certainly the straight-sided flat bottomed containers resemble Orange Plain or undecora
ted St. Johns Incised vessels. The inclusion of brown lumps--crushed sherds--is another
similarity. The above description, except for the lack of decoration, would best fit the
Florida chronological scheme around 1000 B. C. when ceramics were changing from Orange
to St. Johns paste, but straight-sided flat-bottomed containers were still the vogue.
At or immediately after 1000 B. C. St. Johns Incised and Plain pottery covered Florida
from the Chattahoochee River to Miami and left an impression on Tchefuncte Plain and In
cised vessels. Indeed, radiocarbon dates suggest that early St. Johns ceramicsor know
ledge of themmay have reached Louisiana before the crystalization of Tchefuncte culture
as recorded at the Tchefuncte site. Very likely proto-Tchefuncte sites, other than Poverty
Point, will be found. St. Johns and Tchefuncte pastes are similar. Some early St. Johns
vessels show a few casts of fibrous materials as do some examples of Tchefuncte Incised
(Ford and Quimby 1945:58). Tchefuncte Incised designs (Ford and Quimby 1945: PL 3,
a-j) are simple, like those of Orange 4 times. The beaker form with outward slanting walls
and a flat base resembles the usual St. Johns Incised container except for size. However,
Tchefuncte punctated and pinched areas are reminiscent of Georgia fiber-tempered cer
amics. Probably, influences from both Georgia and Florida fused with others from Mexico
to form the Tchefuncte culture.
Via this mechanism Orange period decoration was passed on to Alexander Incised and
other pottery types. Ford (1966:794) in this connection mentions also the Fourche-Maline
ceramics of eastern Oklahoma (Griffin 1952: Fig. 131, C-D, I-J). The illustrations of
Goose Creek Incised decoration, used by Willey (1966:380-82) in his Introduction to Amer-
ican Archaeology, are extraordinarily close to those found in Florida on Orange Incised
sherds while the paste, in the picture, looks like that of St. Johns Incised. The designs,
ORANGE PERIOD, FLORIDA
including the use of ticks, are closer to Orange Incised decoration than they are to those
of Coles Creek with which they are sometimes compared.
The paste (Suhm and Krieger 1945: 378, PI. 72) of Goose Creek Incised vessels some
times contains crushed sherds. A neglected pottery type of the St. Johns River valley is
Tomoka Plain (Griffin and Smith 1949:349; Goggin 1952:101), a soft St. Johns paste con
taining fragments of sherds. Originally thought to be late in time, we now know that this
paste concentrates in the Florida Transitional and early part of the St. Johns I period
(FSM records). Frequently used in the manufacture of St. Johns Incised vessels and found
along the Gulf coast as well as in the St. Johns River valley, it, together with the decora
tive features, forms a bond across the Gulf Coastal Plain.
I have discussed ceramic events of the Florida Transitional period and their relation
ships to the Gulf Coastal Plain in some detail in a recent publication (Bullen 1969a) and do
not propose to go into much detail here. This is the first period exhibiting major extra
territorial communication. The Zabski site (Atkins and MacMahan 1967) may be consider
ed the type site. This period is also well represented at Sunday Bluff (Bullen 1969a) beside
a small stream flowing into the St. Johns River, at the multicomponent Bluffton and Tick
Island sites, at the Johns Island and Battery Point sites on the Gulf coast of the peninsula
(Bullen and Bullen 1950, 1953), and at the base of many shell middens along the upper part
of that coast. It is definitely a period of geographical expansion and of the movement of
people and of ideas.
The importation of steatite vessels increased. Tammany Pinched decoration and
other Tchefuncte ceramic traits are found at the Zabski site on Merritt Island and at Tick
Island, Bluffton, and other large sites on the St. Johns River. Tchefuncte influences or
connections have also been suggested for La Venta punctated designs found near Vera Cruz,
Mexico (Griffin 1966:122). If correct, this would indicate an interaction area extending
from that city all the way to Cape Canaveral, recently renamed Cape Kennedy!
Limestone-tempered ceramics (Fig. 9, c-d) were developed at the beginning of the
Florida Transitional period along the Gulf coast of Florida north of Tampa Bay. Perico
Linear Punctate of that area may reflect in its technique Tchefuncte influences. Slightly
later simple-stamped vessels, lacking any suggestion of tetrapods, were introduced in
both ends of the Gulf Coastal Plain. In peninsular Florida this simple stamping shortly
took the form of "Cross-Stamped" (Phelps 1966b), with wide, deep lands and grooves
(Fig. 9, a-b), and was unquestionably introduced from Georgia or northwest Florida.
At about this time, at least before 500 B. C. Tchefuncte or Tchula people diffused
rocker-stamped tetrapodal vessels eastward along the coastal plain. They have been
found, both zoned and unzoned, as far south as Tarpon Springs (Florida State Museum
records). Somewhat simultaneously, linear- and check-stamping were developed in Geor
gia. Deptford potters rejected the rocker stamping but accepted the tetrapod in their de
velopment of Deptford period pottery which in a short time covered much of the Gulf
Coastal Plain and north and northeast Florida. This consolidation probably coincided with
the adoption of a stable form of horticulture.
Culture dynamics slowed down for a while and Indians in Florida--as well as their
cousins to the west and north--enjoyed the blessings of agriculture, village life, and bur
ial mound ceremonialism. Apparently, the latter had been around, at least in an incipient
form, for a long time to judge from the groupings of bundle burials and rows of postholes
found at Tick Island in apparently preceramic deposits with four radiocarbon dates before
3000 B. C. (Bullen 1962; Samples M-1264, 1265, 1268, 1270, letters of June 10, 1963 and
March 2, 1964 from James B. Griffin and Roscoe Wilmeth).
As outlined above, it is believed that a significant portion of the cultural developments
of Florida Indians during the Late Preceramic Archaic, Orange, and Transitional periods
were passed on to influence many of the succeeding cultural developments of southeastern
The resultant picture in Florida, after the close of the Florida Transitional period,
differs in certain respects from that of the rest of the Gulf Coastal Plain. Deptford cul
ture with sand-tempered pottery clearly held sway over the northern part of the peninsula
and west Florida, and was followed later by the Weeden Island and Ft. Walton-Safety Har
bor cultures. In the east large burial mounds were built (Bullen, Bullen, and Bryant 1967),
but chalky pottery--first plain or red-painted, later check~stamped~-and the shell midden-
village way of life continued until the coming of the Spanish. This probably reflects lack
of good agricultural land in that area. People were undoubtedly lineal descendants of those
living there during the Orange period.
On the Gulf coast north of Tampa Bay, the manufacture of limestone-tempered vessels
continued until Mississippian times, and Weeden Island period pottery may be sand- or
limestone-tempered or made of temperless St. Johns paste. Around Lake Okeechobee,
the distinctive Belle Glade Plain pottery was developed and made without variation from
before the time of Christ until the disappearance of the aborigines. But this drabness in
ceramics was accompanied by extensive ceremonial centers with mounds, circles, and
roadways. Further to the south in the Glades area, which is ringed by shell middens,
hard sand-tempered pottery was developed with distinctive incised decoration unique in
the southeast. To some extent pottery decoration, tools, and way of life, may be traced
back to the Orange period.
ORANGE PERIOD, FLORIDA
Fig. 6. Artifacts from Orange 1 and 2 levels, Bluffton, Florida.
a, bone pin; b, bone awl; c-d, Orange Incised, concentric diamond design; e, Tick Is
land Incised; f-h, Orange Plain; i, shell celt; j, chert point; k, shell gouge.
Fig. 7. Orange period shell tools, Hill Cottage midden, Palmer site, Osprey, Florida.
a., columella (Fasciolaria) chisel; b-d, columella (Busycon) hammers; e, shell pendant
or net sinker; f, Fasciolaria hammer; g, Busycon pounder.
ORANGE PERIOD, FLORIDA
Fig. 8. Specimens from Orange 3 deposits, Summer Haven, Florida.
a-1, Orange Incised (c-f, ticked designs; g-h, punctated side lugs;_i, punctated rounded
lip; j-k, decorated flat lips); m-n, Busycon (carica) picks.
Fig. 9. Sherds from Sunday Bluff, Florida.
a-b, Deptford Cross-Stamped; c-d, Pasco Plain; e, fragment, steatite
vessel; f-g, St. Johns Incised; h-k, Orange Incised; 1, Orange punctat
ed; m, Orange Incised with flat decorated lip.
ORANGE PERIOD, FLORIDA
Fig. 10. Specimens from Zone 9, Site J-9. Florida.
-b, bone; c-f, h, chert; g, steatite sherd; i, St. Johns linear check stamped(?)
k, St. Johns Incised; 1-n, Orange Plain.
Fig. 12. Orange Incised (a-c) and semi-fiber-tempered incised
(Norwood ?) from Bolen Bluff, south of Gainesville, Florida.
Fig. 11. Chipped tools from Zone 9> Site J-5, Florida.
a-j, hafted scrapers; k-o, medium~sized stemmed projectile points; p-q, hafted
knives; r-s, small ovate knives.
FIBER-TEMPERED POTTERY IN THE TENNESSEE VALLEY
John W. Griffin, Chief
Southeast Archeological Center
National Park Service, Macon, Georgia
In general, fiber-tempered pottery in the Southeast is confined to the coastal plain -~
the area below the Fall Line. There are modest exceptions to this generalization in sever
al areas, and one major exception, the middle portion of the Tennessee Valley. Presence
of fiber-tempered pottery in the Tennessee Valley has been known for quite some time, and
until recently this occurrence was with some justification embodied in the literature as one
of the three centers of fiber-tempering -- the others being the Stallings material of the
Georgia coast and the Orange material from Florida. The data presented at this sympo
sium considerably broaden this older concept, and provide us with the opportunity to re
examine the position of the Tennessee Valley fiber-tempered pottery in relation to the
total Southeastern picture.
Our initial knowledge of this pottery was derived from salvage excavations in the
Tennessee Valley Administration reservoirs in the years prior to World War II, and our
major source of data remains the reports on those excavations. There have been relative
ly few additions to the data, and I, therefore, have little new to report. Also, I should
make it clear at the outset that I am not going to describe this pottery in detail, nor am I
going to refine the types. The original type descriptions were presented by Haag (1939),
and they have been restated with some changes in nomenclature by Sears and Griffin (1950).
The series, made up of four types, is usually called the Wheeler series. The types take
their names from the surface treatment: Plain, Simple-Stamped, Punctated, and Dentate-
Sears and Griffin have noted that the fiber inclusions are not the same as those in the
Stallings and Orange series. Casual observation indicates that the amount of tempering is
less than in the classic Stallings and Orange, approaching somewhat the "semi-fiber-tem
pered" of Florida. Surfaces on sherds of the Wheeler series that I have seen appear to be
smoother and better finished than on the other series, and the pottery is somewhat harder.
I am aware, of course, of the suggestion by Weaver (1963) that fiber-tempering may
not truly be tempering at all, but rather inclusions in the clay from which the pottery was
fashioned. I find the idea appealing, yet am unwilling at this point to either accept or dis
card it. Geographical variation in the nature of the included fibers cannot, per se, give us
much insight into this problem, since plants also might be expected to vary biogeographi-
cally. What is needed here are more intensive technological studies both on the botany of
the fibers and the composition of available clay sources.
We should mention next the geographical distribution of fiber-tempered pottery in the
area under examination. So far as the literature is concerned, the maximum concentration
is in the Pickwick Basin where 6. 4 per cent of the 47,966 sherds tabulated belong to the
Wheeler series (Haag 1942). The next reservoir upstream, impounded by Wilson Dam, was
constructed before the emergence of detailed salvage, and we consequently know little or
nothing concerning the relative quantity of fiber-tempered pottery. The pottery from the
Florida Anthropological Society Publication Number 6.
Wheeler Basin was analyzed by J. B. Griffin, but he was provided only a selected sample,
of which about 3 per cent of the 3,749 sherds were fiber-tempered (Griffin 1939). Still
further upstream in the Guntersville Reservoir, only 22 sherds of over a quarter of a mil
lion were fiber-tempered (Heimlich 1952). Beyond Guntersville, and up into Tennessee, ar
an occasional sherd is found.
Turning the other direction, downstream on the Tennessee River, but once again
northward into the state of Tennessee, we do not have published data on the total ceramic
content in the Kentucky Lake Reservoir. Various statements indicate that fiber-tempering
dies out rather rapidly in this direction, too.
It would seem to be no accident that the concentration of fiber-tempered pottery along
the Tennessee River is in just that portion which is approached most closely by the coastal
plain area below the Fall Line. It is but a short jump from the Tennessee River in the area
of Pickwick Basin and Wilson Dam to the headwaters of Mobile Bay. I will be looking for
ward to intensified research in the Tombigbee drainage, a circumstance which will be
forced upon us if and when the proposed Tombigbee waterway project reaches the construc
Of course, it may be that influences came this way and left little evidence, if, as
Ford (1966) suggests, the cultures producing fiber-tempered pottery are maritime in orien
tation and capable of rather dramatic jumps leading to discontinuities. It might be argued
that if the cultures were truly heavily adapted to the use of shellfish there might be great
discontinuity between the coast and the mussel-rich areas of the portion of the Tennessee
River. And, it is still an open question as to whether the presence of fiber-tempered pot
tery in the Tennessee Valley represents the spread of people or ideas, or both.
The temporal placement and range of this pottery remains to be discussed. It has
long been maintained that fiber-tempering lay at the base of the Tennessee Valley ceramic
sequence, and most workers seem to have assumed that it was absolutely earlier than the
sand-tempered pottery which was believed to succeed it. This conclusion was based on
percentage distribution of the various wares in stratigraphic cuts, particularly on some of
the larger shell heaps. However, there is not a single instance in which a level containing
only fiber-tempered pottery can be isolated. Workers in the area have begun to ask whether
the fiber-tempered and the sand-tempered, which seem to appear together in many site con
texts, may not be, at least in part, coeval (personal communications, D. L. DeJarnette and
Richard Marshall). This is another one of those questions which is going to have to be
solved in the field.
No radiocarbon determinations are available for the fiber-tempered pottery of the
Tennessee Valley, in marked contrast to the situation in Georgia and Florida. On the basis
of similarities it has been held that all of the fiber-tempering is roughly contemporaneous.
However, we should remind ourselves of the long time span assigned to the Stallings and
Orange material. There is no reason why the Tennessee Valley material should have to be
contemporaneous with the entire range elsewhere, and, in fact, the very quantity (or lack
of it) in the Tennessee Valley might be taken to argue for a shorter total time range. Let
us add to this Ford's (1966:783) impression that the decorations on the Tennessee Valley
fiber-tempered pottery belong to the 500 B. C. time level. This would be late in the total
Southeastern range of the ware.
Long Branch Fabric-Marked is dated at 150 200 B. C. at Russell Cave (J.W. Grif
fin, Ms.) and 340 150 B.C. at the Westmoreland-Barber site in Tennessee (Faulkner and
Graham 1966). In both of these instances both fiber-tempering and sand-tempering are vir
tually absent from the sites, and the Long Branch seems to come in directly atop a non
ceramic Late Archaic. But further downstream, fiber-tempered and perhaps sand-tem
pered pottery could have been present at 500 B. C. or slightly earlier. Taking the admitted
ly slender evidence of dates following fiber-tempering, the relative scarcity of the types
in the Tennessee Valley, and Ford's impressions of time based on design, I am inclined
to see fiber-tempering in the Tennessee Valley as a very late aspect of the total South
eastern fiber-tempered picture.
I have said nothing about the artifact complex associated with fiber-tempered pottery
along the Tennessee for the very simple reason that I do not know of a single instance that
could be regarded as unquestionable association. This is, of course, in vivid contrast to
the Stallings and Orange situations.
In summary, I view the presence of the Wheeler series along the Tennessee River as
a late fiber-tempered manifestation which probably reached that area through the Tombig-
bee drainage (although there are other alternatives), and which proceeded to exploit the
riverine environment, with heavy emphasis upon the shellfish resources, in much the same
way that other peoples producing fiber-tempered pottery exploited the resources of the sea,
estuaries, and other rivers of the Southeastern Coastal Plain.
Whether its rather restricted distribution in the Tennessee Valley was due to con
trasts in available resources or the resistance of other Late Archaic peoples is not clear.
However, the time period involved was not long, and any further spread seems to have been
truncated by the appearance of the Early Woodland tradition marked by limestone-tempered
pottery, which, at least in Russell Cave, marks a sharp break with the Late Archaic.
THE LATE ARCHAIC IN THE SAVANNAH RIVER REGION
James B. Stoltman, Associate Professor
Department of Anthropology, University of Wisconsin Madison
The concern of this paper is the culture history and culture dynamics of a delimited
region centering on the Savannah River during a delimited period of time, 3000 B. C. to
1000 B. C. The selection of the particular region and period for discussion is dictated by
the theme of this volume. The lower Savannah River has long been recognized as a major
center of fiber-tempered pottery manufacture. Contrasted with the ceramics of other
Southeastern regions, the fiber-tempered pottery of the Georgia-Carolina coast has enough
distinctive features to merit classification into a separate ceramic series, the Stallings
series (Sears and Griffin 1950). The currently known distribution of the Stallings ceramic
series in space and time provides the point of departure for the ensuing discussion.
The temporal limits of discussion are partly based on nine currently available radio
carbon dates (from five sites) associated with fiber-tempered pottery of the Stallings series
(Table 1). These dates indicate that the Stallings ceramic series had its inception in the
middle of the third millennium B. C. and persisted into the early second millennium B. C.
A broader time range, beginning about 3000 B. C. and ending about 1000 B. C. was se
lected as the temporal framework for discussion in order to bracket the earliest and latest
Stallings pottery dates, thus permitting discussion of the culture-historical problems of
origin and demise of the way of life that fostered this ceramic tradition.
The Savannah River Region: Definition and Ecology
Geographically, discussion shall be limited in this paper to what is henceforth refer
red to as the Savannah River Region. The boundaries of this region as illustrated in Figure
13 are defined on the basis of the distribution of currently known sites of intensive habitation
by makers of the Stallings ceramic series. Fiber-tempered pottery of the Stallings series
has been reported beyond these limits, but usually as surface finds and always in small
frequencies (Holder 1938, Griffin 1943:164, Caldwell 1950:19, South 1964, Wauchope 1966:
231, Waring in Williams 1968:185, 255).
The ecology of the Savannah River Region is heterogeneous. Within the region we may
recognize five ecological zones that are relevant to understrnding its human exploitation:
1. Piedmont upland outside the Savannah Valley: topography rolling to rugged;
mainly red clay soils underlain by Precambrian igneous and metamorphic rocks;
rocks and minerals of value to prehistoric man include quartz, slate, phyllite,
gneiss, gabbro, and steatite; dominant vegetation is a mixed oak-pine forest
2. Interior Coastal Plain outside the Savannah Valley: gently rolling to nearly flat
terrain nearer the coast; underlain by unconsolidated marine limestones, sands,
Florida Anthropological Society Publication Number 6.
SAVANNAH RIVER REGION
Fig. 13. The Savannah River Region.
and clays of Cretaceous to Recent age; mainly sandy soils; minerals and rocks
of value to prehistoric man rare, consisting mainly of localized occurrences of
chert -where the Barnwell or Flint Creek Formations outcrop (Cooke 1943:62,
98, Plates 1 and 2); vegetation is predominantly "Southeastern Evergreen For
est" (Braun 1950).
3. Atlantic Littoral: differs from the rest of the Coastal Plain mainly because of
the presence of extensive brackish swamps and salt marshes protected from the
open sea large islands; oysters, clams, and mussels abound in these protected
areas; rocks and minerals of value to man scarce.
Fig. 14. Major fiber-tempered pottery sites in the Savannah River Region.
4. Savannah Valley, Coastal Plain segment: broad, swampy floodplain subject to
seasonal inundation; heavily forested with water-tolerant trees; floodplain fea
tures of significance to man include levees, sand bars, and oxbow lakes, often
at considerable distances from the modern channel; the river here is meander
ing, sluggish and rich in fish and shellfish; the silty to sandy alluvium is devoid
SAVANNAH RIVER REGION
5. Savannah Valley, piedmont segment: here the river is swifter, more deeply en
trenched with a narrower, rockier flood plain than on the Coastal Plain; swifter
currents mean limited mussel resources; shoals near the fall line are excellent
musseling stations with elevated, dry upland nearby for safe, convenient habita
tion sites; differs from Coastal Plain segment in ease of access to rich mineral
resources of the piedmont upland.
At least 24 sites within the Savannah River region have produced evidence of intensive
occupation by makers of fiber-tempered pottery (Table 1 and Fig. 14). Of these, 12 have
been excavated more-or-less scientifically: Stalling's Island, Lake Spring, White's Mound,
Rabbit Mount, Chester Field, Jones Island, Refuge, Bilbo, Oemler, Meldrim, Dulany, and
Sapelo Island. While some information has been published on all of these excavations, rea
sonably complete site reports that describe both the excavations and the artifacts are cur
rently available only for Stalling's Island, Lake Spring, Bilbo, Refuge, and Sapelo Island.
A brief note recording artifact frequencies from the Dulany site has been published by
Williams (1968:208) based on Waring's rough field notes.
SAVANNAH RIVER REGION SITES PRODUCING FIBER-TEMPERED POTTERY
Sites Radiocarbon Dates References
1. Little River
2. Lake Spring
3. Great Kiokee Creek
4. Stalling's Island
Daws Island (38
Miller 1949: 39
Jones 1873: 334-39; Miller 1949: 46-47
Claflin 1931; Fairbanks 1942; in Williams
1968: 331; Bullen & Greene 1970
Claflin 1931: 40-42
Phelps and Burgess 1964
Stoltman 1966 and In Press
in Williams 1968: 198-208
in Williams 1968: 152-197,330
in Williams 1968: 208; 329
Flannery 1943; Griffin 1943
Flannery 1943: 152
in Williams 1968: 182-83; Caldwell 1952:314
in Williams 1968: 182-83; Caldwell 1952:314
in Williams 1968: 182-83; Caldwell 1952:314
in Williams 1968: 263-278; 329
The Stallings Ceramic Series
Fiber-tempered pottery of the Savannah River Region has been well-described in the
literature (Griffin 1943; Sears and Griffin 1950), so we need only review some of its more
significant features here. Needless to say, the presence of a vegetal rather than mineral
temper is one of its unique features. Another diagnostic characteristic is the mode of ves
sel construction, which involves modeling from a single mass of clay rather than coiling.
Although no complete vessels have yet been reported in the Savannah River Region, the
wide-mouthed bowl form appears to be universal. Bases, perhaps because of the modeling
technique, are never conoidal, but are typically rounded or flattened. From these base
forms, the vessel walls rise in a gentle arc so that near the lip the rim orientation is slight
ly out-curving or sometimes approaches the vertical. A form peculiar to the Stalling's Is
land site is cazuela-like bowl with a prominent, angular shoulder and a sharply in-curving
Most fiber-tempered sherds found in the Savannah River Region are undecorated, but
punctated, incised, and simple stamped decoration occurs, apparently on the same paste
and vessel forms as the plain type. Punctates were made with various instruments, in
cluding fingernails and hollow reeds, that were either impressed into the clay vertically
(Fig. 16,a.-b) or, more commonly, dragged and jabbed around the vessel producing parallel
rows of "linear punctations" (Fig. I6,d-g). Incised motifs are most commonly concentric
horizontal rings or cross-hatched lines (Fig. 15, c-d). Rarely incising and punctation occur
together. Simple stamping, which may not be decoration at all, seems to have been accom
plished by impressing a long edge of a blunt instrument repeatedly into the clay (Fig. 15, h).
The irregularity and crudity of the impressions argues against the use of a carved paddle
for simple stamping by Stallings people.
Based upon the absence of decoration or upon the presence of one of the above three
modes of decoration, four fiber-tempered pottery types may be recognized in the Savannah
River Region: Stallings Plain, Stallings Punctate, Stallings Incised, and Stallings Simple
Stamped. The first two have been formally defined by Griffin (1943) and Sears and Griffin
(1950); the latter two, although never formally defined, have long been recognized as in
tegral constituents of some Stallings assemblages.
Since I am using these type names on a region-wide basis, a practice in conflict with
some earlier terminologies, a few words of explanation are in order. In particular, the
term St. Simons has often been applied to all or part of what I am here calling Stallings.
Use of the St. Simons prefix was an outgrowth of Holder's excavations on that island in
1936-37. In 1941 Caldwell and McCann assigned the fiber-tempered sherds found at the
Irene site (and at Stalling's Island as well) to two-types: St. Simons Plain and St. Simons
Incised and Punctated. This represents the first application of formal type names to fiber-
tempered pottery within the Savannah River Region. It is important to note, however, that
while these terms were introduced, the types were never defined.
In 1943 J. B. Griffin analyzed a sample of fiber-tempered sherds excavated by W. K.
Moorehead from the Chester Field site and, upon comparing them with Stalling's Island
sherds, formally defined the types Stallings Plain and Stallings Punctate. He selected the
Stallings prefix over St. Simons because ". .the Stalling's Island site is the only one now
fully described in the literature. ." (Griffin 1943:159). Subsequently, Sears and Griffin
(1950) continued this usage while formally subsuming the St. Simons types in their Stallings
SAVANNAH RIVER REGION
EXCAVATION SIZES AND NUMBER OF ARTIFACTS,
SAVANNAH RIVER REGION SHELL MIDDENS
NUMBER OF ARTIFACTS
(sq. ft. )
(cu. ft. )
per cu. ft.
ca. 4. 0
ca. 6. 0
ca. 4. 5
ca. 5, 962
ca. 4. 0
ca. 7. 0
ca. 5. 0
A-Claflin 1931. B-Bullen and Greene 1970.
Although the Stallings designation gained wide acceptance (Miller 1949), A. J. Waring,
whose research focused on the Savannah River mouth, retained the St. Simons designation
in describing his material. Unlike Caldwell and McCann, however, he restricted the St.
Simons designation to the coastal ware. He felt the fiber-tempered pottery from Stalling's
Island differed from that of Bilbo and other coastal sites in its ". (1) thinner and more
uniform vessel walls, (2) smaller, neater, and more varied forms of punctation, (3) the
presence of a crude form of 1 simple stampling' on the bases of many of the vessels, and
(4) the presence of the flanged, carinated bowl. (Williams 1968:160). In the absence
of supportive quantitative data (e. g. ranges of variation, etc.), the first two differences
are ineffective criteria for sorting St. Simons types from analogous Stallings types. The
latter two, because they are distinctly minority traits at Stalling's Island, appear to me to
be relatively inconsequential differences, particularly in light of the many over-riding sim
ilarities of a more fundamental nature. Indeed, considering Waring's own observation after
comparing fiber-tempered sherds from Bilbo with over 900 Stalling's Island sherds "This
relationship approaches identity." (Waring 1 939:7)--I favor assigning all indigenous fiber-
Tempered pottery of the Savannah River Region to the Stallings ceramic series.
Williams (1968:105) has recently suggested applying type-variety terminology to these
ceramics. He suggests Stallings be retained as the type name within which local varieties--
for example, Bilbo, St. Simons and Stallings-- be recognized. Conceptually, this sugges
tion is compatible with the points of view underlying the various nomenclature differences
Before discussing the non-ceramic materials found in association with fiber-temper
ed pottery, a few words are in order concerning the relative popularity of the various Stall
ings ceramic types. This problem has three aspects, spatial, temporal, and functional
and each must be considered.
Ceramic frequencies are available for eight excavated Stallings sites in the Savannah
River Region, seven of which are included in Table 3. The Chester Field data are omitted
because the 158 fiber-tempered sherds on which Griffin (1943) based his analysis constitute
a highly selected sample of a much larger collection. From Table 3, it is clear that Stall
ings Plain predominates at all sites, except, interestingly, the type site. Stallings Punctate
is the leading decorated type at each site, with Stallings Incised minority type at all sites
except possibly Refuge and Stalling's. Simple Stamped is reported from only two sites.
STALLINGS CERAMIC TYPES AT SEVEN SITES
873 (59. 9)
82 (5. 6)
988 (67. 8)c
104 (4. 1)
728 (28. 7)
106 ( 3.3)
149 ( 4. 6)
1,138 (98. 6)
12 ( 1.0)
4 (0. 4)
1 6 ( 1.4)
127 (72. 2)
2 ( 2. 2)
12 ( 2. 8)
13 ( 3.0)
a Stalling1 s Island data from Bullen and Greene 1970.
b Percentages of total fiber-tempered sherds per site,
c Includes 17 miscellaneous decorated.
While much of the site-to-site variability in frequencies of ceramic types would ap
pear to be due to regional variation, the impact of time differences must also be assessed.
A basic temporal problem concerns the existence of a period during which Stallings Plain
was the only pottery made in the Savannah River Region. The only site where there is un
ambiguous archaeological evidence of a component with Stallings Plain as the sole pottery
type is Sapelo Island. At this site Waring and Larson excavated 17 six-inch arbitrary
levels in the deepest portions of the shell-ring wall. They recovered decorated fiber-
tempered sherds from only the upper five levels, while the lower 12 levels had Stallings
Plain as the sole pottery type (Williams 1968:274, 278).
The evidence from Bilbo and Rabbit Mount is suggestive though equivocal on this
point. The sites are similar in that the basal levels of some excavation units in the shell
SAVANNAH RIVER REGION
middens produced only Stallings Plain pottery, while other excavation units in the same
middens produced decorated fiber-tempered sherds as well (Stoltman 1966; Williams 1968:
179-185). Depending upon one's biases, these data could be used in support of the contra
dictory views that the basal midden levels at these sites were deposited by makers of Stal
lings Plain pottery alone (in this case the decorated sherds would be interpreted as later
intrusions from above) or that the complete middens were deposited by makers of both plain
and decorated fiber-tempered pottery (in which case the absence of decorated sherds in the
basal levels of some pits would be attributed to sampling error). Whatever the correct in
terpretation, a clear trend of increasing relative importance of decoration from the midden
base upward is evident at both sites, as indicated in Table 4 by ratios of plain to decorated
Stallings pottery. At Stalling1 s Island the same trend is also evident (Table 4) despite the
presence of decorated sherds in, apparently, all ceramic levels of Greene's six test pits.
Indeed, decorated sherds even predominate in the lowest ceramic level of one pit (Bullen
and Greene 1970:20).
Although Stalling's Island, Bilbo and Rabbit Mount manifest a similar trend of in
creasing use of decoration within the Stallings series through time, the incidence of dec
orated sherds at Stalling's Island and at Bilbo is significantly greater, whether measured
absolutely or relatively (Tables 3 and 4). Some of the observed between-site variability
may be attributed to spatial separation, but most of it appears to be the result of differ
ences in age. Thus, Rabbit Mount, with its lesser proportion of decorated to plain sherds,
might be expected to be the oldest site and the radiocarbon evidence is in line with that ex
pectation (Table 1). Somewhat confusing, however, is the Sapelo Island date -- 1750250
B. C. (M-39) -- which derives from an unspecified level, but beneath all decorated pottery
(Williams 1968:329). Considering that elsewhere in the Savannah River Region there are at
least seven radiocarbon dates associated with fiber-tempered pottery that predate the Sa
pelo Island date, either the date is in error (it was run on shell) or else the site, perhaps
because of its geographic marginality, had a conservative ceramic tradition. Despite these
admittedly fragmentary data, I nonetheless suspect that a plain-pottery phase or sub-phase
of the Stallings culture does exist in the Savannah River Region. Current knowledge of such
a phase is restricted to the Coastal Plain and, in fact, to the Atlantic Littoral if the Rabbit
Mount shell midden is assigned entirely to a decorated pottery phase. I am presently un
willing to do this, however. Considering the antiquity of the Rabbit Mount dates and the
fact that only 6 sherds of a total of 300 from the lower half of the midden were decorated
(Table 4), I feel it most likely that the dates apply to a plain-pottery-only Stallings phase.
The variable of function must also be considered in attempting to understand inter-
and intra-site differences in ceramic assemblages. Exactly what the function of Stallings
pottery might have been is unknown for certain. Pottery generally is well-suited for both
food storage and cooking, and presumably it was so employed by Stallings people. Charred
encrustations occur on inner surfaces of some sherds, which suggests they functioned as
cooking pots. There is a positive correlation between Stallings ceramics and shell mid
dens indicating considerable dependence upon shellfish. Palatability of mollusks depends
upon their being kept alive until they are to be eaten, which thus increases the likelihood
that techniques for live storage would be devised. The shellfish could, of course, have
been gathered just before they were to be eaten, thus eliminating the need for storage, but
considerable inconvenience would be avoided by keeping a supply near at hand in, say, a
pot full of water. Since my young daughter recently kept two "pet" river mussels alive in
a fish bowl for over a month, it seems reasonable that Stallings people might have similarly
used their pottery.
RATIO OF PLAIN TO DECORATED SHERDS AT BILBO, RABBIT MOUNT,
AND STALLING' S ISLAND
Bilbo a Rabbit Mount b Stalling's Island c
Upper Half 546
Lower Half 358
a Stratigraphic block. b Portioned excavated c Greene's Pits 1,2,4, 6.
While the use of Stallings vessels for the cooking or storage of shellfish offers no ob
vious explanation of the site-to-site ceramic variability recorded in Table 3 nor the within-
site variability recorded in Table 4, these functions may be the explanation for the paucity
of Stallings ceramics in the Piedmont and Coastal Plain outside the environs of the Savan
nah Valley and the Atlantic Littoral. These ecological zones are characterized by small
streams of limited mussel potential and no reported archaeological sites -with either large
shell middens or large yields of Stallings sherds. Taking these negative and positive cor
relations at face value, it appears that Stallings peoples left their primary riverine or lit
toral environments for short-term forays into the uplands to hunt (most likely deer) and to
gather plant foods along with various raw materials such as flint. While engaged in such
activities, their ceramics would have been of little or no use and, indeed, would even have
been an encumbrance to their mobility. I offer this as possible explanation as to why the
Stallings series is so poorly represented in ecologic zones 1 and 2 as defined above.
Stone Industry Associated With Stallings Ceramics
While a rich lithic industry is normally associated with Stallings pottery, reasonably
comprehensive descriptions of the stone artifacts are available for only four sites: Stall
ing' s Island, Lake Spring, Bilbo, and Rabbit Mount. Fortunately, these sites span the Sa
vannah River Region's geographic diversity, which provides the opportunity to investigate
the role of ecology as a contributing factor to intra-cultural variability. In the ensuing
discussion the presence or absence of various artifact types at sites within the Savannah
River Region are noted without bibliographic documentation; to do so would be needlessly
repetitious, for all such information comes from the few basic sources listed by site in
The only artifacts of chipped stone found at all four of the above sites are stemmed
projectile points, drills (including the distinctive cruciform type reported so far only from
Piedmont sites), leaf-shaped blanks, and probably unifacially retouched flakes. Stemmed
scrapers, though distinctive, are reported only from Stalling's Island and Bilbo.
Especially diagnostic are Savannah River Stemmed projectile points (Fig. 17, a-e;
Coe 1964:44-5). A minority type at the Stalling's Island site has a contracting stem
SAVANNAH RIVER REGION
reminiscent of Coe's Morrow Mountain I and II types (Coe 1964:37-39). Elsewhere in the
Savannah River Region such points supposedly characterize preceramic components (Cald
well 1954; Neill 1966), but Bullen and Greene's recent study (1970:13-14) reveals them to
be relatively more abundant in the ceramic levels at Stalling's Island.
Flake tools have not been described at any of the sites except Rabbit Mount, although
their presence at all sites is virtually certain (Claflin 1931:34; Miller 1949:44). At Rabbit
Mount 82 of the 320 stone artifacts attributed to the Stallings component are retouched uni
facial flakes (Fig. 17, m-o). Sixteen have definite graver spurs (Fig.l7,_p); such gravers
are also reported from Lake Spring.
At Rabbit Mount we found two flaked stone artifact types that have not been reported
before in a Stallings context: narrow, bifacial knives and chert adzes (Fig. 17, _f-l). Both,
like the retouched flakes, are unspectacular, but they are nevertheless significant. The
eight narrow knives (4 to 44 mm. wide with a mean of 27. 6 mm. ) have distinctively par
allel edges, and all have been transversely broken. From these data, I infer their past
function to have been the cutting and prying open of mussel shells. Three flaked chert ar
tifacts I have inferred to be adzes (i. e. wood-working tools) because of their transverse,
gouge-like, cutting edges and lateral edge grinding (from having been hafted? ).
Selection of rock for flaking normally reflects available local resources. Thus at
Lake Spring on the Piedmont 97.7 per cent of the flaked stone artifacts and debitage are of
"aplite" (rhyolite?) and quartz (Miller 1949:40), whereas at Rabbit Mount on the Coastal
Plain over 99 per cent are of chert. By contrast, "slate" appears to predominate at Bilbo
suggesting close contact (trade?) with the Piedmont (Williams 1968:173-4). Material used
in the manufacture of flaked stone tools is significant also because it can profoundly influ
ence artifact sizes and forms. For example, igneous and metamorphic rocks of the Pied
mont tend to outcrop in more massive chunks than either quartz in the same area or cherts,
which occur as fragmented modules on the Coastal Plain. As a result, we could expect,
say, Savannah River Stemmed points made of igneous rocks to be longer and proportion
ately narrower on the average than comparable points of quartz or chert.
A comparison of Savannah River Stemmed points from the North Carolina Piedmont
(mean length, 100 mm.), where the main material was igneous rock (Coe 1964:44), with the
Rabbit Mount points of the same type (mean length, 63.2 mm.), that were mainly made
from chert, supports this expectation. Recently, Bullen and Greene (1970:13-14) have
noted similar size differences related to material within their sample of 177 projectile
points from Stalling's Island.
The only ground or pecked stone artifacts reported from all four of the above sites
are both grooved and perforated "netsinkers" (Fig. 18, e-g). The latter, typically made
of steatite, are among the most diagnostic Stallings artifacts. Their distribution within the
Savannah River region clearly reflects proximity to the natural outcrops, for they become
progressively less common as one moves downstream from Stalling's Island on the Pied
mont (over 2500 steatite netsinkers) to Rabbit Mount (86 fragments) to Bilbo near the Sa
vannah mouth (only 3 steatite netsinkers). It has been suggested that these so-called net
sinkers actually functioned as baking stones much the same way that "Poverty Point ob
jects" of clay supposedly did. Some support for this view may be seen in the complimen
tary distribution of the two, the baked clay balls so far known only from the coast (Dulaney
and Sapelo Island) where the steatite slabs are rare. It should also be noted that, despite
the widespread, if uneven, use of steatite by Stallings people for fashioning various arti
facts, bowls of this material have so far been reported only from Bilbo and Meldrim
Carefully polished bannerstones, grooved axes, and gorgets occur at some (though
not all) Stallings sites, mainly on the Piedmont. Since such artifact types are most easily
fashioned from igneous and metamorphic rocks, it comes as no surprise that they are rare
on the Coastal Plain only two bannerstones (Rabbit Mount and Sapelo Island) one axe
(Rabbit Mount), and no gorgets have been reported from Stallings sites on the Coastal
More roughly made pecked or polished stone artifacts reported commonly from Stal-
lings sites include abraders, mortars, discs and pitted stones (Fig. 18). The latter
three categories, along with the generally amorphous, rarely reported category of ham-
merstones (Fig. 18, a b, ^d), could be interpreted as primarily plant food processing arti
facts. There is, however, currently no evidence other than typology to sustain this inter
pretation, and it must be pointed out that such artifacts, with the possible exception of pit
ted stones at Stallings Island, are uncommon at all sites.
From the foregoing discussion of stone artifact types found in association with Stal
lings ceramics it is apparent that considerable site-to-site diversity exists. Much of this
can be attributed directly to spatial differences in natural resources. Thus, the greater
incidence of artifacts made of steatite and of igneous or metamorphic rocks in Piedmont
sites of the Stallings Culture is here viewed as primarily a reflection of local ecology.
This point merits emphasis for such between-site differences have often been used instead
to support claims of differential age. Age differences no doubt are the source of some of
this variability, but a careful consideration of local ecology has led to the suggestion that
some of this variability--for example, few steatite "netsinkers, shorter (chert) projec
tile points and ungrooved (chert) axes or adzes on the Coastal Plain--is best attributed to
local adaptations to the available resources rather than evidence of age differences.
Bone-Antler-Shell Industries Associated With Stallings Ceramics
A rich bone-antler industry is normally associated with Stallings ceramics, thanks
to the favorably alkaline environment afforded by the shell middens found at most Stallings
period sites. Shell artifacts are rare, although reported from five separate sites. At the
type site all shell artifacts were found with burials. Most of these were demonstrably in-'
trusive, leaving a residue of three ear plugs and seven strings of marine shell beads as
possible Stallings culture grave goods. Two perforated conch shells from Bilbo, one bead
from Sapelo Island, a conch gouge from Dulany, and a perforated conch with a beveled colu
mella from Rabbit Mount complete the inventory of known shell artifacts. The Rabbit Mount
conch is notable, for its three distinctive traits (perforation, strategic breakage, and bevel
ed columella are also found at Puerto Hormiga and on shell "picks" in Florida (Reichel-
Dolmatoff and Bullen, this volume).
The most distinctive bone artifact from Stallings sites is the polished pin (Fig. 19,
a.~e). Their slenderness and carefully polished surfaces distinguish them from awls. They
may be either plain or decorated with finely engraved lines. Nail-like heads may be present
on both plain or decorated types.
Awls carved and polished from deer bone splinters are the most numerous bone arti
facts found at Stallings sites (Fig. 19, f-h). A single needle, the only one reported from a
Stallings site, was found at Stalling1 s Island. Distinctive, though rare, bone artifact types
occurring at more than one Stallings site include cut, sometimes-engraved deer mandibles
with heavily worn inferior surfaces; blunt-ended, beveled, deer bone "spatulae"; beveled,
deer bone (projectile?) points (Fig. 19, i); atlatl hooks (Fig. 19, k); and fishhooks.
SAVANNAH RIVER REGION
Items of personal adornment, reported so far only from Rabbit Mount, are a split and per
forated beaver incisor, a cut and polished piece of turtle plastron, and four polished bone
tubes that presumably are beads.
A variety of antler artifacts reoccur in Stallings sites, the most distinctive of which
is the socketed projectile point (Fig. 19, j). Interestingly, while "only a few arrow points
made from antler tips" were found at Stalling1 s Island (Claflin 1931:26), thirteen are re
ported from Bilbo. Less distinctive antler artifacts reported or expectable at most Stal
lings sites are tines that have been transversely cut (drifts), sharpened (punches), battered
(flakers), or beveled (chisels) and antler bases or segments that have been hollowed at one
end (hafts-Fig. 19, 1). Atlatl hooks of antler are reported only from Stalling's Island and
Burials and Features Associated With Stallings Ceramics
Intentional human internments have been found at the Stalling's Island, Lake Spring,
and Daws Island sites. In addition scattered scraps of human bone were recovered from
Stallings contexts at Bilbo and Rabbit Mount. At Lake Spring and Daws Island single, pri
mary, flexed burials were found unaccompanied by grave goods. The Lake Spring indivi
dual is slightly brachycranic (cephalic index 80.9), although the presence of some artificial
occipital flattening must certainly be considered a contributing factor (Miller 1949:49)
At least eighty-four burials were found at Stalling's Island, mainly in the central,
high portions of the site. While four of these are obviously post-Stallings intrusives and
the ages of others are indeterminate, a majority were attributed to the Stallings occupation.
Most of the internments were primary burials in semi-flexed or flexed positions within
shallow, round sub-midden or within midden pits. Six secondary, bundle burials were also
found. Of eighty burials that could be of Stallings age, twenty-six were accompanied by
mortuary offerings although, exactly what was interred is rarely mentioned. In four cases
"a small animal with canine teeth" (Claflin 1931:44) accompanied the dead. The only shell
artifacts found at the site occurred as grave goods. At least four burials with shell beads
are datable to the Stallings occupation with reasonable certainty.
Only two complete human skulls were recovered but observations on numerous frag
mentary ones were made. Claflin (1931:43) states that "in general. .. the Stalling's Island
people were brachycephalic and did not as a rule practice cranial deformation". Assuming
Claflin's otherwise undocumented observations are correct, the tendency toward brachy-
cephalism is interesting and nearly unique in the Southeast at this general time level.
Lewis and Kneberg (1959:169) have called attention to this, noting that only in the Kays
phase of west Tennessee do similarly brachycephalic skulls occur in the late Archaic.
Clearly, imperfect data on cephalic indices is a tenuous basis for inference about popula
tions and their histories, but it is nevertheless interesting to note that, as discussed below,
this gross anatomical similarity is accompanied by cultural similarities as well.
Pits (either storage or refuse) and hearths are frequent features at most Stallings
sites. Their presence scattered throughout middens indicates that occupation directly ac
companied the deposition of refuse. Other than this, little further of significance about
these features can be gleaned from the reports. The number of such features at Stalling's
Island is noteworthy -- 110 pits and 24 fireplaces. Some of the latter were plastered with
clay that was later fired hard. At Bilbo three pits lined with unfired clay found below water
table were classed as storage pits. The base and sides of a number of storage or refuse
pits at Stalling's Island showed evidence of fire, suggesting a prior use (clam bakes?) be
Other than scattered post molds at Bilbo, the only structure reported from a Stal
lings site is a possible lean-to from Rabbit Mount. It consists of an irregular, packed
clay floor covering an area of no more than ten square feet accompanied by an adjacent
row of three probable post molds. The small size indicates we may be dealing with only
the vestiges of a larger structure, although intensive excavation failed to uncover further
evidence of it.
Subsistence Base and Settlement Pattern
All available evidence indicates that the makers of Stallings ceramics were heavily
dependent upon shellfish -- mainly mussels inland, mainly oysters on the coast -- for
their subsistence. Shell middens are a characteristic feature at every site listed in Table
1 except White's Mound. At the latter site, while a true shell midden was lacking, it is
significant that "the only evidence of concentrated shell deposits so far" was found in the
three squares "where Stalling's Island occupation seems to be heaviest" (Phelps and Bur
At inland sites the available data suggest that intensive shellfish exploitation was a
practice unique to the Stalling's Island culture within this part of the region. Both Claflin
(1931:7-8) at Stalling's Island and Miller (1949:39-40) at Lake Spring found that all non-
Stallings ceramics were found on top of the major Stallings middens or in intrusive pits.
The White's Mound reference cited above also implies that the post-Stallings visitors to
that site did not intensively gather shellfish.
On Groton Plantation, where I conducted a survey, the evidence indicates that the
Stallings people had an economy heavily geared toward the exploitation of shellfish and that
they were the only occupants of the locality to ever do so. Rabbit Mount, a natural sand
knoll within the Savannah floodplain, was the only site of 21 catalogued on the plantation
that had a shell midden. Eighty-seven percent of the sherds from these are of the Stal
lings ceramic series. The remaining 207 sherds, all sand-tempered, derive from I jq
components at the site subsequent to the Stallings occupation (i.e. Thom's Creek
through Savannah II phases) and are interpreted as intrusive.
The distribution of pottery types among the 21 Groton Plantation sites further demon
strates the close correlation between the Stallings ceramic series and shellfish gathering.
From all sites, we amassed a combined ceramic sample of 11,274 sherds. Nineteen of
the sites, all located in the upland adjacent to the Savannah Valley, produced 3,001 of these
sherds, while the two sites in the Savannah floodplain produced the remaining 8,273 sherds.
Of the 3,001 sherds from the upland sites, only 37 (1.2%) are fiber-tempered. By con
trast, 3,307 of the 8,273 sherds (40.0%) from the two floodplain sites are fiber-tempered.
Sherds in our Groton Plantation sample came from the two floodplain sites. These data in
dicate a decided preference by the makers of Stallings ceramics for the Savannah River
Closer scrutiny of the data reveals that it was not simply the floodplain, but specific
floodplain conditions to which the Stallings people were adapted. The two floodplain sites,
Rabbit Mount and Clear Mount, are similar in many ways. Both are natural sand mounds
whose summits rise high and dry above the surrounding lowland, even during the worst of
the annual floods. They are less than one-half mile apart and may be reached on foot dur
ing most of the year either from the river or from the upland. Where they differ signifi-
SAVANNAH RIVER REGION
cantly is in the presence of an oxbow lake adjacent to Rabbit Mount. It was when this lake
was part of the Savannah River system that the Stallings people presumably occupied the
site, for an analysis of a sample of mollusks from Rabbit Mount indicates they were gather
ed from "somewhat slow moving water with a substrate of sand, or sandy mud. (William
J. Clench, personal communication). There can be little doubt that before being cut off
from the Savannah system, the oxbow lake afforded just those conditions and, consequently,
must have been the source of the shellfish found on Rabbit Mount. The point of this is that,
all other factors approximately equal, Rabbit Mount differed from Clear Mount mainly in
its accessibility to shellfish-rich waters. As the following data dramatize, it was the for
mer site that was preferred by the Stallings people:
Total Sherds Fiber-tempered Sherds
Rabbit Mount 6,508 3,257 (50. 0%)
Clear Mount 1,765 50 ( 2.8%)
In our six test pits scattered on Clear Mount plus our survey of the site no sign of a shell
midden nor any evidence of an intensive Stallings occupation was discovered. All these
various threads of evidence support the same conclusion: on Groton Plantation the makers
of Stallings ceramics had a subsistence base heavily dependent upon shellfish. Consider
ing the close correlation between the Stallings ceramic series and shell middens through
out the Savannah River Region, the same conclusion is warranted for the entire region.
Before leaving the topic of shell middens, an especially interesting type must be
mentioned, the shell ring. At least eight have survived (Williams 1968:253). All are
found on the coast where their range extends from Sapelo Island on the south to near
Charleston, South Carolina, on the north. They are ring-shaped middens whose walls
range from 2 to 9 feet high and from 20 to 50 feet broad at the base. Enclosed by the walls
are flat, central areas whose diameters average between 100 and 150 feet. The walls are
composed mainly of oyster shells, but also contain pottery and other artifacts as well as
vertebrate food refuse. The enclosed area, however, is surprisingly devoid of occupa
tional debris other than that secondarily washed down from the walls.
Excavations have been conducted at six of these shell rings, Sapelo Island No. 1,
Chester Field, Large and Small Ford Shell Rings, Sea Pines, and Yough Hall (Williams
1968; Waddell 1965; Calmes 1968). The first two are, of course, fiber-tempered pottery
sites, but the last four are characterized by a preponderance of sand-tempered pottery
variously classified as Awendaw or Thom's Creek. Plain and punctated types predomin
ate. The latter are obviously closely related to one another and in turn to Stallings Punc
tate. It appears that shell rings are positively associated with the Stalling's Island cul
ture and with closely related complexes that, depending upon one's point of view, could be
considered variants of the same culture. What the function of these shell rings might have
been is unknown. Waring regarded them as "certainly ceremonial enclosures of some
sort" (Williams 1968:246), although the possibility that they represent the carefully dis
carded refuse of a centrally located residential group cannot be discounted. It is interest
ing to note that similarly-shaped shell accumulations containing fiber-tempered pottery
have been reported from Florida, Colombia, and Mississippi (Bullen, Reichel-Dolmatoff,
this volume; Gagliano, personal communication). The common imperative to dispose of
inedible shells may be sufficient explanation of these parallelisms, but a historical ex
planation (e.g. stimulus diffusion) cannot be lightly dismissed.
Besides shellfish, the Stallings people exploited a wide range of game (Stoltman: in
press; Williams 1968:275; Flannery 1943:149; Claflin 1931:12). In all cases white-tailed
deer was by far most abundant with beaver a distant second and opossum, muskrat, otter,
raccoon, gray fox, skunk, rabbit, and bobcat of les ser importance. Turtles and fish were
taken in some number while turkey was the principle bird. Quail is reported only from
Any discussion of subsistence would be deficient without consideration of plant food
but that is difficult to evaluate. The available evidence suggests, however, that plant
foods were overshadowed in their importance by deer and mollusks. Caches of charred
corncobs found at Stalling's Island were originally assigned by Claflin (1931:12) to the
Stallings component, but subsequent work at the site has produced more corn that has been
radiocarbon dated to the 18th century A. D. (Bullen and Greene 1970:10). Other than these
cobs, direct evidence of plant remains from Stallings sites is non-existent.
Indirect evidence in the form of artifacts that presumably were used to process wild
plant food is available to suggest that this food source was not entirely neglected. As noted
earlier, mortars, ground stone discs, pitted stones, and variously shaped hammerstones -~
all of which could have served for processing plant foods -- occur in small frequencies at
some Stallings sites. Since all of these tool types are generalized forms capable of per
forming multiple functions, it is impossible to infer from them any more than that an un
specified amount of wild plant food was being processed. Considering the total ecological
and cultural context of these tools, it is my hunch that wild plants were of secondary im
portance in the Stallings diet.
Of course, it is possible that the currently known Stallings sites sample a limited
segment of a diversified settlement system. Perhaps the relative ease of finding shell
middens coupled with their archaeological productivity has encouraged us to overlook Stal
lings sites representing different, task-specific activities. In this regard, the dearth of
archaeological investigations in the uplands bordering the Savannah Valley is to be parti
In an effort to reduce the affects of this sampling bias, we surveyed as intensively as
we could in part of a summer the uplands adjacent to Rabbit Mount in South Carolina. The
results of this survey can hardly be considered definitive, for our time was limited and the
obstacles to discovering even large sites in the heavily wooded terrain were formidable.
With these reservations in mind, the relevant results of the survey follow.
As noted previously, the yield of Stallings ceramics from the upland sites was small,
and I have raised the possibilities that this might be explained by sparse occupation or by
activity differences of the Stallings people or by sampling error. It should also be added
that any of these factors may be simultaneously operative. Due to the nature of the sample,
the probability of sampling error remains high. Nevertheless, it is significant to note that
lithic artifacts typically associated with Stallings pottery are likewise rare in our upland
collections. Of the 386 stone artifacts found in the upland, only 118 may be of Stallings
age. These include one drill and one stemmed scraper, 14 Savannah River Stemmed pro
jectile points, and a variety of less diagnostic retouched flakes, cores, choppers, blanks,
and knives. One is left with the impression that Stallings people visited the upland infre
quently or for short periods and, considering the kinds of stone artifacts recovered, main
ly to hunt and butcher deer and to collect raw materials such as chert. If plant foods were
being gathered in the uplands, the current archaeological evidence suggests they were
being brought back to the main floodplain sites for processing.
SAVANNAH RIVER REGION
Considering the distribution and relative sizes of tailings sites on Groton Planta
tion, I concur with the Beardsley et. al. (1956:138-39) classification of the Stallings set
tlement system into their "Central-Based Wandering type. Intensively occupied sites
like Rabbit Mount and Stalling's Island must have served as the focal points or "base
camps" around which the annual gathering cycle revolved. From such base camps,
smaller groupd no doubt departed to perform specific tasks such as to gather certain
nuts in season, hunt deer, or quarry chert, only to return again from their temporary
"work camps" to the central base. The current data, at least, support such a view
of the Stallings settlement system.
The Stalling's Island Culture
Throughout the foregoing discussion, my goal has been to erect as complete a picture
as the archaeological data allow of the way of life that fostered the tradition of making Stal
lings ceramics within the Savannah River Region. I have loosely used the terms "Stallings
people" to refer to the makers of Stallings pottery. Now that much of the relevant data has
been presented, I should like to define these terms more rigorously.
While fiber-tempered pottery is unique and interesting, its significance is better un
derstood if we view it in proper perspective within the context of a total cultural system.
As already noted, the available archaeological evidence within the Savannah River Region
demonstrates a strong, positive correlation between Stallings pottery and shell middens.
I have taken these data at face value to indicate that the intensive exploitation of shellfish
along the Savannah River and the Sea Island section of the Atlantic coast permitted a semi-
sedentary settlement system of the Central-based Wandering type that in turn permitted
the successful adoption of ceramic vessel manufacturing.
However, as noted earlier, intensive exploitation of shellfish preceded and outlasted
fiber-tempered pottery in the Savannah River Region, a fact obscured by a narrow consid
eration of these ceramics alone. Since the subsistence base was surely of more fundamen
tal importance to the people than the nature of their containers, I prefer an archaeological
nomenclature that emphasizes these priorities. Accordingly, for the historically related
occupations of the Savannah River Region by pre-agricultural peoples whose way of life in
volved the intensive exploitation of shellfish, regardless of whether or what kind of cera
mics were made, I formally propose the name Stalling's Island Culture.
This terminology is not precisely like that of Claflin (1931) or Fairbanks (1942), so
a few words of clarification are necessary. While Claflin (1931:37-42) actually employed
the terms "Stalling's Island culture", his usage is more restrictive than that proposed
here. It is clear from his report that he considered fiber-tempered pottery an essential
ingredient, one might even say an index fossil, of this culture. Fairbanks (1942:230), in
adopting the nomenclature of the McKern system, referred the pre-ceramic and fiber-
tempered ceramic components of Stalling's Island to a "Savannah River focus". Miller
(1949) and Caldwell (1952) have subsequently continued this terminology. The Stalling' s
Island Culture designation proposed here is more inclusive than this Savannah River focus,
for it also subsumes historically related complexes that share the same way of life but are
characterized by different (i. e. sand tempered ) ceramics. Within this culture, various
temporally or spatially distinct "phases" in the usage of Willey and Phillips (1958) may be
recognized. Another advantage of superceding the Savannah River focus (or phase) termi
nology is that it avoids confusion with the younger and altogether different Savannah phase
of the same region.
Culture History: Origins of the Stalling1 s Island Culture
The origin of the Stalling's Island Culture, I would attribute to Archaic inhabitants
of the Georgia-Carolina Piedmont who, under influences emanating from the Tennessee
Valley of Alabama and Tennessee, moved more or less permanently into the Savannah
River Valley about 3000 B. C. Prior to this time the available archaeological evidence
suggests the Savannah Valley was occupied only intermittently by generalized hunter-
gatherers. However, with the adoption of a subsistence-economy heavily dependent on
shellfish, ancestors of the Stalling1 s Island people began to intensively occupy the Savan
nah Valley for the first time.
The presumed local ancestor of the Stalling's Island Culture is the so-called Old
Quartz Culture or Industry of the Georgia-Carolina Piedmont. While over 200 small, sur
face sites are known (Caldwell 1958:8-9), only two Old Quartz sites, Lake Spring (Caldwell
1954) and Westo Bluff (Neill 1966), both in Columbia County, Georgia, have so far been
published. Assemblages normally are dominated by flaked quartz bifaces and flakes.
Among the more characteristic artifacts are contracting stemmed, Morrow Mountain-like
projectile points and side-notched projectile points, sometimes with beveled edges. Shell
middens, pottery, and polished stone, with the exception of one quartz banner stone and 18
steatite "netsinkers" surface collected at Westo Bluff (Neill 1966), have yet to be reported
from an Old Quartz context. Although the age of Old Quartz is unknown, evidence of its
relative earliness is seen at Lake Spring where it was found beneath three to four feet of
sterile river sands that underlay the Stallings shell midden (Caldwell 1954).
The problem of transition between Old Quartz and the Stalling's Island Culture is
obscured by the paucity of data concerning the earliest or pre-ceramic phase of the latter
culture. A pre-ceramic phase has been recognized in the basal levels of Stallings shell
middens at Stalling's Island, Lake Spring, and Bilbo. It is poorly known because the non
ceramic artifacts recovered from these levels were not described (Claflin 1931: Fair
banks 1942) or the sample size was so small that the probability of bias must be consider
ed high (Miller 1949:40; Williams 1968:155). The Lake Spring data, at least, indicate that
such characteristic Stallings artifacts as Savannah River Stemmed projectile points, stea
tite "netsinkers", polished stone discs, cruciform drills, and bannerstones do occur in the
pre-ceramic component. However, recent work at the type site has revealed some unex
pected differences between the ceramic and pre-ceramic components with respect to pro
jectile points (Bullen and Greene 1970:14). In six test pits Bullen and Greene (1970) pro
duced no contracting stemmed projectile points in their deepest pre-ceramic levels (55-
72 inches). Rather, these points, 48 of 55 of which were made from quartz, were found
mainly in the ceramic levels, while the straight-stemmed, "classic", Savannah River
Stemmed points (all made of rhyolite) were recovered mainly from pre-ceramic levels.
Similarly, at Lake Spring the use of quartz clearly increased from the pre-ceramic into
the ceramic levels. Considering the stratigraphic evidence for the earliness of Old
Quartz, these trends within the Stalling's Island Culture constitute an interesting and un
explained reversal of what one would expect.
These fragmentary data considered, there is no basis for serious disagreement with
Caldwell's (1958:15) earlier statements that "There is clear evidence of cultural conti
nuity from the pre-ceramic levels of the shell heaps through the pottery-bearing zone.
Earthenware was just another innovation in a continuing way of life. The last sentence is
an important one. It emphasizes that whatever artifact differences there may have been
between pre-ceramic and ceramic components of Stalling's Island and Lake Spring, they
shared the same basic way of life, an intensive adaptation toward the Savannah River Valley
SAVANNAH RIVER REGION
with particular emphasis on shellfish gathering. In my view it is this way of life, rather
than differences in artifact types or materials, that best distinguishes the Stalling's Island
Culture from its presumed Old Quartz ancestor. Two radiocarbon dates from the pre
ceramic levels of Stalling's Island demonstrate that the new culture was in existence by
about 2750 B. C. (Samples M-l 279 and M-1277 gave 275 0 150 B.C. and 250 0 150 B.C).
Considering the broader archaeological context of the Southeast during the fourth
millennium B. C. I believe that the Stalling's Island Culture was not an entirely local de
velopment, but was influenced in its formative stages by cultures of the Tennessee Valley
to the west. There, as exemplified by the Eva site, a way of life based upon the intensive
exploitation of shellfish had begun by 5200 B.C. (Lewis and Kneberg 1961:13).
Not only are the shell middens of the Tennessee River demonstrably older than those
of the Savannah River, thus raising the possibility of ancestry, but the cultural inventories
have enough in common to suggest it. Early Lauderdale, Kays, and Three Mile are pre
ceramic phases found in Tennessee Valley shell middens. Radiocarbon dates for the for
mer two both approximate 2800 B.C. (Lewis and Kneberg 1959:169 and 178). That is,
these phases seem to precede somewhat and then parallel in time the Stalling's Island
Culture. Lewis and Kneberg (1959), in recognition of a host of cultural similarities,
have lumped the Kays and Lauderdale phases with the Stalling's Island Culture into their
"Eastern Tradition". Common traits of this tradition, and of the Three Mile phase as
well, include straight-stemmed projectile points, stemmed drills, bannerstones, atlatl
hooks, and socketed antler points. In addition, it is interesting to note that stemmed
scrapers and flaked stone adzes occur in both the Kays and Three Mile phases (Lewis and
Kneberg 1959:164, 169).
During the fourth and early third millennia B. C. while such shellfish gatherers
flourished in the Tennessee Valley, small groups of upland-adapted hunter-gatherers no
doubt sparsely inhabited much of the Appalachians to the east. Old Quartz seems to have
been a local variant of this non-shellfish-using, upland way of life shortly before 3000 B. C.
Their geographic position along the sole route of easy access through the Appalachians into
eastern Georgia, the Rabun Gap between the headwaters of the Little Tennessee and Savan
nah rivers (La Forge et al. 1925:100, 119), no doubt brought the Old Quartz people know
ledge of, if not actual contact with, the shellfish gatherers of the Tennessee Valley.
Shortly after 3000 B. C. upland-adapted, Old Quartz peoples of the Georgia-South Carolina
Piedmont, under influences emanating from the Tennessee Valley, began to exploit in
tensively Savannah River shellfish and in so-doing became the founders of the Stalling's
As for Stallings ceramics, their origins remain a mystery. They are still the old
est radiocarbon-dated ceramics in eastern North America (Bullen 1961; Stoltman 1966).
This, however, I feel is more a reflection of the incompleteness of our data than of the real
situation. Assuming fiber-tempered pottery throughout the Southeast is historically re
lated, the Savannah River Region's geographic marginality precludes it as a likely ances
tor. In view of the finds from Puerto Hormiga, Colombia (Reichel-Dolmatoff 1965),and
evidence that fiber-tempered pottery sites along the Florida Gulf Coast have been sub
merged (Bullen 1 969b; Phelps 1966a), I look for earlier sites eventually to be found in Florida
(See also Ford 1966 and Emery and Edwards 1966). In any case, the Stallings ceramic se
ries appears as a late and non-disruptive additive trait (origin unknown) in an on-going
shellfish-based economy within the Savannah River Region. Because its appearance is un
accompanied by any significant changes in the archaeological record, I doubt that its arriv
al on the Georgia-Carolina coast can be attributed to the appearance of new peoples.
Owing to the absence or presence of this trait, -what I am here calling the Stalling's
Island Culture may be conveniently subdivided into pre-ceramic and ceramic phases.
Based on the supposition that a plain-pottery-only phase preceded a decorated pottery
phase within the region, I have previously (In Press) followed Waring (Williams 1968:216-
221) in recognizing a bipartite division of the ceramic phase. I am currently ambivalent
about the advisability of this. As discussed earlier, only at Sapelo Island is there un
ambiguous evidence of a component with Stallings Plain as the sole pottery type; depending
on how the Rabbit Mount and Bilbo data are interpreted, such a component could also be
inferred for these sites. However, if the current radiocarbon dates are reasonably accu
rate, the lateness of basal Sapelo Island suggests that ceramic changes were not synchro
nous throughout the region. In that case a Stallings I II III trichotomy would be erro
neous as a regional framework. In this paper I thus offer an alternative terminology, a
dichotomy, to be tested further. I offer two models because the ambiguities of the avail
able evidence prevent us from convincingly rejecting one or the other of them. These two
terminologies, are as follows: Trichotomy Stallings I (pre-ceramic), Stallings II (plain
fiber-tempered pottery), and Stallings III (plain and decorated fiber-tempered pottery)
and Dichotomy Stallings I (pre-ceramic), Stallings II (both plain and decorated fiber-
tempered pottery). If the ceramic criteria distinguishing Stallings III from II phases fail
to appear synchronously throughout the region, then the first model should be abandoned.
What Happened to the Stalling's Island Culture?
Upstream from the vicinity of the Savannah River mouth, middens of intensive shell
fish gatherers are unreported subsequent to the Stalling's Island Culture. It would appear
that the Stallings way of life was abandoned earlier over the inland portions of the Savannah
Valley, than along the coast. Data from Rabbit Mount suggest abandonment of that site may
have been precipitated by a reduction in mussel resources due to the shift from stream to
oxbow lake conditions adjacent to the site.
Although the local geology is too imperfectly known to confirm or reject the above in
ference, sea level data suggest that the kind of phenomenon inferred for Rabbit Mount about
2000 B. C. could have been more widespread in the Savannah Valley at this time than it had
been before. According to Emery and Edwards (1966:733) sea levels in the Atlantic at the'
end of the Pleistocene rose more or less continuously until "At present the sea level is com
paratively stable and it has been sofor the past 3,000 to 4,000 years. It thus follows that
prior to 3,000 to 4, 000 years ago the Savannah1 s gradient would have been greater and that
after, because of reduced gradients, the river's propensity to meander and to overflow its
banks would have increased. It can thus be postulated on the basis of independent geologic
evidence that during the second millennium B. C. (the time of the dissolution of the Stalling's
Island Culture in much of the Savannah River Region) a modern stream regimen characteriz
ed by frequent and violent floods and fluctuations in channel position came into being. This
is not to say that floods and meandering failed to occur before 3,000 years ago, but only that
the frequency of such disruptive phenomena increased enough to be a more effective deter-
rant to human settlement of the valley.
Along the coast a more complex picture is emerging. Here shellfish were being ex
ploited by makers of fiber-tempered pottery and by people who tempered their pottery with
sand. In some cases, for example at the Refuge site (Williams 1968:198-208), the latter
are clearly post-Stallings in age. In other instances, however, the two wares appear to be
contemporary. Thus the following radiocarbon dates purport to date Awendaw and/or
Thom's Creak ceramics (Fig. 1 6, g-k) at four South Carolina shell rings (Waddell 1965;
SAVANNAH RIVER REGION
The two oldest of these dates fall comfortably within the range of the dates associated with
fiber-tempered pottery at the nearby Bilbo site, while the other four dates are younger than
all fiber-tempered pottery dates from the Savannah River Region. (This statement assumes,
of course, that the mean value of each date offers the best estimate of the dated sample's
The interpretation of these dates is complicated by the absence of site reports. All
we know of the Yough Hall shell ring is that the "prevalent" pottery there belongs to the
Awendaw complex, a coastal variant of Thom1 s Creek (Waddell 1965:82; Williams 1968:321).
Of the other three sites, we know only that fiber-tempered sherds constitute a minority of
the ceramic sample from two, while Thom's Creek types (Awendaw?) comprise the total
sample of 87 sherds from the Small Ford Shell ring (Calmes 1968:46). As long as multiple
components are indicated at the Yough Hall, Large Ford, and Sea Pines sites, and the ce
ramic types found in association with the dated charcoal or shell samples remain unreport
ed, it would be premature to accept the dates at face value. The Small Ford site date
(1-3047) is a different matter, however, for the ceramic sample, small though it is, sug
gests a single-component site. Assuming the sample is unbiased, we are left with the
possibility that sand-tempered, punctated ceramics on the South Carolina coast coexisted
with the Stallings series of the Georgia coast. This does not preclude the generally ac
cepted view that Thom's Creek/Awendaw ceramics are descended from the Stallings series,
but it requires us to allow for the possibility that such a descent was a localized rather than
a pan-regional phenomenon. Indeed, there is every reason to believe that, while the
Thom's Creek/Awendaw ceramics may have emerged before fiber-tempered pottery ceased
being made, the former out-lasted the latter in the Savannah River Region.
The differences between Thom's Creek and Awendaw are poorly defined. Waring,
who at one time referred to the latter as Horse Island, (Williams 1968:218-19, 330) re
garded it as a coastal variant "related to but not identical to the Thom's Creek Ceramic
Complex." Waddell's( 1 965:84) map shows Awendaw sites restricted to the South Caro
lina coast, which is in contrast to Thom's Creek sites that are found scattered well inland
(Phelps 1968; Griffin 1945a). While the precise ceramic differences between the two com
plexes remain obscure, at least one significant difference exists -- coastal (Awendaw?)
sites often have shell middens, whereas inland Thom's Creek ceramics are unreported in
shell mound contexts. At Rabbit Mount, for example, we found 71.7 per cent (38) of the
site's 53 Thom's Creek Punctate sherds in portions of the site where shell midden was
lacking. And, of those 15 sherds recovered from portions of the site with shell midden,
12 came from above the midden, 2 from the midden (which I attribute to mixture), and one
from uncertain provenience. Both the vertical and horizontal distributions indicate clearly
that the Thom's Creek people who visited Rabbit Mount were not engaged in shellfish
Conceptually, I feel it is useful to maintain an Awendaw/Thom's Creek distinction be
cause their life-ways appear to have been drastically different, unless, of course, these
should prove to be seasonal variants of a single culture system. Whether or not the sup
posed ceramic differences hold up under closer scrutiny, the current evidence suggests
that, while the coastal (i. e. Awendaw) people persisted in the old life-way based heavily on
shellfish gathering, the inland (i. e. Thom1 s Creek) people embarked on a new way of life.
Thus, I consider both Thom's Creek and Awendaw descendants of the Stallings Ceramic
Phase, but only the latter part of the Stalling's Island Culture.
Following the Awendaw phase, shellfish gathering continued on the coast, but whether
or not there is historical and cultural continuity between the peoples is unclear. It is poss
ible that the Refuge phase represents a terminal manifestation of the Stalling's Island Cul
ture on the coast. By Deptford times, however, it is clear we are dealing with a new way
of life; some would even say Deptford represents an intrusive cultural entity into the Sa
vannah River Region (Phelps 1968:29). In any case about 1000 B.C. on the coast, a way of
life whose historical roots we have traced back to about 3,000 B. C. in the inland portions
of the Savannah Valley can be seen to be waning. In the ensuing millennium new cultural
traditions developing elsewhere in the Eastern Woodlands began to intrude upon the Savan
nah River Region inaugurating a new era.
SAVANNAH RIVER REGION
Fig. 15. Fiber-tempered sherds from Rabbit
a-b, Stallings Plain; c-d, Stallings Incised; e-h,
Mount, South Carolina.
Stallings Simple Stamped.
Fig. 16. Fiber- and sand-tempered sherds from Rabbit Mount, South Carolina.
a-f, Stallings Punctate; g-k, Thom's Creek Punctate.
Fig. 17. Flaked stone artifacts from Rabbit Mount, South Carolina.
a-e, Savannah River Stemmed projectile points; f-i, narrow, bifacial knives;
j 1, chert adzes; m-o, retouched unifacial flakes; p, flake with engraving spur.
Fig. 18. Pecked, battered, and polished stone artifacts, Rabbit Mount.
a~b, discoidal chert hammerstones; c, mortar fragment; d, spheroidal chert ham-
merstone; e, grooved sandstone net sinker; f-g, perforated steatite slabs (net sinkers);
h, j, irregular quartz discs with ground edges; i, fragment three-quarter grooved axe.
SAVANNAH RIVER REGION
Fig. 19. Artifacts of bone and antler from Rabbit Mount, South Carolina.
a, long, plain bone pin; b-e, decorated bone pins; f, long bone point;
g-h, bone awls;_i, wide pointed bone tool (spatulate form); j, socked
antler point; k, bone atlatl hook; 1, cut and socketed antler handle.
Atkins, Steve, and Jeannie MacMahan
1967 The Zabski site, Merritt Island, Florida. Florida Anthropologist, vol. 20,
nos. 3-4, pp. 133-45. Tallahassee.
Beardsley, Richard K. et al
1956 Function and evolutionary implications of community pattering. In Robert
Wauchope (ed.), Seminars in Archaeology: 1955. Memoirs of the Society
for American Archaeology, no. 11, pp. 129-55. Salt Lake City.
Braun, E. Lucy
1950 Deciduous Forests of Eastern North America. The Blakiston Co.
Brush, Charles F.
1965 Pox pottery: Earliest identified Mexican ceramics. Science, vol. 149,
no. 3680, pp. 194-5. Washington.
Buckley, James D. and Eric H. Willis
1969 Isotopes' radiocarbon measurements VII. Radiocarbon, vol. 11, no. 1,
pp. 53-105. New Haven.
Bullen, Ripley P.
1954 Culture changes during the fiber-tempered period of Florida.
Southern Indian Studies, vol. 6, pp. 45-48. Chapel Hill
1955 Stratigraphic tests at Bluffton, Volusia County, Florida. Florida
Anthropologist, vol. 8, no. 1, pp. 1-16. Gainesville.
1958a The Bolen Bluff site on Paynes Prairie, Florida. Contributions of the Florida
State Museum, Social Sciences, no. 4. Gainesville.
1958b Six sites near the Chattahoochee River in the Jim Woodruff Reservoir area,
Florida. River Basin Surveys Papers, no. 14, Bureau of American Ethnology,
Bulletin 1 69, pp. 315-57. Washington.
1959 The transitional period of Florida. (15th) Newsletter, Southeastern Archaeo
logical Conference, vol. 6, pp. 43-53. Chapel Hill.
1960 The earliest pottery in southeastern United States, 2000-1000 B. C. and its
case as an independent invention. Des Actes du VI Congres International des
Sciences Anthropologiques et Ethnologiques, Tomo 2, (Ier vol.), pp. 363-67.
1961 Radiocarbon dates for southeastern fiber-tempered pottery. American Anti
quity, vol. 27, no. 1, pp. 104-6. Salt Lake City.
1962 Indian burials at Tick Island. American Philosophical Society Yearbook 1961,
pp. 477-80. Philadelphia.
1969a Excavations at Sunday Bluff, Florida. Contributions of the Florida State
Museum, Social Sciences, no. 15. Gainesville.
1969b Further Comments on Emery and Edwards' "Archaeological Potential of the
Atlantic Shelf". American Antiquity, vol. 34, no. 3, pp. 331-332.
Salt Lake City.
Bullen, Ripley P. and Adelaide K.
1950 The Johns Island site, Hernando County, Florida. American Antiquity, vol.
12, no. 1, pp. 23-45. Menasha.
1953 The Battery Point site, Bayport, Hernando County, Florida Florida Anthro-
pologist, vol. 6, no. 3, pp. 85-92. Gainesville.
1961 The Summer Haven site, St. Johns County, Florida. Florida Anthropologist,
vol. 14, nos. 1-2, pp. 1-15. Gainesville.
MS The Palmer site, Sarasota County, Florida. Manuscript at Florida State
Museum, Gainesville, Florida.
Bullen, Ripley P. and Adelaide K. and William J. Bryant
1967 Archaeological investigations at the Ross Hammock site, Florida. William
L. Bryant Foundation, American Studies, no. 7. Orlando.
Bullen, Ripley P. and Adelaide K. and Carl Clausen
1968 The Cato site near Sebastian Inlet, Florida. Florida Anthropologist, vol. 21,
no. 1, pp. 15-16. Tallahassee.
Bullen, Ripley P. and Harold K. Brooks
1967 Two ancient Florida dugout canoes. Quarterly Journal of the Florida Academy
of Sciences, vol. 30, no. 2, pp. 98-107. Gainesville.
Bullen, Ripley P. and William J. Bryant
1965 Three Archaic sites in the Ocala National Forest, Florida. William L.
Bryant Foundation, American Studies, no. 6. Orlando.
Bullen, Ripley P. and H. Bruce Greene
1970 Stratigraphic tests at Stalling' s Island, Georgia. Florida Anthropologist,
vol. 23, no. 1, pp. 8-28. Gainesville.
Bullen, Ripley P. and John W. Griffin
1952 An archaeological survey of Amelia Island, Florida. Florida Anthropologist,
vol. 5, nos. 3-4, 37-64. Gainesville.
Caldwell, Joseph R.
1950 A preliminary report on excavations in the Allatoona Reservoir, Early
Georgia, vol. 1, no. 1, pp. 4-21. Athens.
1952 The archeology of eastern Georgia and South Carolina. In James B. Griffin
(ed. ) Archeology of Eastern United States, pp. 312-321. Chicago.
1954 The Old Quartz Industry of Piedmont Georgia and South Carolina. South
Indian Studies, vol. 5, pp. 37-39. Chapel Hill.
1958 Trend and tradition in the prehistory of the eastern United States, American
Anthropological Association, Memoir, 88. Menasha.
Caldwell, Joseph and Catherine McCann
1941 Irene Mound Site, Chatham County, Georgia. University of Georgia Press.
1968 Test excavations at three late Archaic shell-ring mounds on Hilton Head
Island, Beaufort County, South Carolina. Southeastern Archaeological
Conference, Bulletin 8, pp. 45-48. Morgantown.
Childe, V. Gordon
1951 Man Makes Himself. Mentor Books. New York.
Claflin, William H. Jr.
1931 The Stalling1 s Island mound, Columbia County, Georgia. Papers of the
Peabody Museum of American Archaeology and Ethnology, Harvard
University, vol. 14, no. 1. Cambridge.
Coates, Gordon C.
1955 Recent tests at the Battery Point site, Bayport, Hernando County, Florida.
Florida Anthropologist, vol. 8, no. 1, pp. 20-27. Gainesville.
Coe, Joffre Lanning
1964 The formative cultures of the Carolina piedmont. Transactions of the
American Philosophical Society, N. S. vol. 54, part 5. Philadelphia.
Cooke, C. Wythe
1943 Geology of the coastal plain-of South Carolina. U. S. Geological Survey,
Bulletin 941. Washington.
Emery, K. O. and R. L. Edwards
1966 Archaeological potential of the Atlantic continental shelf. American
Antiquity, vol. 31, no. 5, pp. 733-737. Salt Lake City.
Evans, Clifford, and Betty Meggers
1965 Cronologia relativa y absoluta en la Costa del Ecuador. In Cuadermos
de Historia y Arqueologia. Casa de la Cultura Ecuatoriana. Ncleo del
Fairbanks, Charles H.
1942 The taxonomic position of Stalling's Island, Georgia. American
Antiquity, vol. 7, no. 3, pp. 223-231. Menasha.
Faulkner, Charles H. and J.B. Graham
1966 Westmoreland-Barber Site (40 Mi-11), Nickajack Reservoir, Season II.
Department of Anthropology, University of Tennessee. Knoxville.
Ferguson, Vera Masius
1951 Chronology at South Indian Fields, Florida. Yale University Publications
in Anthropology, no. 45. New Haven.
1943 Some notes on a few sites in Beaufort County, South Carolina. Bureau of
American Ethnology, Bulletin 133 (Anthropological Papers, no. 21),
pp. 143-153. Washington.
Ford, James A.
1966 Early formative cultures in Georgia and Florida. America Antiquity,
vol. 31, no. 6, 781-99. Salt Lake City.
1969 A comparison of formative cultures in the Americas, diffusion or the psychic
unity of man. Smithsonian Contributions to Anthropology, vol. 2. Washington.
Ford, James A. and George I. Quimby, Jr.
1945 The Tchefuncte culture, an early occupation of the lower Mississippi Valley.
Memoirs of the Society for American Archaeology, no. 2. Menasha.
Goggin, John M.
1952 Space and time perspective in northern St. Johns archeology. Yale
University Publications in Anthropology, no. 47. New Haven.
1954 A preliminary statement on the archeology of the Cape Haze area, south
western Florida, mimeographed. Department of Sociology and Anthro
pology, University of Florida. Gainesville.
Gregorie, Anne King
1925 Notes on Sewee Indians and Indian Remains of Christ Church Parish,
Charleston County, South Carolina, The Charleston Museum. Charleston.
Griffin, James B.
Report on the ceramics of Wheeler Basin. In An Archeological Survey of
Wheeler Basin on the Tennessee River in Northern Alabama, by W. S. Webb.
Bureau of American Ethnology, Bulletin 122. Washington.
An analysis and interpretation of the ceramic remains from two sites near
Beaufort, South Carolina. Bureau of American Ethnology, Bulletin 133
(Anthropological Paper, no. 22), pp. 155-168. Washington.
1945a Ceramic collections from two South Carolina sites. Papers of the Michigan
Academy of Science, Arts and Letters, vol. 30, pp. 465-476. Ann Arbor.
1945b The significance of the fiber-tempered pottery of the St. Johns area in Flor-
ida, Journal of the Washington Academy of Sciences, vol. 35, no. 7.
Culture periods in Eastern United States archeology. In J. B. Griffin, ed. ,
Archeology of Eastern United State s, pp. 352-364. Chicago.
Some prehistoric connections between Siberia and America. Science,
vol. 131, no. 3403, pp. 801-812. Washington.
Mesoamerica and the eastern United States in prehistoric times. Handbook
of the Middle American Indians, vol. 4, pp. 111-30. Austin.
Griffin, John W
Archeological investigations at Russell Cave, Alabama. Archeological
Research Series, National Park Service. (In press. )
Griffin, John W., and Hale G. Smith
Nocoroco, a Timucua village of 1605 now in the Tomoka State Park.
Florida Historical Quarterly, vol. 22, no. 4, pp. 340-61. St. Augustine.
The Cotten site: an archaeological site of early ceramic times in Volusia
County, Florida. Florida State University Studies, no. 16, pp. 27-60.
Haag, William G.
Type descriptions. Southeastern Archaeological Conference, Newsletter,
Vol. 1, No. 1.
A description of the Pickwick pottery. In An Archaeological Survey of the
Pickwick Basin in Adjacent Portions of the States of Alabama, Mississippi
and Tennessee, by W. S. Webb and D. L. DeJarnette. Bureau of American
Ethnology, Bulletin 129. Washington.
Heimlich, Marion Dunlevy
1952 Guntersville Basin pottery. Geological Survey of Alabama, Museum
Paper 32, University of Alabama.
Hemmings, E. Thomas
1969 An early ceramic site near Beaufort, South Carolina. Notebook, vol. 1,
nos. 6-9. Institute of Archeology and Anthropology. Columbia.
1938 Excavations on Saint Simons Island and vicinity (winter 1936-37).
Proceedings of the Society for Georgia Archaeology, vol. 1, no. 1, pp. 8-9.
Hole, Frank, Kent Y. Flannery, James A. Neely
1971 Prehistory and human ecology of the Deh Luren Plain (excerpts). In
Prehistoric Agriculture (Stuart Struener, ed.), pp. 252-311).
Natural History Press, Garden City, N. Y.
Holmes, William H.
1894 Earthenware of Florida. Journal of the Academy of Natural Sciences of Phila-
delphia, vol. 10, no. 1, pp. 106-28. Philadelphia.
Jones, Charles C., Jr.
1873 Antiquities of the Southern Indians, Particularly of the Georgia Tribes.
D. Appleton and Company. New York.
Kenyon, Kathleen M.
1967 Jericho: Oldest walled town. In Archaeological Discoveries in the Holy Land,
pp. 19-28. Thomas Y. Crowell Company. New York.
La Forge, Laurence, Wythe Cook, Arthur Keith, and Marius R. Campbell
1925 Physical Geography of Georgia. Geological Survey of Georgia, Bulletin
No. 42. Atlanta.
Lewis, T.M. N. and Madeline Knberg
1959 The Archaic culture in the Middle South. American Antiquity, vol. 25,
no. 2, pp. 161-183. Salt Lake City.
1961 Eva, An Archaic Site. The University of Tennessee Press. Knoxville.
Libby, Willard F,
1952 Chicago radiocarbon dates III. Science, vol. 116, pp. 673-681.
Marks, Anthony E. Joel L. Shiner, and T. R. Hays
1968 Survey and excavations in the Dongola Reach, Sudan. Current Anthropology,
vol. 9, no. 4, pp. 319-23. Chicago.
McDowell, L. L. J. C. Stephens, and E.H. Stewart
1969 Radiocarbon chronology of Florida Everglades pest. Proceeding Soil
Science Society of America, vol. 33, no. 5, pp. 743-45"! Madison.
Meggers, Betty J. Clifford Evans and Emilia Estrada
1965Early formative period of Coastal Ecuador: The Valdivia and Machililla
Phases. Smithsonian Contributions to Anthropology, vol. 1. Washington.
Miller, Carl F.
1949 The Lake Spring site, Columbia County, Georgia. American Antiquity,
vol. 15, no. 1, pp. 38-51. Menasha.
1950 Early cultural horizons in the Southeastern United States. American
Antiquity, vol. 15, no. 4, pp. 273-288. Menasha.
Moore, Clarence B.
1892-94 Cerain shell heaps of the St. John's River, Florida, hitherto unexplored.
American Naturalist, vols. 26-28. Philadelphia.
Neill, Wilfred T.
1966Westo Bluff, A site of the Old Quartz culture in Georgia. Florida
Anthropologist, vol. 19, no. 1, pp. 1-10. Gainesville.
1967The dawn of history in southern Palestine. In Archaeological Discoveries in
the Holy Land, pp. 3-8. Thomas Y. Crowell Company. New York.
1965 The Norwood series of fiber-tempered ceramics. Proceedings of the 20th
Southeastern Archaeological Conference, Bulletin No. 2, pp. 65-69.
1966a Early and late components of the Tucker site. Florida Anthropologist, vol. 19,
no. 1, pp. 11-38. Gainesville.
1966b Deptford Cross-Stamped: A preliminary statement. Newsletter South-
eastern Archaeological Conference, vol. 10. Morgantown.
1968 Thom's Creek ceramics in the central Savannah River locality.
Florida Anthropologist, vol. 21, no. 1. pp. 17-30. Tallahassee.
Phelps, David Sutton, and Rebekah Burgess
1964A possible case of cannibalism in the Early Woodland period of eastern
Georgia. American Antiquity, vol. 30, no. 2, pt. 1, pp. 199-202.
Salt Lake City.
I960Prehistory and evolutionary theory. In Sol Tax, editor, Evolution
After Darwin, vol. II, pp. 85-97. Chicago.
1961Puerto-Hormiga: un complejo prehistrico marginal de Colombia (Nota
preliminar). Revista Colombiana de Antropologa, vol. 10, pp. 349-54.
1965Excavations Arquelogicas en Puerto Hormiga. Ediciones de la Universidad
de los Andes, Antropologa 2. Bogata.
Ritchie, William A.
1962The antiquity of pottery in the Northeast. American Antiquity, vol. 27,
no. 4, pp. 583-4. Salt Lake City.
1951 A survey of Indian River archeology, Florida. Yale University Publica-
tions in Anthropology, no. 44. New Haven.
Sears, William H. and James B. Griffin
1950 Fiber-tempered pottery of the Southeast. In James B. Griffin (ed.),
Prehistoric Pottery of the Eastern United States. Ann Arbor.
1964 Reports on current field work: North Carolina. Southeastern Archaeo
logical Conference, Bulletin No. 1, p. 5. Cambridge.
Stoltman, James B.
1966New radiocarbon dates for Southeastern fiber-tempered pottery. American
Antiquity, vol. 31, no. 6, pp. 872-874. Salt Lake City.
In Press Groton Plantation: An archaeological study of a South Carolina locality. Pea
body Museum of Archaeology and Ethnology, Harvard University. Cambridg
1965 The Gulf Stream. 2nd. Edition. University of California Press. Berkeley.
Suhm, Dee Ann, and Alex D. Krieger
1954 An introductory handbook of Texas archeology. Bulletin of the Texas
Archeological Society, vol. 25. Austin.
Waddell, Eugene G.
1963 Thom's Creek Punctate. Newsletter of the Southeastern Archaeological
Conference, vol. 9, no. 2, pp. 3-5. Cambridge.
1965 A C-14 date for Awendaw Punctate. Southeastern Archaeological Con
ference, Bulletin no. 3, pp. 82-85. Cambridge.
Waring, Antonio J. Jr.
1939 Fiber-tempered wares on the Georgia coast. Southeastern Archaeological
Conference Newsletter, vol. 2, no. 1, pp. 7-8.
Warren, Lyman O. William Thompson, and Ripley P. Bullen
1967 The Culbreath Bayou site, Hillsborough County, Florida. Florida Anthro
pologist, vol. 20, nos. 3-4, pp. 146-63. Tallahassee.
1966 Archaeological survey of northern Georgia. Memoirs of the Society for
American Archaeology, No. 21. Salt Lake City.
Weaver, Elizabeth C.
1963 Technological analysis of prehistoric lower Mississippi ceramic materials:
A preliminary report. American Antiquity, vol. 29, no. 1, pp. 49-56.
Salt Lake City.
Webster, William J.
1970 A new concept for the Busycon shell receptacle Florida Anthropologist,
vol. 23, no. 1, pp. 1-7. Gainesville.
Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast. Smithsonian Miscellaneous
Collections, vol. 113. Washington.
An Introduction to American Archaeology. Prentice-Hall, Inc.
Willey, Gordon R. et al
1956 An archaeological classification of culture contact situations. In Robert
Wauchope (ed. ) Seminars in Archaeology: 1955. Memoirs of the Society
for American Archaeology, no. 11, pp. 1-30. Salt Lake City.
Willey, Gordon R. and Philip Phillips
1958 Method and Theory in American Archaeology. University of Chicago
Williams, Stephen (ed.)
1968 The Waring papers: The collected works of Antonio J. Waring, Jr.
Papers of the Peabody Museum of Archaeology and Ethnology, Harvard
University, vol. 58. Cambridge.
Wimberly, Stephen B.
I960 Indian pottery from Clarke County and Mobile County, Southern Alabama.
Alabama Museum of Natural History, Museum Paper 36. University.
1875 Fresh-water shell mounds of the St. John's River, Florida.
Peabody Academy of Science, Memoir, vol. 1, no. 4. Salem.
PUBLICATIONS OF THE FLORIDA ANTHROPOLOGICAL SOCIETY
PUBLICATIONS of the Society appear at irregular intervals as funds permit. These are
provided by Sustaining Membership dues, special gifts, or subsidies provided by the au
thor. PUBLICATIONS are distributed to all members of the Society on issuance. Addi
tional copies may be secured as follows:
1. Two Archeological Sites in Brevard County, Florida, by Hale G. Smith,
31 pp. 2 figs. 4 pi. 1949. Out of print.
2. The Safety Harbor Site, Pinellas County, Florida, by John W. Griffin and
Ripley P. Bullen, 42 pp. 2 figs. 4 pis. I960. Out of print.
3. The Terra Ceia Site, Manatee County, Florida, by Ripley P. Bullen,
48 pp. 6 figs. 7 pis. 1951. Out of print.
4. The European and the Indian by Hale G. Smith, 150 pp. 6 maps,
1 plate, 1956. Out of print.
5. Florida Anthropology edited by Charles H. Fairbanks, 81 pp. 1 chart,
4 summary articles, 18 pp. bibliography, 1958. Out of print.
6. Fiber-tempered Pottery in Southeastern United States and Northern Co
lombia: Its Origins, Context, and Significance edited by Ripley P. Bullen
and James B. Stoltman, 72 pp. 3 figs. 16 pis. 1972 $2. 00
PUBLICATIONS 1-5 are out of print but may be secured from Johnson Reprint
Corporation, 111 Fifth Avenue, New York, N. Y. 10003, at $3.50 per number.
Number 6 may be secured from the Treasurer (see inside front cover).