An Archeological Survey of Chert Quarries in Western Allendale Co

University of South Carolina
Scholar Commons
Research Manuscript Series
Archaeology and Anthropology, South Carolina
Institute of
7-1984
An Archeological Survey of Chert Quarries in
Western Allendale County, South Carolina
Albert C. Goodyear
University of South Carolina - Columbia, goodyear@mailbox.sc.edu
Tommy Charles
Follow this and additional works at: htp://scholarcommons.sc.edu/archanth_books
Part of the Anthropology Commons
Tis Book is brought to you for free and open access by the Archaeology and Anthropology, South Carolina Institute of at Scholar Commons. It has
been accepted for inclusion in Research Manuscript Series by an authorized administrator of Scholar Commons. For more information, please contact
SCHOLARC@mailbox.sc.edu.
Recommended Citation
Goodyear, Albert C. and Charles, Tommy, "An Archeological Survey of Chert Quarries in Western Allendale County, South Carolina"
(1984). Research Manuscript Series. Book 187.
htp://scholarcommons.sc.edu/archanth_books/187
An Archeological Survey of Chert Quarries in Western Allendale County,
South Carolina
Keywords
Excavations, Indians of North America, Stone tools, Chert quarry survey, Allendale County, South Carolina,
Archeology
Disciplines
Anthropology
Publisher
Te South Carolina Institute of Archeology and Anthropology--University of South Carolina
Comments
In USC online Library catalog at: htp://www.sc.edu/library/
Tis book is available at Scholar Commons: htp://scholarcommons.sc.edu/archanth_books/187
B
D
u
D
D
o
ti

o
".
AN ARCHEOLOGICAL SURVEY OF CHERT QUARRIES
IN WESTERN ALLENDALE COUNTY, SOUTH ....
by
Albert C. Goodyear and Tommy Charles
with an Appendix by Sam B. Upchurch
Research Manuscript Series 195
This project was supported, in part, by a grant from
the United States Department of the Interior, National
Historic Preservation Act of 1966 (80 Stat. 915) and
amendments, administered through the South Carolina
Department of Archives an4 History, Memorandum of
Agreement dated June 7, 1983.
Institute of Archeology and
. Columbia, South Carolina 29208
July 1984
. ;
......
, ~
~ '
'-
The University of South Carolina Hoffers equal opportunity in its
employment, admissions and educational activities, in accordance
with Title IX, Section 504 of the Rehabilitation Act of 1973 and
'other. 'civii rights laws.
ii
U
D
D
D
D
o
D
D
o
o
D
D
D
D
o
D
o
[J
L
"I observed[,] rideing over A rint of A hill[,],
prodigious numbers of flints[, of as many dimen-
tions asshapes[,J &many great rocks of good flint
A little above y common surface[.] I admired at
first how thay came to be broke into various
shapes[ ,] most of them with A smooth polish: I
asked one that rode A mile with us if he had not
often found in ye fields Indian arrow JOints &
flint knives &hatchets: I was answered[,["]very
frequently[tt]: this accounted for ye numbers that
lay scattered on ye surface[,] which did not imme-
diately answer thair purpose[.] but A gentleman
observed to me another cause which seemed alIso
probable[:] that thair fires in ye woods[,]
which thay set in 1 spring to burn up ye leaves &
matterL,] which heats ye mass of flints
near y surface:".
Quote by
John Bartram
Diary of a Journey Through the Carolinas,
Georgia, and Florida
From July 1, 1765 to April 10, 1766
iii
iv
v
TABLE OF CONTENTS
38AL141 Johnsons Landing
Other Lithic Outcrops and Quarries in
Allendale County, South Carolina
38AL149 County Road 66 Outcrop
38AL150 Millett Recreation Center..
38AL1 Little Hell Landing
38AL151 Fastwater Outcrop. .
38AL148 Kerwin Outcrop .
38AL14 The Rice Quarry ("Red Bluff Flint Quarries").
Other Outcrops and Quarries in Burke and Screven
Counties, Georgia io .'.
U.S. 301, Allied Chemical Sylvania Terminal
Screven County, Georgia
Wade Plantation Quarry No.1, Screven County, Georgia.
9BUR6 Stony Bluff.
9BUR59 .
9BUR11
CHAPTER I: THE ALLENDALE COUNTY CHERT QUARRY SURVEY:
HISTORICAL BACKGROUND, GOALS, AND METHODS.
1
4-
8
8
13
15
x
vii
15
15
18
29
29
32
42
51
55
63
78
79
80
82
93
96
97
97
97
97
98
101
101
106
106
107
107
107
109
Page
xiii
... .
.
. . .
. . . . . . .
. . . . . .
e.
SITE DESCRIPTIONS
Sl1lllD1ary . . . . . . . . .
38AL140.
38AL136.
38AL138. .
38AL1.52.
38AL137 The Swag Site.
38AL146 "The Gulley Site"
38AL145 "Boundary Marker 232".
38AL161.
38AL143 Big Pine Tree Site
38AL135 .
Summary . . . .
38AL142 White Woman's Point.
38AL139 The Topper Quarry
38AL23
Introduction.
The Allendale Chert of South Carolina
Purpose of This Report
Survey Objectives and Methods.
,Artifact Analysis.
ACKNOWLEDGMENTS.
LIST OF FIGURES.
CHAPTER II:
LIST' OF TABLES
o
u
u
u
u
u
u
u

D
D
U
D
11,;.
I.t
D
D
U
U
o
TABLE OF CONTENTS (Cont.)
Page
CHAPTER II (Cont.)
9EUR60 Roy Hillis Road Quarry.
9BUR61 Ellison Bridge Road
. . . . . . . . . .
CHAPTER III:
REFERENCES
APPENDIX A
CONCLUSIONS.
. . . . . . . .
. . . . . .
. . e.
. . . .
. . . . . . . . .
. . . . .
J09
109
113-
119
125
vi
vii
LIST OF FIGURES
6
16
12
11
17
10
35
36
57
58
43
64
65
17
30
30
Page
Hatched
Hatched
Hatched 2c. Map showing location of Chert Quarry survey.
areas indicate land and roads surveyed
2a. Map showing location of Chert Quarry survey.
areas indicate land and roads surveyed
2b. Map showing location of Chert Quarry survey.
areas indicate land and roads suryeyed
FIGURE 1. Location of Red Bluff Flint Quarries, or Rice Quarry
. ( 38AL14) in Allendale County, South Carolina.. Also
shown is the geographic distribution of Allendale Chert
artifacts in private collections
FIGURE
FIGURE
FIGURE 5. Chert boulders and debris on hillside of 38AL136,
the Beech Tree Site
FIGURE
FIGURE 3. Locator map for sites 38AL140, 38AL136, 38AL138,
38AL152, and 38AL137
FIGURE 4. Chert boulders and debris on hillside of 38AL140
FIGURE 9. Site map showing auger test locations at 38AL146,
the Gulley Site.
FIGURE 8. Locator map for sites 38AL146, 38AL145, 38AL161,
38AL143, and 38AL135
FIGURE 7. Inside of unusual pit showing bedrock at 38AL152
Boundary Marker 232 si te
Big Pine Tree site
FIGURE 12. Lithic r t i f t ~ recovered from auger testing
at 38AL143, Big Pine Tree site
FIGURE 13. Profile showing natural stratigraphy from eroded bank
at 38AL135. Trowel points to clusters of flakes lying
in lower most zone of site
FIGURE 6. Chert boulders and debris on hillside of 38AL152
FIGURE 14. Closeup view of lowermost zone in bank profile of
38AL135. Note the thin microlenses of alluvially
deposited sand and weathered flakes lying flat
LIST OF FIGURES (Cont.)
Page
...
FIGURE 15. Weathered, light colored unifacial and bifacial
artifacts' presumed washed out from the lowermost zone
on the beach of 38AL135. .. 66
FIGURE 16. Topographic map and locations of subsurface testing
at 38AL135
...
68
FIGURE 17. Weathered, light colored unifacial and bifacial
artifacts recovered from lower most zone of
38AL135 by excavation. 69
FIGURE 18. Test pit N205, R197 at 38AL135. PhotograPh shows
placement of two meter squares within backhoe pit.
Bottom of square is 97 cm below surface. 71
FIGURE 19. Lithic artifacts from excavation of N205, R197
at 38AL135 . 73
FIGURE 20. Lithic artifacts recovered from the beach (38AL135).
Specimen h found in situ in the bank at 77 77
FIGURE 21. Bricks and chert debris in the roadbed at 38AL142,
Whi te Woman's Point site . 79
FIGURE 22. Large chert boulder at 38AL142, White Woman's Point
site ....... 80
FIGURE 23. Locator map for sites 38AL142, 38AL139, 38AL23
and 38AL141. 81
FIGURE 24. Small unnamed creek filled with chert boulders and
debris running through 38AL139, the Topper Quarry. 82
FIGURE 25. Pot sherds from surface and excavations at 38AL123 84
FIGURE 26. Lithic artifacts from surface and excavations
at 38AL23. 85
FIGURE 27. Profile of north wall of x 2 meter pit at 38AL23 89
FIGURE 28. View of chert outcrop in bank of Savannah River
at 38AL141, Johnsons Landing. 96
FIGURE 29. Silicified sandstone boulders at 38AL149, where
County Road 66 crosses Lower Three Runs Creek. 98
FIGURE 30. Locator map for other outcrops and quarries in
Allendale County, South Carolina, and Burke and
Screven counties, Georgia.
viii
99
LIST OF FIGURES (Cont.)
Page
FIGURE 31. Silicified sandstone bouiders at 38AL150, near
Millet, South Carolina. 100
FIGURE 32. Silicified sandstone boulders at 38AL1, Little
Hell Landing
FIGURE 33. Sandstone and chert boulders at 38AL151, Fastwater
Outcrop on the Savannah River.
FIGURE 34. Chert and limestone boulders at 38AL148, the
Kerwin Outcrop near Barton, South Carolina
100
102
102
FIGURE 35. Profile of 5 x 5 foot test pit at 38AL14, the Rice
Quarry. (Photo by James L. Michie, 1971). 103
FIGURE 36. Selected lithic artifacts from test excavations at
38AL14, the Rice Quarry. 104
FIGURE 37. Chert boulders from outcrop in swamp of Watch Call
Creek, 38AL14, the Rice Quarry. 105
FIGURE 38. The outcrop and quarry at Stony Bluff, 9BUR6 on the
Savannah River, Georgia. 1Off
FIGURE 39. Chert debris on recently cleared land at 9BUR59,
Georgia. . . . . . . . . . . . . . . . . . . 108
FIGURE 40.
FIGURE 41.
Chert debris in roadbed of 9BUR11 , Georgia
Profile in road cut of 9BUR11 showing depth of
artifacts in sand on west" side of road
. . . . .
110
110
FIGURE 42. Chert debris at Quarry 9BUR61 on Ellison Bridge
Road, Georgia. . . . . .. 111
ix
0
0
LIST OF TABLES
0
Page
D
-TABLE 1- Artifact distributions by depth
--
from Auger Test 1,
38AL136, The Beech Tree Site.

--
- .

.-
. .
_.
19
TABLE 2. Artifact distributions by depth from Auger Test 2,
U
38AL136, The Beech Tree Site. -

. . . 20
D

21
U

22

2'3
U TABLE 6. Artifact distributions by depth from Auger Test 7,

24

TABLE 7. Artifact distributiomi -by depth from Auger Test 1,
38AL138

26
U
38AL138

27
E
38AL138

28

38AL137, The Swag Site.

31
U
TABLE 11- Artifact distributions by depth from Auger Test 2,

33
U

34
U
38AL146

37
U
38AL146

38
38AL146

39
U
38AL146

40
U
0
x
xi
LIST OF TABLES (Cant. )
Page
TABLE
33. Vertical distributions
of artifacts .from Auger Test
1,
38AL23. . . . .

. .

.

86
TABLE 34.
Vertical distributions
of artifacts
from 1 X 2M
test excavation,
38AL23

..

87
TABLE 35.
Vertical distributions
of artifacts
from Auger Test 2,
.38AL23.

91
TABLE 36.
Vertical distributions of artifacts
from Auger Test
3,
38AL23.

92
TABLE 37.
Vertical
distributions
of artifacts
from Auger Test
4,
38AL23.

94
TABLE 38.
Vertical
distributions of artifacts from Auger Test
5,
38AL23.

95
xii
D
B

U
D
D
D
U
D
D
D

D
D
D
D
U
D
D
ACKNOWLEDGMENTS
The scope and value of the Allendale County chert quarry survey was
certainly extended beyond what could have been accomplished by Goodyear and
Charles - alone. 'rhe following people and agencies contributed greatly
toward the success of the project.
The survey and report were made possible bya Historic Preservation
survey and planning matching grant funded by the United States Department
of the Interior and the University of South Carolina. The grant was
administered by the South Carolina Department of Archives and History
through the office of Charles Lee, State Historic Preservation Officer. At
the South Carolina Department of Archives and History, Langdon Edmunds,
Susan McGahee, Suzanne Wylie, Nancy Brock, Mary Edmonds and Dr. Kenneth E.
Lewis all provided able assistance in preparing the archeological district
nomination and this report.
The Sandoz-Martin Works plant in Martin, S. C., and its employees,
were gracious and enthusiastic hosts for our research. Coincidentally,
most of the new quarries found were located on their property. At Sandoz,
Mike Anderson, Personnel Manager, and Charles Dunbar were most helpful in
every way. We also acknowle4ge the assistance of Sylvio Rodriguez,
Sandoz-Martin Works Manager, Howard Williams, Engineering Manager, John
Thompson, and Donna Youmans. Eleanor B. Owens, an amateur archeologist and
local expert, - was also most helpful in orienting us to the archeological
resources of the plant. The Sandoz company provided a backhoe and operator
for testing site 38AL135.
Many people in the Allendale County area facilitated our research.
These are Bert and Becky Kerwin of Barton, S.C., David Topper of Fairfax,
S.C., Dr. Larry Strong of the University of South Carolina-Salkahatchee,
Blanchard Poole, Manager of Butterfield plantation owned by Continental
Group, Inc., Richard Johnson, Manager of Wade plantation owned by W. S.
Morris, and E. R. Bostick of Estill, S.C.
The Savannah River Plant Archeological Program, under the direction of
Glen Hanson, is also to be thanked. Glen and his able crew spent a day
helping excavate 38AL135 and did an admirable job. Glen and Richard Brooks
have also shared with us their considerable knowledge of lithic raw mate-
rial types and sources from their experience tn surveying the extensive
Savannah River Plant property located immediately north of Allendale
County.
Testing sites 38AL23 and 38AL135 was made possible through volunteer
labor provided by the Steen family, Frank, Andee, Chris, Mathew and Van
from Kershaw, S.C., members of the Archeological Society of South Carolina.
Also prOViding assistance were Joe DaVis, Elizabeth Haynes, and Sammy Lee.
Our brief foray into Burke and Screven counties, Georgia, was aided by
George S. Lewis of Augusta, Georgia. George took us to several quarries
that he had previously mapped for the University of Georgia. At the Uni-
xiii
versi ty of Georgia, Dr. David Hally provided Georgia site numbers in a
timely fashion.
The staff of the Institute of Archeology and Anthropology is also to
be thanked. Dr. Robert L. Stephenson, Director, served as chief adminis-
trator of the Institute during the time of this project. Bob is to be
thanked for his support and encouragement during this time. and years
passed. Dr. Bruce E. Rippeteau, succeeding Director,. is also thanked for
his interest in the survey. Stanley South provided valuable consultation
on certain enigmatic hafted bifaces recovered during this survey. James L.
Michie also provided consultation on lithic technology and gave his
opinions about chert sources in South Carolina. Glen Hanson also helped
identify some of the ceramics. Alan Albright and Ralph Wilbanks confirmed
the chert source at 38AL135 by diving in Smiths Lake. Their efforts are
greatly appreciated. Dorothy Alford kept up with the budget, allowing us
the greatest use of our grant. Laura McGuire oversaw the cleaning and
cataloging of the artifacts. Mary Joyce Burns, Kenn Pinson, Rosalyn
Randall, Gordon Brown and Darby Erdare all to be thanked for their quality
production of the manuscript. Gordon provided excellent photography and
Darby did the line drawings. Mary Joyce typed the manuscript tirelessly
and accurately. Thanks to Azalee Swindle and Pat Busbee who typed the
rough draft of the manuscript. To Kenn and Rosalyn went the task of
assembling and editing the report and keeping up with the rest of us. To
all a sincere thanks.
xiv
CHAPTER I
THE ALLENDALE COUNTY CHERT QUARRY SURVEY:
HISTORICAL BACKGROUND, GOALS, AND METHODS
Introduction
The archeological study of prehistoric stone tools has proven to be a
valuable area of research to anthropologists. In many areas of the world
and for certain periods of time, stone artifacts constitute the only sur-
Viving record from which to understand man's past. Stone tools formed by
flaking are a subset of a much broader technological inventory used by
prehistoric people. The physical properties of stone that can be flaked in
a controlled manner are becoming increasingly better understood (Crabtree
1972), as well as the uses to which these tools were put (Semenov 1964;
Keeley 1980). The study of chipped stone technologies offers a Wide
variety of culturally interesting sources of information including how such
tools were made, the uses of the tools, the place of the lithic technology
in the overall cultural organization, the tracing of mobility patterns by
raw material source exchange and many other important subjects.
For most areas of the eastern United States, the only surviving
records of Paleo-Indian and Archaic groups is that of their lithic tech-
nologies. In the majority of cases all that we shall ever know will be
based on interpretations of chipped stone assemblages and the geoarcheo-
logical contexts within which they are found. Much important work, how-
ever, has been done in this seemingly constrained situation in terms of
studying raw material types and sources, raw material selection patterns
through time, reduction trajectories for both manufacture (Johnson 1979)
and maintenance and recycling (Goodyear 1974), and studying human adapta-
tion through technological organization (Binford 1979; Goodyear 1979;
Claggett and Cable 1982; Sassaman 1983).
It has become obvious to many investigators that a significant factor
operating to produce patterning and variability in Iithic assemblages is
related to the type of raw material and its availability. The of
a raw material to be flaked by percussion and pressure techniques is known
to be directly related to the physical properties of the material. Fine-
grained cryptocrystalline lend themselves to the production of well-made
chipped. stone tools. Such material provides flexibility all the way
through the reduction trajectory from preform production through multiple
phases of maintenance and recycling (Goodyear 1979).
It is well known throughout North America that Paleo-Indian groups
relied heavily on the better grade fine-grained flints and cherts for the
production of their bifacial and unifacial tools (Wormington 1957; Wilmsen
1970; Gardner 1977). Some of these types of flint and chert are highly
distinctive and have been named. The Alibates flint of west Texas, which
was so heavily used by Paleo-Indians over the southern plains, and Onondaga
flint, originating in western New York and spreading over the Northeast,
are good illustrations of this pattern. In the lower southeast, the Fort
Payne and various Ridge and Valley cherts are well known for their exploi-
tation by Appalachian Paleo-Indian and Archaic groups.
One major area and subject of the Southeast that are not well known,
however, are the Paleo-Indian and Archaic occupations of the southeastern
Coastal Plain extending from South Ca:rolina through Georgia; Florida and
.Alabama. Except for Florida, there have been only limited surveys of the
Coastal Plain and even fewer excavations of key sites that would enable
recogni tion of Paleo-Indian and Archaic adaptations of this zone. Know--
ledge to date consists of a few point types, often recognized as dating to .
the Late Pleistocene and Early Holocene because of work done in adjacent
states.
The raw material of the Coastal Plain assemblages is typically that of
marine cherts (Caldwell 1958: 10). Although the Coastal Plain is practi-
cally a stoneless environment, consisting of unconsolidated marine sedi-
ments, in certain places well-formed cherts outcrop. Other lithic raw
materials, quartz and metavolcanics, may be also present but limited to
major river valleys which drain the mountains and Piedmont. Regardless of
these non-chert resources, Coastal Plain chert dominates lithic assemblages
on the Coastal Plain of eastern Alabama, Georgia and South Carolina (Kelly
1954: 14; Charles 1981: 49).
The cherts of the southeastern Coastal Plain are tertiary in age and
consist of two basic forms created by two processes of silicification
(Upchurch, Strom and Nuckels 1981: 38-40). One type of chert is opaline in
makeup and formed in marine sediments by the breakdown of diatoms which
yield silica to form chert within the same beds. These opaline cherts are
chemically unstable, tend to be brittle and weak when flaked, and tend
toward cortication. They were not as often used by the Indians because of
their weaker condition, although they were utilized. Examples in South
Carolina are the silicified siltstones of the Hawthorn formation found in
Beaufort County and the highly fossiliferous cherts found in Calhoun and
Sumter counties which are silicified coquina (Upchurch, Appendix A).
More common and technologically valuable cherts are those made from
replaced limestone. They are formed by the groundwater transport of silica
originally formed in diatoms. These cherts are chemically and physically
stronger because the silica was in the form of quartz during replacement of
the host rock rather than opal. Furthermore, the host rock being replaced
is limestone which provides a tougher, more homogeneous. fabric. The fine
cherts of the Flint River Formation in Georgia and South Carolina are exam-
ples of this process, as are the limestone-replaced cherts of Florida
(Upchurch, Strom and Nuckels 1981).
Some of the earlier investigators of Georgia recognized the importance
of chert to the prehistoric Indians who left extensive lithic sites along
the Flint, Chattahoochee, and Savannah rivers. C. C. Jones (1880) was
impressed by the extensive workshops along the Savannah near the Fall Line,
and C. B. Moore visited the outcrop and quarry of Stony Bluff (9BUR6) along
the Savannah River (Moore 1898). Earlier than this, the traveling natural-
ist John Bartram visited the Vicinity of Brier Creek, Georgia, and observed
" prodigious numbers of flints, ... and many great rocks of good flint"
2
o
o
o
o
Ill ...... '.;. ..
II
Q
D
H
~
I
U
n...
I
U
U
U
D
D
D
o
(Bartram 1942: 24). Bartram's traveling o m p n ~ o n also seems to have noted
that contemporary forest fires had thermally altered chert artifacts lying
at the ground surface causing them to have a "smooth polish."
A. R. Kelly, beginning in the 1930s, initiated investigations into
. most of the fundamental research problems concerning Coastal Plain chert
assemblages that are being pursued today (Kelly 1938: 4-8). As a result of
his excavations at Macon Plateau near the Fall . Line of Georgia, Kelly
detected a preceramic "flint industry" that manifested a number of fasci-
nating traits. First, he correctly noted the European Paleolithic charac-
ter of the chipped stone artifacts, especially the formalized unifaces
often made on blade-like flakes (Kelly 1938: Fig. 4). He ascribed this to
an early hunter period, which we know today as Paleo-Indian and Early
Archaic. Further, he noted the strong tendency for the artifacts to be
highly weathered or decomposed. He observed that the degree of decomposi-
tion was seldom matched by other American collections. He attributed this
to great antiquity observing that chert specimens of less than 1 ,000 years
do not evince this weathering (Kelly 1938: 4). Kelly also correctly stated
that the weathered condi tionof the chert was not patination, as it is
usually referred to in the current literature of southeastern archeology,
but a process of cortication (Kelly 1938: 4; cf. Upchurch, Strom,'and
Nuckels 1981: 76-77; Appendix A). This patterning in degree of weathez.ing
by time stimulated work on developing geological means of developing
chronological techniques for archeology (Kelly and Hurst 1956). Kellywas
also interested in the study of chert quarries and adjacent sitesrelated
to the processing and use of chert (Kelly 1950, 1954). He realized the
distinctiveness of cores and implements found near sources, particularly
their large sizes. Finally, he (1954: 14) also realized the greatinflu-
ence of the type of lithic raw material available to prehistoric Indians in
determining variability in assemblages between areas.
The cherts of the Coastal Plain are quite amenable to thermal altera-
tion to improve their flaking capacity. The first modern attempt to
vigorously verify the effect of heat treating on chert or flint was done by
Barbara A. Purdy (Purdy and Brooks 1971) using Florida cherts. Since that
time David G. Anderson (1979a) has thermally altered chert- taken from jthe
Rice Quarry in Allendale County, South Carolina, and the Theriault site in
Burke County, Georgia, noting color changes and increased luster.
One of the great needs to facilitate settlement analyses of southeast-
ern Coastal Plain sites is that of chert source identification or prove-
nance. For purposes of studying mobility patterns and exchange, tools and
debi tage must be traced back to locations in which they originated. In a
preliminary study of the major chert types in Georgia, including the Ridge
and Valley and Coastal Plain provinces, Goad (1979) submitted samples for
neutron activation analysis. Statistical evaluation by discriminant
analysis revealed that 87.5% of the samples could be correctly classified
into seven groups. In the seven group solution, the cherts or flints from
the Ridge and Valley province were completely separate from groups formed
by Coastal Plain samples (Goad 1979: Table 14). Furthermore, the Coastal
Plain samples were separated into three groups by geography: a southwest
group from Sumter, Colquitt and Coffee counties; a north-central Coastal
Plain group from Laurens County; and an eastern group from Burke County
(Goad 1979: Table 14). This. work suggests some potential for chemicaXly
3
differentiating quarry sources on the Coastal Plain of Georgia at the
interregional level.
The most thorough description and source analysis has been done for
the state of Florida (Upchurch, Strom and Nuckels 1981). At the time of
their study, 175 locations where chert is known to occur were available for
study. They used fossils, mineralogy (X-ray diffraction), fabric, and
chemical composition (X-ray florescence) of minor elements to develop cri-
teria for the recognition of what they call "quarry clusters." A total of
19 quarry clusters were defined for the state of Florida. A quarry cluster
is " an area known to contain numerous exposures of chert, some of which ,
must have been used by early man, and in which the chert is expected to be
relatively uniform in fabric, composition, and fossil content. Quarry
clusters are usually wi thin the exposure belt of a single formation or
facies" (Upchurch, Strom and Nuckels 1981: 9). At this level of analysis
it is usually not possible to assign an artifact to a particular quarry;
however, depending on the variation wi thin the host rock of a quarry
cluster and the size of the formation or facies, it is often possible to
assign provenance within a few miles. Two raw material source studies have
been done on Paleo-Indian artifacts from the 'Tampa Bay region (Goodyear et
al. 1983; Daniel and Wisenbaker 1984) using the quarry clusters defined by
Upchurch (Upchurch, Strom, and Nuckels 1981).
The geographic distribution of the Coastal Plain cherts can be
generally described as beginning, around Tampa Bay, Florida, proceeding up
the western half of the Florida peninsula including the panhandle, north-
ward iIlto extreme southern Alabama, then northeastward on a diagonal
following the upper Coastal Plain of Georgia. It is significant that the
extreme northern tip of this chert distribution barely extends east of the
Savannah River into South Carolina. It is of relevance to archeologists,
not to mention the prehistoric inhabitants of South Carolina, that this
high quality chert was only available to any significant degree in extreme
western Allendale County. Thus, as will be seen as a result of the follow-
ing study, we are in a fortunate position analytically to delineate a rela-
, tively small area in South Carolina from which much of the chert utilized
prehistorically was no doubt obtained.
The Allendale Chert of South Carolina
Archeologists working in South Carolina have recognized a distinctive
light colored fossiliferous Coastal Plain chert that wa.s commonly used
throughout prehistory for the manufacture of chipped stone tools. This
lithic material has been known for some time by artifact collectors and
archeologists alike to have originated somewhere in the southern part of
the state because of the abundance of chert artifacts in that area (Michie
1977: 16-18; Anderson, Lee, and Parler 1979: 10; Charles 1981: 49; Anderson
et a1. 1982: 126). Based on Cook's (1936) description of the Oligocene in
South Carolina, it has generally been thought that this chert is derived
from the Flint River Formation.
4
o
U
I.... .... ....
II
I
D
U
D
o
D
D
D
H
U
D

o
Until recently, the only source and quarry related to this chert known
in South Carolina by archeologists was the Rice Quarry, 38AL14, located in
Allendale County, near the Savannah River on King and Watch Call creeks
(Fig. 1) (Stoltman 1974: 175). Early in 1971, Dr. E. Thomas Hemmings of
the Institute staff nominated this 20-acre quarry to the National Register
of Historic Places under the name "Red Bluff Flint Quarries." Later in
that year, James L.Michie, Sammy T. Lee, and A. Robert Parler, members of
the Archeological Society of South Carolina,coriducted test excavations in
the extreme southern end of the quarry, south of King Creek. Their work
revealed a substantial utilization of the chert by Late Pleistocene-Early
Holocene Indians. The artifacts and notes (Michie 1984) from their excava-'
tion have been placed with the Institute and the results of their testing
are briefly reported in this report.
Since the beginning in the 1970s, research on Coastal Plain chert has
been characterized by two basic questions. First, is all the better grade,
sparsely fossiliferous chert known in the literature as "Allendale" one
lithic material, petrologically speaking, or is there more than one dis-
tinct chert represented? Second, is the so-called Allendale chert Cor
cherts) from one basic source area on the Coastal Plain, or can other
sources be found north, east and south of Allendale County?
Wi th the exception of some preliminary petrologic work initiated by
David G. Anderson and his associates, identification of Allendale chert has
proceeded on a macroscopic basis. Four characteristics of the chert have
formed the criteria used to identify Allendale chert: (1) homogeneity with
few fossils to disrupt isotropy; (2) earth. colors ranging from dark brown,
red, orange, yellow, tan and white, often translucent; (3) thermal altera-
tion which changes the colora to pink, red and blue and increases the
vitreousness resulting in a glossy appearance (Anderson 1979a); and (4) a
strong tendency to weather, causing loss of silica and color, and resulting
in the forming of cortex yielding light tan, yellow and even white colors.
This tendency toward weathering can be extreme among older artifacts such
as Clovis, Suwannee and Dalton points (Michie 1977: 19), resulting in arti-
facts that are light in weight and color, taking on the appearance of cor-
tex.
As part of the lithic analysis of the Southeastern Columbia Beltway
excavations, Anderson" (1979b) submitted five samples of chert derived from
three different Coastal Plain sources to Dr. Gerald R. Baum. Two flakes,
one natural and one thermally altered, were analyzed from the Rice Quarry.
A third flake of a fossiliferous chert from site 38CR33 was analyzed which
was taken from an outcrop near Lake Marion. The fourth was taken from the
fossiliferous chert outcrop at Buyck's Bluff (38CL17) in Calhoun County on
the Congaree River. The fifth chert sample was a nodule of fossiliferous
chert from an outcrop on the Wateree River in Sumter County. Baum (1979:
33) reported that in spite of the known different origins of each, all five
were approximately similar in terms of macro and microscopic composition.
In a similar study of the Mattassee Lake sitea in Berkeley County,
South Carolina, Baum analyzed two chert flakes from site 38BK226 which were
classified by archeologists as "Allendale chert." Baum reported that both
samples agreed with previous descriptions of Allendale chert as determined
in the Southeastern Beltway study (Anderson et ale 1982: 126). Baum was
5
tJuillmarnlina !.an11.tt'tinns uruty ..
Percentage of Coastal Plain Chert
RED BLUFF
FLINT QUARRIES--"'"
38AL/4
LEGEND
I
ss
. isi S6
:=i
.=.
...... 3
.... 0
Figure 1. Location of Red Bluff Flint Quarries, or Rice Quarry
(38AL14)in Allendale County, South Carolina. Also shown is
the geographic distribution of Allendale chert artifacts in
private collections.
6
not told that an extra-local source was possible among the samples. For
one of the Allendale chert specimens it was reported that the source area
was unknown, for the other "possibly Santee Limestone Formation?" (Anderson
et al. 1982: 126)
In the present study, a chert sample from 38CL17, Buyck's Bluff, and a
sample from 38SU42, a quarry north of Rimini in Sumter Collnty, were sub-
mitted to Dr. Sam B. Upchurch for thin section analysis. These cherts are
known to archeologists to be highly fossiliferous. One is so distinctive
that it has been named "Buyck's chert" for the outcrop present at Buyck's
Bluff. Upchurch (Appendix A) easily differentiates these two samples from'
those' of the Flint River Formation of Allendale County and states they are
derived from the Black Mingo Formation. The material from the Black Mingo
is classified as "opaline coquina," whereas the Flint River cherts are
classified as "silicified grainstones" (Appendix A). As will be seen in
this report, based on the geological analysis of Upchurch, the highly
fossiliferous cherts of the Black Mingo Formation with mollusc fabrics are
distinct from the Allendale cherts of the Flint River Formation. Further,
the cherts that have been called "Allendale" by South Carolina archeolo-
gists are derived from the Flint River Formation and are in nearly every
case examples of silicified grainstone.
The question of geographic distributions of the sources of Allendale
chert was also one addressed by the present survey. As mentioned, the Rice
Quarry (site 38AL14),' until rece'ntly, had been the only known outcrop and
quarry in South Carolina. This quarry and those immediately across the
Savannah River in eastern Georgia .. were thought to be the primary sources
for chert artifacts found in South Carolina (e.g., Anderson, Lee and Parler
1979: 10-12; Goodyear, House and Ackerly 1979: 199). This was based on the
historic fact that no other chert quarries had been found in the South
Carolina Coastal Plain that produced the Allendale chert, although no sys-
tematic surveys had been done to probe this issue.
In 1979, Tommy Charles of the Institute staff began traveling the
state, meeting private artifact collectors and examining their collections.
In the process of developing a network of collectors, he made the acquain-
tance of Mr. David Topper of Fairfax, South Carolina. Topper told Charles
that more quarries existed to the south of Martin, South Carolina. In June
1981, he took Charles and Goodyear to two chert outcrops wi th evidence of
aboriginal utilization. These were on the Sandoz-Martin Works property,
just north of Johnsons Landing. Upon realizing that more quarries existed
in Allendale County,plans were laid to do a comprehensive survey of the
western edge of the county near the Savannah River. In 1983, a grant pro-
posal to survey western Allendale County to search for chert quarries was
submitted to the South Carolina Department of Archives and History. This
proposal was accepted and a Survey and Planning matching grant was awarded
to conduct this survey and to nominate appropriate quarries to the National
Register of Historic Places. At the conclusion of this study, 14' quarries
and sitea related to quarries were nominated together as an Archeological
District to the National Register.
7
Purpose of This Report
The purpose of this report is to present the findings of the quarry
survey which was conducted during the late summer of 1983 and the winter of _
1984 by these writers. Described herein are the methods employed and the
locations of where survey activity was conducted in Allendale County and
. the methods which were used to evaluate sites. Archeological and petro-
logical data gathered from each site are presented as site descriptions.
As a" part' of data documentation, the report submitted by Dr. Sam B.
Upchurch, complete with microphotographs, is presented as Appendix A.
Although the discovery of additional quarries is possible, especially, in
the Savannah River, this report presents a fundamental inventory of
quarries for South Carolina and petrologic knowledge for what is called
Allendale chert, and it also forms a baseline from which to derive much
important work and interpretations.
Survey Objectives and Methods
The survey objectives were to discover chert outcrops and quarries in
western Allendale County and gather data sufficient to evaluate their
eligibility for the National Register. As part of the quarry evaluation,
information on site size, depth, degree of utilization, occupational
history and type of raw material was gathered. Unaltered chert pieces were
also collected from each outcrop for geological analysis. The objective of
the petrologic study was to analyze the variability contained within those
cherts called n Allendale" by archeologists to see if it is all one kind of
material or if significant varieties can be discerned that would warrant
the creation of new types. To help differentiate the Allendale chert from
other often similar fossiliferous Coastal Plain cherts, additional cherts
and silicates from South Carolina and Georgia were submitted for petrologic
analysis (Appendix A).
The survey strategy can best be described as opportunistic. The sur-
vey was basically conducted by Tommy Charles in a truck and by boat. Sites
were searched for in dirtroads, firebreaks, creeks and other types of
exposed or eroded surfaces. A total of 18 days was spent searching western
Allendale County (Fig. 2a-c). The hatched areas (Fig. 2a-c) indicate roads
and fields that were inspected.
The survey for quarries was directed to a large extent by our previous
knOWledge of quarry locations. The Rice Quarry, 38AL14, for example, is
located on King and Watch Call creeks in a floodplain situation close to
the Savannah River. Reports of chert boulders being observed in the Savan-
nah River by a University of South Carolina geologist and geology graduate
students, plus the two quarries shown to us by Topper on the Sando'z-Martin
Works land near Smiths Lake Creek, all pointed to situations where chert
was exposed by flowing water. Thus, the Savannah River and streams and
swamps draining into the Savannah were all likely places where Indians
could have found chert exposed.
Given that at least two quarries had been previously observed on the
Sandoz-Martin Works land, this property was earmarked for intensive vehicle
and pedestrian survey (Fig. 2b). This was a wise decision, for nine quar-
8
o
D
I
D
I
n.l;
II
D
I
D
D
I
III
I
11
II
I
D
I...
I
D
o
ries were located on or near the plant property which had not been pre-
viously recorded. The Sandoz-Martin Works plant and adjacent lands (Fig.
2b) were thoroughly driven over by truck and walked, including all exposed
ground such as dirtroads, ditches, creeks and firebreaks. The company
graciously provided us with excellent topographic maps drawn in two-foot
contour intervals with a scale of 1 inch to 500 feet. Plant employees who
had hunted, fished, and collected artifacts on the Sandoz property also
told us of places where they had seen chert exposed.
The Butterfield plantation, a four thousand-acre tract, was also
intensively surveyed by truck and on foot. The Butterfield plantation
(Fig. 2b, c) is located along U.S. 301 where it crosses the Savannah River.
It is split up by U.S. 301 and County Road 102 and by South Carolina state
roads which run north and south. The plantation, used primarily for pri-
vate bird hunting, is laced with dirtroads, firebreaks, and small plots
cultivated for bird food. No quarries were found on this land.
The Savannah River (Fig. 2a-c) from CohenI s Bluff in lower Allendale
County to one mile north of Steele Creek in Barnwell County, South Caro-
lina, was surveyed by boat. This took two days. All exposed banks were
checked as well as old cutoffs of the river and creeks that could benavi-
gated by boat. Three outcrops of chert bearing evidence ofaborigimil
exploi tation were discovered. Two of these, Fastwater Outcrop (38AL151)
and Johnsons Landing (38AL141) weFe found on the South Carolina side of the
Savannah River, while a third un"named and unnumbered source bearing evi-
dence of use was found on the Georgia side near the mouth ofSmiths Lake.
The boat survey was done in early August 1983 when the water level in the
Savannah River was high. It is likely that more chert outcrops
ries are present within the river. .
Lower Allendale County was also surveyed, concentrating efforts along
South Carolina State Highway 3 and adjacent areas (Fig. 2c). This was done
by driving roads and walking exposed ground where available. Much of the
land along the Savannah River in lower Allendale County is contained within
large plantations or timber companies and is posted or has locked gates.
This severely limited efforts to survey the southern portion of the county.
Numero\lS dirtroads and firebreaks that intersect State, Highway 3 were_
driven and walked (Fig. 2c). This was done all the way to Hampton County
line. Several prehistoric sites were encountered which were rich in lithic
remains, but no chert boulders or outcrops were seen.
Informants who were familiar with the Savannah River and Allendale
County area were also profitably utilized in the survey. Mr. Charles Dun-
bar, a resident of the Martin community and employee of the Sandoz-Martin
Works company, shared his knowledge of where chert and other rocks were
known to outcrop. This resulted in locating two silicified sandstone out-
crops north of Martin. Mr. and Mrs. Bert Kerwin also showed us a deposit
of chert which had been dragged up by a dragline behind their home on the
Coosawhatchie River near Barton, South Carolina.
Upon finding a quarry or site near a quarry, some idea of lateral and
vertical extent was obtained. Usually enough ground was exposed to esti-
mate the size of the site, although in some cases the vegetation was so
dense it was not determined. Testing of site depth in an effort to dis-
9
Figure 2a. Map showing location of Chert Quarry survey.
indicate land and roads surveyed.
10
Hatched areas
I
L
L
r
I
L
Figure 2b. Map showing location of Chert Quarry survey.
11
II
Hatched areas
Figure 2c. Map showing location of Chert Quarry survey.
12
Hatched areas
o
D
D
D
o
o
cover stratified remains was a high priority. This was done by using an
8-inch bucket auger which could excavate in 15-cm levels with little con=
tamination from the walls of the hole. All sites were tested in 15-cm
levels except where specified, and all sediments were sifted through
quarter-inch screen. Test pits were excavated at some sites and these are
discussed by site. In most sites, auger testing was purposefully not done
in the dense, rocky areas. Testing in these locations would only yield
large quantities of natural rock and undiagnostic debris . Places for test-
ing' were chosen near the outcrop where site depth might be expected because
of colluvial and alluvial deposition. These areas often produced evidence
of core reduction, tool manufacture and tool use, which allows temporal
differentiation by depth.
Artifact Analysis
Ceramic and lithic analyses follow conventional approaches and
descriptions for Coastal Plain assemblages (Stoltman 1974; Anderson, Lee
and Parler 1979; Hanson, Brooks, and White 1981). The bulk of material
recovered was chert debitage, cores and tools. Stage concepts of core and
biface reduction are the basis for the flake and core categories (House and
Wogaman 1978; Goodyear, House and Ackerly 1979). These publications maybe
consulted for attribute definitions and justifications.
Some additional comments are' necessary to explain the 1ithic typolo-
gies. Primary decortication flakes have their dorsal surfaces fully cov-
ered with cortex; secondary decortication flakes simply have the presence
of cortex regardless of how much or little; tertiary flakes are morphologi-
cally indistinct, except that they have no cortex and they are not biface
thinning flakes or other distinct flake types; the defini tion of biface
thinning flakes follows that set forth by House and Wogaman (1978: 59-60).
The biface thinning flake is the most common flake fonD. and it is assumed
that, although some core trimming flakes may be indistinguishable from
biface thinning flakes, most are truly from bifaces. The majority of cores
recovered are bifacial in form, further justifying this assumption. The
most common flake type on the Savannah River Plant located a few miles
north of Martin is the biface thinning flake (cf. Hanson, Brooks ,and Wnite
1981). In the debi tage analysis, no distinction was made as to whole or
broken flakes.
One difficulty encountered regarding thermal alteration concerned the
accurate recognition of what was not heat treated. Many later stage reduc-
tion flakes and bifaces exhibited glossy surfaces and color changes, s ~
cially to reds and pinks so characteristic of the thermally altered Allen-
dale chert (Anderson 1979a). It .was not possible, however, to exclude some
pieces from being thermally altered, for some flakes exhibited color
changes and glossiness on one end but not the other. Thus, except for
bifaces, points and cores, no attempt was made to record systematically
thermal alteration. More experimental work needs to be done to develop
better criteria. Also, some of the upland quarries exhibited large ther-
mally altered boulders which have been exposed at the ground surface due.to
forest fires. It is not possible to rule out that some of the heat treated
chert seen on the quar:ries is a result of natural causes ( cf. Gregg and
Grybush 1976).
13
14
CHAPTER II
SITE DESCRIPTIONS
38AL140
This site is located on a severely eroded hilltop on the south side of
Smiths Lake Creek immediately north of old county road 102 (Fig. 3). S.C.
102 in fact cuts through the western edge of the site. All topsoil has
been removed by erosion, leaving great quanti ties of chert exposed (Fig.
4) The chert here is quite fossiliferous and of poor chipping quality
compared to the usual Flint River chert. Some of the chert, however, is of
average quality and it may be this grade that was sought by prehistoric
quarrymen. Compared to other quarries in the area, this one received less
use. Excavations here would not be possible because of erosion.
A chert sample was taken for petrological analysis. Upchurch (Appen-
dix A) classes' this material wi thin Category 7, silicified grainstone. He
further subdivides this sample into Category 7c, a ttspotted silicified"
grainstone. The sample from 38AL140 is the only "spotted" grainstone he
encountered. Depending on the geographic content and physical variability
of this material and the other, subcategories with Category 7 (Upchurch,
Appendix A), it may be possible' to discern specific sources wi thin the
Allendale Chert Quarry Cluster. The chert from site 38AL140 is an expres-
sion of the Allendale chert.
38AL136
This site was auger tested in two places along a terrace which exists
on the south side of Smiths Lake Creek at the foot of a hill. The :terrace
is relatively expansive here and was formed by colluvial action at the base
of the uplands (Fig. 3). Chert boulders, cores and flakes were observed in
dense concentrations in the dirtroad which runs above the terrace (Fig. 5).
Because of this evidence of quarry activity, it was decided to auger test
on the terrace where sediments would be protected by forest. The terrace
drops off fairly rapidly overlooking the Smiths Lake Creek floodplain,
forming a good situation for a site with chert raw material available on
the hillside and good access to the creek bottom. The terrace was auger
tested on the extreme southwestern end and the northeastern end where the
terrace widens.
Auger tests 1-4 were excavated on the southwestern portion where the
terrace narrows near the road (Fig. 3). A section of pipe was driven into
the ground for a surveying datum, located about 1 m off the dirtroad
toward the swamp. From this datum point an angle and distance were deter-
mined back to D.o.E. permanent marker (C34) using a transit and stadia rod.
A large beech tree, 1.8 m in circumference at breast height, is located in
front of the datum. The only' other large beech tree on the terrace is
located to the east where auger tests 5-7 were dug (Fig. 3).
15
o 500 1000 FEET
METERS
Figure 3. Locator map for sites 38AL140, 38AL136, 38AL138, 38AL152, and
38AL137.
Figure 4. Chert boulders and debris on hillside of 38AL140.
Figure 5. Chert boulders and debris on hillside of 38AL136,
the Beech Tree Site.
17
Durin,g the few days prior to excavating auger tests 1-4, there had
been considerable rain causing the soil to be saturated down to clay sub-
soil. It was not possible to recover the sandy sediments in the bucket
auger below about 60 cm because the sand would melt back into the water
which was standing in the bottom. of the hole. In auger test 1, a grey
sandy sterile clay was contacted at about 60 cm. In auger tests 2 and 3
this clay was not encountered, although it likely exists immediately below
this level. _On the southwestern end of the s1te-it is unlikely that
artifact-bearing sediments exist below 60 cm. Auger test 4 was dug at the
edge of the swamp and was abandoned because of highly organic soil and
water. No temporally diagnostic artifacts of any sort were found in these
three auger holes to give an indication of time depth or cultural periods
(Tables 1-3).
Auger tests 5, 6, and 7 were excavated at a later date in the project.
These tests were placed around another large beech tree. This part of the
terrace is also in woods but has a larger area of flat ground suitable for
human occupation (Fig. 3).
The suspected colluvial or1g1n of sediments overlying lithic debris on
the terrace was well confirmed in these tests. Artifacts were covered from
30 to 60 cm with sterile sand which had buried A soil horizons. In auger
test 6, two separate dark A horizons were auger tested hitting flakes at 60
cm. The sterile sand zone is deeper going up slope. Interestingly, the
fact that no artifacts were found in this overlying sand indicates that
sand, but not chert, was transported and the lithic remains on the terrace
are in situ. The sterile colluvial overburden probably relates to historic
farming activity on the adjacent hills.
Auger tests 5 through 7 reveal a lithic stratum approximately 60 cm in
thickness (Tables 4-6). There are some brightly colored heat treated and
presumably later flakes in the upper portion of the holes, but the over-
whelming majority of flakes, especially from the bottom 30 cm, are very
yellow and weathered. The latter are suggested to be of some antiquity,
presumably preceramic. The crude bifaces (Tables 4, 5) also came from the
lower levels. The lithic remains at 38AL136 are indicative of raw material
reduction close to the source. The presence of cortical material, split
chunks, and tertiary Iittle of which looks thermally al tered, is
expectable given the existence of an outcrop on the adjacent
hillslope. Very little river cortex was seen in the debris here, a fact
reinforcing the interpretation that aboriginals were exploiting a hillside
terrestrial source. This site offers a good opportunity to study raw
material procurement and utilization of terrestrial chert outcrops.
Upchurch (Appendix A) has classified the chert as Category 7b,
"banded, silicified grainstone." This is typical Allendale chert, and is a
member of the Allendale Chert Quarry Cluster.
38AL138
This site was initially discovered in a dirtroad which runs along the
northern rim of a high hill overlooking Smiths Lake Creek Valley (Fig. 3).
This hill produces natural exposures of chert wherever it is deeply eroded.
18
c::=J c::::J 11:::1 a:::J i,Lfi,J r::::J E:::I 'dCL, i r::::I i, . . 1 r::=:l C:J
TABLE 1
en
en en QJ
QJ
()
ClI en
ClI
en r-I en QJ
r-I
QJ J".r.l
en J".r.l 01-1 .
ClI QJ
QJ
:3: '0 .
I-l ClI .
ClI r-I ()
r-I I-l r-I QJ I-l 0 r-I s:: r-I J".r.l r-I ClI ,.Q QJ J".r.l
0 ()
J".r.l oM J".r.l 0 4-4
r-I () () QJ ..c: '0 0 'r!
J".r.l ClI QJ A
E-t '0 QJ
E-t P=l p... 01-1
I-l A QJ ..c: QJ . ,
e
:>. z :3: QJ U
.
ClI QJ N () ()
QJ '0 I-l N
CO .
s:: 'r! ()
'r! ::s ClI r-I QJ 0
QJ 01-1 r-I r-I ClI QJ

0 01-1 ClI r-I 0 4-4 QJ ,Q 01-1 4-4 01-1 I-l ClI ClI I-l I-l ()
I-l 4-4 'r! 01-1 .r!
I-l ,Q 4-4 QJ 01-1
.
01-1 01-1 QJ oM ..c: I-l QJ QJ 'r! 01-1 QJ s:: 0 0 ClI I-l r-I 0 0 0
J".r.l U p...
til E-t P=l P P u u l:J:l p... j:tl p...
0' E-t E-t
0-15cm
15-30cm
1a
3
4 11 .9 3.0 ....
\0
30-45cm 1 1 1 14
17 7.1 .42
45-60cm
4b 1 2
54
61 57.3 .94
60-75cm
1 7 15
23 24.3 1.06
75-90cm
90-105cm
105-120cm
120-135cm
Key - a. MetavOlcanic (?)
b. One limestone (?) cortex
:;';;::5:;'::" .,' ,..... Ji!L,. .' .__ .a.. .." _ ,,,.;&1&b,L,.. 5iJ!#",'#. _. ".&: ..-., "". H J.SUSL, ".sanz,_""" ,,_,.,# J$'"... ". .. ",".A,Jiilla .\$4. 4. ".,2-.-,&1.,$$1$." , . ",t4,.1"J . !Dh", ML __ .,..1 ' "",,,,,,_ ,a:pV".-",
oz
x
..... .....
1.0 -:J 0'1 -!'>- \.>l
.....
0
(l) I\;)
0 0 VI 0 VI 0 VI I
'=< 0 VI I I I I I I
.....
I I
.....
1.0 -:J 0'1 -!'>- . \.>l VI
I ..... .....
0 0 VI 0 VI 0 Q
\.>l I\;)
VI Q Q Q Q Q s.
P, VI 0 Q S S S S S
.
Q Q
S
S S
t:-I
1-'-
S
(l)
to
c+
0
::s
(l)
>
"""
Fire Cracked Rock

-."
1-3
'-' H
.....
Chunks
I:j
Q
P, >
(") 0
1-3 ti
c+
t::t
(l)
><
Prim. Decor.
H
t/)
1-3

Secon. Decor.
H
to
c:::
1-3
H
0
z
Tertiary Flakes
t/)
0'1
.....
-!'>-
.....
I Biface Thin. Flakes

t::t
t?::l
'"d
Utilized Flakes
I:j

0
:s::
Retouched Flakes 1-3
> >
c:::
Q
Uniface
t?::l t?::l

I\;)
1-3
t?::l
Core
t/)
1-3
I\;)
..
Cobble Tool
\.>l
CD
>
Hafted Biface
t:-I
.....
\.>l
0'1
..
Preform
1-3
::t:
t?::l
to
Blank
t?::l
t?::l
(")
::t:
1-3
Pottery !:tl
t?::l
t?::l
Quartz Pebbles
t/)
H
1-3
t?::l
,....;J I\;) -!'>-
..... I Total Number Flakes
-!'>-
I Total Wt. Fl. Gms
.

. .
0 I\;)
1.0 \.>l
. . . .
I Average Fl.Wt. Gros VI
.....
I\;)
\.>l
-:J 0 I\;)
r::::::J t=::J E::i c:.:J 1,_1 I.;,j r:::::l c:::l I .... , I r::::J
TABLE 3
ARTIFACT DISTRIBUTIONS BY DEPTH FROM AUGER TEST 3, 38AL136, THE BEECH TREE SITE
CIl CIl CIl
Ql Ql
l5
8
() ell CIl ell
0 CIl r-l CIl Ql r-l
p:: Ql r:r.. Ql
CIl r:r.. +J
.
Ql Ql

"t:l
. I-l ell
.
ell r-l ()
r-l I-l r-l
Ql I-l 0 r-l s:: r-l r:r.. r-l ell .0 Ql r:r..
0
()
r:r..
OM
r:r.. 0 4-l
1
r-l
() () Ql .d "t:l 0 OM
r:r..
ell Ql s=l E-I "t:l Ql E-I j:Q
P-l +J
I-l s=l I-l Ql .d Ql
e
z Ql
(.) CIl .
ell Ql N () () Ql "t:l I-l N bO
.
s::
OM () OM
::s ell r-l Ql 0 Ql +J r-l .-i ell
Ql s::

0 +J ell r-l 0 4-l Ql .0 +J 4-l
+J I-l ell ell I-l

I-l ::s
() I-l 4-l OM +J OM
I-l .0 4-l Ql +J ell +J +J Ql
'M ..c: I-l Ql Ql OM +J Ql s:: 0 0 ell I-l r-l 0 ::s 0 0
t,) P-l
tI)
E-I j:Q P
p::
P
(.) (.)
::0 P-l P-l 0" E-I

0-15cm
5 5 2.8 .56
N
15-30cm 11 11 4.2 .38
I--'
30-45cm 6a 10 65 79 193.5 2.45
45-60cm 1b 4 36 1 40 13.2
33
60-75cm
75-90cm
90-105cm
105-120cm
120-135cm
Key - a. Two limestone (?) cortex
b. Limestone (?) cortex
.. _'5',H!'F.....;;' L "' .,.,_. ..'" ",,' '" ,. Lij.9dF",. ""',_V ."!r!':' _.6"",,. .. ", ... j!ji.r..,... , ".. .t,,..,UE ..,,. ,.. "iIf,,.':H'&i!'Ii":, 1 "JiSi'Lhir t;'iii''J'"'"''
Key - a. Highly weathered, early stage ovoid form
b. Angular chunk, weathered (137.1 gm)
c. Distal fragment, crude, highly weathered
r:=J c::I I!::J II:!.:] c::J E.] EJ r=J r:::::::J L"_"_l r::::::J r:::::J
TABLE 5
ARTIFACT DISTRIBUTIONS BY DEPTH FROM AUGER TEST 6, 38AL136, BEECH TREE SITE
Cf) Cf)
8
Q)
..
,.!l4 ,.!l4 Cf)

CJ (IS Cf)
8
Cf) r-I Cf) Q)
r-I
I'Ll
JJ
Cf)
I'Ll
.j.J
. Q) Q)

't:l . J.l
. (IS
fj
CJ r-l J.l r-I
Q) J.l 0 r-l l=l r-l r-I (IS ,.c Q)
I'Ll
,.!l4 0 CJ I'Ll
.r-l
I'Ll 0 4-l
r-I
CJ CJ
Q) ,d 't:l 0 .r-l
I'Ll
III Q)
A >- E-t 't:l Q)
E-t .j.J
J.l A J.l Q) ,d Q)
e
>- z
Q)
U
Cf) . ell Q) to:! CJ CJ Q)
't:l J.l to:! bO
.
l=l
.r-l CJ .r-l
::J ell r-I Q) 0 ,.!l4 Q) .j.J r-I r-I ell
Q)
.
0 .j.J (IS r-l 0 4-1 Q)
.0
.j.J 4-l
.j.J
J.l ell (IS
k
J.l CJ J.l 4-l .r-l .j.J ..-l J.l .0 4-l Q) .j.J
III
.j.J .j.J Q)
.r-l ,d J.l
Q) Q) .r-l .j.J Q)
l=l 0 0 (IS J.l. r-I 0 ::J 0 0
I'Ll U tI.l E-t D P:l D U u ::Il ct .E-t E-t

0-150m
15-30cm
30-45cm
N
w
45-60cm
60-90cm 5 3 9
52 - 2 1 71 52.4 .74
90-105cm 3 5 15 75
1a 2 98 135.8 1.38
105-120cm 3 2 19 29 53 104 1.96
120-135cm 1 1 2 21 1b 25 28.7 1.15
Key - a. Corner of weathered blank
b. Very weathered, crude square form
:,.;,,},,- &:;,(,,1,. '.c5.'o/,. G,"Wt .,R; 3. ,fL ,,$\' ..... ':+",5'!e,A .** ".. ,iL.. t,,,; -".-.-.;:: "
VZ
o
u
u
U
B
D
D
D

o
D
D
D
D
D
D
D
o
o
Such is the case at this site where on the northern side of the hill, chert
boulders up to a meter in diameter could be seen. This quarry is spatially
quite extensive, up to 300 meters east and west to judge from lithic debris
in the road, and about 100 meters north and south. On the wooded hillside,
large boulders and split chunks of chert bearing cortex, often reddened
from fire, could be seen. In some places flakes and lumps of chert were so
dense they could be heard underfoot under the leaves. It is likely that
chert boulders were naturally exposed from erosion on this hillside such
that quarrying by aboriginal groups was made possible. I No diagnostic arti-
facts were seen from the dirtroad which presented a considerable amount of
ground surface over a 300-meter strip. All varieties of colors are observ- '
able in the chert as well as the full range of weathering. Specimens of
chert were collected for petrologic analysis.
The si te was subsurface tested using a bucket auger in three places
(Fig. 3). Auger test 1 was placed on the extreme western end of the road.
Artifacts were continuously present in the upper 75 cm of the hole. The
upper 45 had three sandy soils. The first 15 cm was a brown coarse sand;
at 30 em this changed to a gold colored sand; between 30 and 45 cm a tan
reddish sand was present. The flakes from these upper zones were not
markedly weathered. At the level of 45 to 60 em a dark red coarse sand
appeared that was full of small quartz pebbles (Table 7). From 45 to 75 cm
this zone of red sand persisted, bearing extremely weathered chert flakes.
The association of highly weathered chert and this red sand has been noted
previously by Tommy Charles from sites in Allendale and Hampton counties.
Auger test 2 was located at about the middle of the site alongside the
dirtroad (Fig. 3). The stratigraphic picture here is virtually the same as
auger test 1 (Table 8). Red coarse sand began at about 20 cm and continued
to the bottom of the test. Again, the chert was highly weathered, nearly
white in some cases and very light in weight.
Auger test 3 was placed in the road on the eastern end of the site in
a location manifesting a high density of flakes on the surface. The soil
here was different from the previous two tests, being a lighter red or tan
in color. Flakes were not nearly as weathered either. The site depth here
was essentially 60 em (Table 9).
The hilltop upon which site 38AL138 is located seems to have a variety
of soil types present. The most interesting association is that of the
coarse bright red sand and the exceedingly weathered chert-. There maybe a
chemical basis for this weathering. This is an extremely large site with
considerable evidence of chert exploitation by aboriginal populations. No
ceramics or diagnostic lithics were found to aid in chronological place-
ment. It seems likely, however, that owing to its size and exposure it was
available for procurement throughout prehistory.
Upchurch (Appendix A) has classed a chert sample from here as Category
7a, silicified grainstone. It is Allendale chert and the quarry is a mem-
ber of the Allendale Chert Quarry Cluster.
25
9Z
1..0 -.3 0'1 .j::>. \.)J .....
0
I\)
0 0 \J'1 0 \J'1
0 \J'1 I
0 \J'1 I I I I I I
.....
I I
.....
1..0 -.3 0'1 .j::>. \.)J
\J'1
..... .....
0 0 \J'1 0 \J'1 0 (')
\.)J I\)
\J'1 (') (') (') (') (')
S
\J'1 0 (')
s s s s s
(') (')
s
s s
fiLl ;;:;b, 1 Wi 1$'.;1 a',',H&'i iii
C
\!4i:J lib",] Eiili\i i:;gil'J it. Iiii!ij El IZ'.i c::::J I.: I c:::;]
TABLE 8
ARTIFACT DISTRIBUTIONS BY DEPTH FROM AUGER TEST 2, 38AL138
to to (/)
III III

,.!ol ,.!il ,.!il to
U
CIS to CIS
0 to r-l (/) OJ .-l
rx: III III ,.!il to
.j,J
.

,.!ol III III III
t3:
'0 .
.

r-l u r-l .-l
JJ
0 r-l r-l CIS ,.0 III
0 0
:a
0
.g
.-l
u u III '0 0 '.-I

III III A E-l '0 III E-l
'"
p.. .j,J
A III ,.c::. III

e
z t3: III
U UI
.
CIS
Q) N 0 0 III '0 N bO
,.!ol .

.-1 0 'M
=='
CIS r-l Q) 0
III .j,J r-I .-l III

CIl

0
.j,J
III .-l 0 CIl
.j,J
.j,J

III
0 '.-1
.j,J or-f $-I III III
.j,J
III .j,J III
'0-1 ,.c::
III III '0-1
.j,J
g
0 CIS r-l 0
6-
0 0
rz. u p.. CI.l E-l P u u l3:I
p.. p..
E-l E-l
0-15cm 2 8 4 14 64.7 4.62
15-30cm 3 18 22 43 114.4 2.66
30-45cm 14 94 69 la 1 177 237.0 1.34
N
2
b
c
....
45-60cm 6 17 44 33 100 313.8 3.14
60-75cm 1 11 15 28 20.8 .74
75-90cm 2 2 3 3
10 8.0 .8
90-105cm 4 4
8 7.0 .87
105-120cm
120-135cm
Key - a. Heat treated, chunky fragment, 211.7 gm
b. Squarish, crude, weathered, 38.3 gm
c. Unidentified core 78.8 gm
,.y"A$i\.'jj,,,;;:&Wih ,,,.,"WLD,;;;:W,z.V. "%iH"U,,,A,.;,O,@t@JV.,,11!'oJt;l,IfflNfil!i-) ,,,.$2,"',4 '#.ihiiNt.",t",!&,. *"IliIWlIiih' '.. ,IlHiffi"T,.t, Mi&:\MJ ;2.4&" ,",JiSt C;;;:J:,W.,,$,,; q,; ;,:z.34Ma:4.;, ;"MMliM'.U!',lI.",&!@!4iili,P "Z$.w', p ",M-,",,, " "",>K"'"""
C\I \0 \0 .....
0
stllf)
=l11 t.i ag-e.:IaAV 0"1 0"1
0 0 0 0
0
0 0
C\I r;--;
t-
t- \0
0
smf)
t.i "=l11 t-e=lOJ, 0 \0 "OJ- 0"1 C\I
C\I C\I \0
t- \0 t- U"\
U"\ 0"1 \0
satqqac:I z=l.:I-enb C\I
A.:Ia:i:iOc:I
CO
.....
H
-=:
CO
E-i
CIJ
r:e::l
E-i

tJ:::
r:e::l
::>
tOOJ, -=:
S
tJ:::
I-i:l
0"1

r:e::l E-i
H A.
r:e::l
-=: A
E-i
paq:>no=lau >;
CIJ
:z;
pazTI1=lfi 0
H
E-i
::>
"U1qJ, a:>-eng:
t- O .....
H
\0 U"\ C\I
tJ:::
E-i
CIJ
H
A.:I-e1=l.:IaJ,
U"\ CO C\I A
U"\ CO C\I
E-i
0
-=:
J:o:>aa uo:>aS
0 \0
I-i:l
H
E-i
tJ:::
-=:
"J:o:>aa
"m1.:IeI
C\I

..... U"\ \0 C\I
C\I "OJ-
a.:I1.i
E: E:
El C) C)
E: E: El E: El
C)
0 U"\
E:
C) C) C) C) C)
U'\ C\I
C)
0 U"\ 0 U"\ 0 0 ..... .....
U"\ "OJ- \0 t- 0"1 .....
I I
I I I I I I U"\ 0
I U"\ 0 U'\ 0 U'\ 0 0 C\I
0 "OJ- \0 t- 0"1
28
38AL152
This site is related to a natural exposure of chert which shows evi-
dence of quarrying. Very large chert boulders, some over a meter in diame-
ter, can be seen strewn along the steep hillside facing south (Fig. 6).
Chert boulders and cobbles occur from the rim of the hill downslope for
about 100 meters (Fig. 3). Burned boulders and cobbles as well as ,chunks
and flakes can be' seen near the outcrops and. flakes can be found under the
leaves. In one location, a sizable hole in the hillside is present which
appears to be cutout of the hill as though quarried by historic period
people (Fig. 7). The eastern wall of this pit appears straight as though
dug systematically. The shape of the hole, however, could be a result of
erosional forces because a gulley leads away from it. An inspection of the
pit wall revealed humanly modified flakes and cores to a depth. of 30 cm.
Below that was rock. No auger testing was done as it is likely that the
cuI tural debris is not deeply vitreous on this rocky, steep hillside. At
least three types of rock were seen occurring here naturally: the vitreous,
fine-grained Flint River Formation chert; a loosely cemented sandstone; and
a highly fossiliferous chert. This quarry was discovered after the chert
sample had been sent off for petrologic analysis.
38AL137 The Swag Site
The Swag site is named for an eroded depression between two prominent
hilltops (Fig. 3). It was shown to us in June 1980 by David Topper of
Fairfax, South Carolina. Chert boulders and chips and cores are densely
distributed where two dirtroads intersect. Evidence of chert workings are
thinly distributed over a distance of 150 meters from the swag along the
roads. The eroded depression or swag is really the upper portion of a
creek bed which widens considerably down slope (Fig. 3). In the dirtroad
which runs west from the swag area, a section of naturally occurring chert
apparently utilized has been exposed by m o ~ n erosion. Next to this road
lies the creek channel with occasional boulders of unidentified rock ex-
posed. Given the rather minimal erosion here which has nonetheless exposed
chert, it is likely that considerable chert lies buried under the hillside.
From the standpoint of aboriginal procurement, however, access may have
been relatively limited. Mo st of the debitage and cortical material is
limited to the swag area suggesting this was their access point. Very few
large chert cobbles are present. The size of the site is not well known,
for trees and leaves form a dense ground cover. There is a mantle of sand
which provides some depth. Numerous chunks and flakes were seen, but no
cuIturally diagnostic artifacts were seen. Chert samples were taken for
petrologic analysis.
The site was tested in three loci (Fig. 3) in order to determine site
depth. Auger test 1 was located on the gentle slope of the hillside north
of the swag. Very few artifacts were found (Table 10), most of the flakes
being later stage reduction debris. There was approximately 90 cm of sand
within which the flakes were contained. The soil began to be clayish (red)
and produced no flakes after that.
Auger test 2 was placed near the heart of the depression (Fig. 3).
There was about 50 cm of sand containing flakes and chunks overlying an
29
Figure 6. Chert boulders and debris on hillside of 38AL152.
Figure 7. Inside of unusual pit showing bedrock at 38AL152.
30
u;
\D -:J 0-. \J'I
--
0
I\)
0 0 IJ'I 0 IJ'I 0 IJ'I I
0 IJ'I I I I I I I
--
I I
-
\D -:J 0-. \J'I IJ'I
- -
0 0 IJ'I 0 IJ'I 0
()
\J'I I\) IJ'I () () () () ()
S
IJ'I 0 ()
s s s s s
() ()
s
s s
o

g
-
I\) \D 0"1 IJ'I
-
I\) \J'I \J'I
. . . . . .
0-.
-
-:J ex> -:J IJ'I
I\) I\)
. . . . . .
0-. 0 IJ'I 0-. -:J IJ'I
IJ'I f\)
\J'I
I\)
I\) -:J 0-.
\J'I
I\)
Fire Cracked Rock
>
Chunks
::0
1-3
H
..."
>
Prim. Decor.
0
1-3
t:::::l
H
Secon. Decor.
m
1-3
::0
H
txl
Tertiary Flakes
c::
1-3
H
0
2:
Biface Thin. Flakes
m

Utilized Flakes
t:::::l
t'::I
"tl

Retouched Flakes ..."
::0 1-3

>
txl
Uniface
l:"'l
>
t'::I
c::

t'::I 0
Core
::0
1-3
t'::I
m
Cobble Tool
1-3
- ..
Hafted Biface
\J'I
ex>
>
l:"'l
-
\J'I
Preform
-:J
..
1-3
Blank
::x::
t'::I
m
.::e::
Pottery
>

m
H
Quartz Pebbles
1-3
t'::I
Total NUmber Flakes
Total Wt. Fl. Gms
Average Fl. Wt. Gms
9

g
g

I
g
n
9
n
n

n
B
o
o
increasingly rocky zone (Table 11). There were almost no flakes between 60
cm and below, and at 85 cm rocks were so thick that augering was impossi-
ble. The rock is a sandstone-like material commonly found in deposits
bearing chert on the Sandoz-Martin Works property. It appears that auger
test 2 contacted the formation bearing the chert.
Auger test 3 was placed a few meters east of auger test 2 immediately
adjacent to thegulley. The main prehistoric deposit here is buried under
a probable historic mantle of eroded sand. Flakes did not appear substan-
tially until the 30 to 45 cm level (Table 12), and an old burned A horizon
was contacted at this level. Flakes continued until about 90 cm where
sterile red mottled clay was reached. Auger test 3 is interesting for the
character of its debitage. Most of it represents the first stage of raw
material acquisition and preparation resulting in a large proportion of
chunks, and decortication flakes (Table 12). Many of the chunks bear cor-
tex, indicating the knappers were right over the source.
No tools or cores were recovered in the three tests, nor any time-
sensi tive artifacts. The period or periods .of utilization are thus far
unknown. The chert debitage from auger tests 2 and 3 was not particularly
weathered and in most cases was very bright and fresh. We found no worked
materials, a fact attributable in part to the heavy surface collecting of
the area.
The chert from the Swag site is classified as Category 7d, "light
brown, silicified grainstone" (Upchurch, Appendix A). This quarry contains
Allendale chert and is a member of the Allendale Chert Quarry Cluster
formulated by Upchurch (Appendix A). .
38A1146 "The GUlley Site"
This site lies immediately east of 38AL145 on the same terrace. Topo-
graphically, it is identical to 38AL145 in that it is situated on a high
shoulder of well-drained land overlooking the floodplain of Smiths Lake
Creek (Fig. 8). Chert cores and flakes can be found in the road which runs
south of the site. This site is separated from 38AL145 by a ravine (Fig.
8)
The site was tested by running a string of auger tests parallel to the
bluff. Five holes were excavated (Fig. 9), three terminating between 55
and 60 cm on red clay. Two other tests, 2 and 3, went considerably further
into the ground before striking sterile red clay. Auger test 2. produced
flakes until a depth of 105 cm; auger test 3 produced flakes until 90 cm
(Tables 13-17). The greater depth of auger tests 2 and 3 initially caused
us to think the site had substantial depth, hence stratigraphic information
of some duration of time. However, auger tests on either side of these two
(Fig. 9) indicate the familiar situation where sterile red clay is reached
at around 45 to 60 cm. The existence of a contemporary gulley only a few
meters to the north of auger tests 2 and 3 strongly suggests that these
tests came down into an old gulley which has been filled with sand and
flakes.
32
v
\.0 -:J 0'\
VI
......
0
I\)
0 0 \J1 0 \J1
0 \J1 I
0 'JI I I I I I I
......
I I
......
\.0 -:J 0'\ VI 'JI
......
--
0 0 'JI 0 'JI 0 (')
VI I\,) 'JI (') (') (') (') (')
S
'JI 0 (')
s s s s s
(') (')
s
s s
Fire Cracked Rock
>
I Chunks
I\,)
-- --
8 CD 0'\ 'JI
H
l-2j
>
I Prim. Decor.
0
'JI
8
t:1
H
l:I2
0
I Secon. Decor.
8 'JI
--
H
tIl
I Tertiary Flakes
c::: I\,) 0'\ VI
8 VI
--
\.0
H
0
!2:
I Biface Thin. Flakes
l:I2 I\,)
\.0 VI 0'\
-- tIl
r<
Utilized Flakes
t:1
t.:zj
'"d
8
Retouched Flakes
l-2j
8
>
tIl
Unifac.e
t-<
>
t.:zj
c:::
Q
t.:zj I\,)
Core

8
t.:zj
Cobble Tool
l:I2
8
VI
..
Hafted Biface
VI
CD
>
t-<
--
Preform
VI
-:J
..
8
Blank

t.:zj
l:I2
:IE;
Pottery
>
Q
l:I2
H
8 VI I\,)
Quartz Pebbles
t.:zj
0'\ VI 0
--
--
Total NUmber Flakes
VI \.>l. \.0
......
\.0 0'\
--

I\,)
'JI
I
Total Wt. Fl. Gms.
CD I\,)
--
\.0
--
-:J \.0 I\,)
. . . .
--
0'\ 0'\
0
I\,)
-- . \J1 0'\ I\,)
I Average Fl. Wt. Gms
. . . . .
.
0 O'\.-:J
-:J 0
--
-:J 0 0'\
/
/
/
/
/
/
I
/
/
AT3
---_.....
I
I
I
r'
/ '
'" \
",'" \
'" ..... _--"
",--""', "," ..... _-
",,- I --
...... I )
-_... .., I
, I
I I
I I
I I
I I
I I
I I
, I
, /
" '/ I I
, I
\ I
\ I
35
o 500 1000 FEET
METERS
\
\
Locator map for sites 38AL146, 38AL145, 38AL161, 38AL143, and Figure 8.
38AL135.
o
D
o
D
U
U
D
D
D
D
D
D
U
U
D
D
D
D
U
~
---+-
It)
~
~
-
Q)
Q,
"a.
-
i
'
:E
:>
~
a
~
<:]
~
~
f.t)
~
~
'3
~
~
~
....,
..CI1
~ N
~ ~
CI) ,."
.'
..
..
lLJ
-;
:Ii
t-
eo
(i)
I-
<0
en
>-
IA.I
s:
lLJ
I-
...J
...J
a:
<t ...J
'IA.I
C)
CD
:::>
-r
~
It)
(.!)
~
lLJ o.:t
~
:r
t-
e
0
Figure 9. Site map showing auger test locations at 38AL146 , the Gulley
Si teo
36.
TABLE 14
co
(I) II)
Q)
Q)
l5
(I) (J
co
l5
co r-I (I) Q)
r-I
C1l rz..
Q)
(I)
rz.. .

Q)
Q)
"C .
H
.
r-I (J
.-f H r-I Q)
H 0 r-I a .-f r-I CIS
..0
Q)
rz..
..!<l 0 (J
o.-f 0
r-I (J (J Q)
"0 0 o.-f
tIS Q)
t=l
>-
Q)
E-l I:Q
Pot
H A H .r:: Q)
e
>- z :3
Q)
u co
CIS
Q) N (J (J Q)
"0
H N tlO .
a -r-!
(J
-r-! .::s
tIS r-I Q)
0
Q)
r-I tIS Q)
s 0 tIS .-f 0 Q)
..0
III I-l I-l -r-! (J

I-l o.-f -r-! I-l ..0 C1l tIS C1l -r-! I-l Q) Q) -r-!
oS::
0 0
I-l r-I 0
<S-
o 0
U Pot CI)
E-l
I:Q
::J ::J u u
Pot I:Q
Pot E-l E-l
0-15cm
2c
1 11
1b 1 12 4.2 .35
15-30cm
1 1 10
1b '12 6.1 .51
30-45cm
1 1 4 58
65 47.9 .74
w
00
45-60cm
6c
4 6 8 233
3 252" 258.2 1.02
60-75cm
3 7 5 166
1d
4 181 196.6 1.09
75-90cm
3 2 28
6
33 11.0
.33
90-105cm
1
3 1 13
1 18 10.6
.59
105-120cm
120-135cm
Key - a. One limestone cortex chunk and one sandstone (?)
b. Plain, fine sand tempered
c. One chert, fine limestone cortex
d. 116 gm%
r::=J jj IJ I I ., E:::j EJ I J EJ C'.:J I> I E:::j Ie"," I r:::::J C:J r::=J
TABLE 15
(/) CIl CIl

Q) Q)
.!:ol

CIl
0 CIl
CIl r-l CIl Q) r-l

Q)
Q)
CIl
.j.J
.

.!:ol
Q) Q)
::s:
'0
.
I-l
.
ClI r-l 0 r-l
r-l
Q) I-l 0 r-l s:: r-l r-l ClI ..0
.!:ol 0 0 .... 0
r-l
0 0 Q) ,.c:: "t:l 0 ....

ClI
Q)
A l E-i "'d
Q)
E-i Pol
.j.I
I-l A I-l
Q) ,.c:: Q)
l Z ::s:
Q)
u (/) .
ClI
Q) N 0 0
Q)
"'d I-l N bO
.!:ol
.
s:: 'M 0 'M ::l ClI r-l CII 0 .!:ol
Q) .j.I r-l r-l ClI
Q)
a 0 .j.I ClI r-l 0 ll-l OJ ..0

.j.J ll-l
fa
.j.J I-l. ClI ClI
I-l 'M 0 I-l 'M
.j.J
'M I-l .0 ll-l Q) .j.J
ClI
.j.I .j.J
'M
,.c:: I-l Q) Q) 'M
.j.J Q)
g
0 0 ClI I-l r-l 0 ::l 0 0
U Pol CI.l E-i P P:l u u lI:l Pol CY E-i E-i
0-15cm 4 1a 4 1.6 .4
15-30cm 16 1a 1 16 9.1 .57
30-45cm 4 16 1b 20 11.3 .56
w
\0
45-60cm 1c 1 50 1 51 32.8 .64
60-75cm 2 2 2 40 46 30.1 .65
75-90cm 1
,2
8 1d 11 7.6 .69
90-105cm
105-'120cm
120-135cm
Key - a. Plain, fine sand tempered
b. Base of small triangular point
c. Limestone (?) cortex fragment
d. 151.6 gm
c::J I::] lei I I::] E::l E::l IE::J I:::) c:::J IE::J c:::l I:::::l E.:::] E.:::] r:=J
TABLE 17
(/)
(/) (/)
Q)
Q)
c!3

(/)
CJ
to
c!3

(/) .... to Q) ....
Q)

Q)

(/)

.....
.

Q) Q)

"0
. k
. .... CJ .... k ....
Q) k 0 .... I=t ....
.... CIS
,.0 Q)
f:a:.l
0 CJ ..-l 0

....
CJ CJ
Q) .r:: "0 0 ..-l
f:a:.l
CIS
Q)
A >- E-l
aJ
Q) E-l
pQ Pol
.....
k A k ,.d Q)
a
>. z
Q)
U
(/) . CIS
Q) til <J CJ
Q) "d k til bO

.
I=t
.... <J ..-l ::s CIS
.... Q) 0

Q) ..... .... .... CIS
Q)

0 ..... CIS .... 0
Q) ,.0 4-l

cd cd k
k

CJ

..-l ..... ..-l k ,.0
Q)
CIS ..... ..... ..... Q)
.... k
Q) ..-l ..... Q)
g
0 0

k .... 0 0 0

Pol CIl E-l
pQ :::> u U Pol
pQ Pol G E-l E-l
0-15cm 4a 1 1 1
2b 1 4 9.4 2.35
15-30cm 2c 2 3 1 41
1d 47 30.3 .64
30-45cm 4 6 6 74
1e 9 90 100.0 1.11
.j::-.
45-60cm
f-I 1c 1 7
8 2.0 .25
60-75cm
75-90cm
90-105cm
105-120cm
120-135cm
Key - a. Three limestone (?) cortex, one chert
b. Plain, fine sand tempered
c. Limestone (?) cortex
d. Quartz cobble fragment
e. Chunk of burned clay
No time-sensitive artifacts were found except plain ceramics and the
base of a small chert triangular point. All sherds were found in the upper
30 cm of the site, and most of these came from the upper 15 cm (Tables 13-
17). The triangular point came from 30-45 cm in depth (Table 15). Inter-
estingly,even though auger tests 2 and 3 are thought to be located over a
gulley, the pottery from these two was restricted to the upper 30cm of the
. site. . This suggests that the filling in of. this supposed gulley occurred
before approximately the last 2,000 years. There are no tools present in
the debitage from the lower levels, however, that would indicate a pre-
ceramic occupation. One unusual characteristic of auger test 2 is the.
enormous number of flakes recovered between 45 and 75 cm (Table 14). Under
the gulley hypothesis, it is possible that flakes were concentrated in the
low spot by winnowing action from running water.
Like 38AL145, the sand on this site was probably ultimately derived
from sheetwash and colluvium originating from the adjacent hills (Fig. 8).
Bioturbation in this soft sand is a viable alternative explanation (Michie
1983). Site 38AL146, like 38AL145, has no definite evidence of a pre-
ceramic occupation, suggesting a relatively late origin for the sand mantle
on the terrace. This site has further research value for studying Woodland
and Mississippian period occupations and utilization of chert at the
source.
38AL145 "Boundary Marker 232"
This site was discovered by auger testing. Flakes and cores are pres-
ent on the dirt road which runs to the south of thesi te along the Savannah
River Plant boundary (Fig. 8). It was decided to subsurface test the
terrace which lies between the road and Smiths Lake Creek because this area
was a high but level area of ground overlooking the floodplain, one which
also had access to chert (Fig. 8). The site is mostly on Savannah River
Plant land and is named after the closest S.R.P. permanent boundary marker,
number 232. To the south, the site extends on Sandoz-Martin Works land at
least as far as the road and for an undetermined distance south of the
road. The site is covered with large pines and is in a well-protected
situation. It is possible it has been plowed, although this has not been
determined. In addition to the prehistoric occupations present on the
terrace, a historic earthen dam was constructed across Smiths Lake Creek in
front of 38AL145 (Fig. 10). Its age has not been determined. The function
was probably that of a mill for there are timbers lying fallen into Smiths
Lake Creek. -The wid th of the dam is between. two and three meters and has
very mature hardwoods growing in the dam. Immediately north of the site
datum was a small rectangular pit (Fig. 10), which was perhaps a borrow pit
related to the dam construction.
Seven auger tests were dug in a line over approximately 100 meters of
the terrace where it overlooks the floodplain (Fig. 10). All tests pro-
duced lithic artifacts as well as pottery (Tables 18-24). Decorated sherds
indicate an occupation span from Savannah I (A.D. 700-1200) to Savannah II
(A.D. 1200-1600) (Stoltman 1974: 27-31). Ceramic evidence for this includes'
a single cord marked sherd, four check stamped sherds, three curvilinear
complicated stamped sherds and five burnished sherds. The base of a small
42
r:::::J I'u; ..] . :.',
I:::i r::::::l r::::::J
c:J

\j
-
41J.
4-
4
.4 4-
OLD EARTHEN----....
DAM
4-
4-
4
4- 4.
\JOLo BORROW PIT
.4-
4-
41/-

.p-
I.>,)
eAT 7
.eAT5 eAT4
38ALl45
BOUNDARY MARKER 232 SITE
o . 15 Meters
. , , .
,AT3
eAT6
ATI ..
,AT2. &oATUM (pipe)
00
TWIN PINE
TREES
e AUGER TEST
I!'igure 1o. Site map showing auger test locations at 38AL145, Boundary
Marker 232 si te.
+-:"(l 1)
..;::::<,,,_... .__?'i,E._" ,. "a. .'_'".". ,.L.k.. , .. ES...." .iE,....kRlU ,...;SL..,L .3!se .."i .. ..&JU, .. ..o,,, . ,,, ..i.S.,._,,,
TABLE 18
CJ)
CJ) CJ)
OJ
OJ

&
,.!.:\
CJ) tJ CJ)
.:'1 &
a to r-I CJ)
OJ
IXl OJ OJ
CJ)
.j.J . ,.!.:\

OJ
OJ
"0 .
1-1
'"
.
'"
r-I
tJ
r-I 1-1 r-I OJ 1-1 0 r-I s:: r-I r-I Cll
..0 OJ
,.!.:\ 0 tJ 'rl 0 lJ.l ,.0

r-I tJ tJ OJ .c "t:l 0 'rl OJ

'"
OJ A
t'
E-l "t:l OJ
E-l pQ
::1 .j.J
1-1 A OJ .c OJ
>. z OJ
t) CJ) .
'"
OJ N tJ tJ OJ "t:l 1-1 N
be
.
s:: 'rl tJ 'rl ::1
'"
r-I OJ 0 OJ .j.J
r-I r-I
'"
OJ
0 .j.J
'"
r-I a lJ.l OJ ..0 .j.J lJ.l
g
.j.J
1-1
'"
'"
1-1 1-1 tJ 1-1 4-l 'rl .j.J
rl 1-1 ..0 4-l OJ .j.J
'"
.j.J .j.J
OJ 'M .c 1-1 OJ OJ 'rl .j.J
s:: 0 0
1-1 r-I 0 ::1 0 a
t)
CI)
E-l pQ
l:J l:J
t) t) j:4 pQ
0- E-l E-l
0-15cm
4
8a
4 2.2 .55
15-30cm
12
7b 2 12 9.8 .82
30-45cm
1 14
3c 1 15 5.6 .37
-l:'-
45-60cm
5
5 4.6 .92 -l:'-
60-75cm
75-90cm
90-105cm
105-120cm
120-135cm
Key - a. Five plain, fine sand tempered; 2 checked stamped, 1 burnished
b. Five plain, fine sand tempered; 1 cord marked, 1 burnished
c. Three plain, fine sand tempered
r::::::::J I'll] i",' I!!!! I!!! I!I!J Il!!!I!J E::'J r:::::! E:] r:::J
TABLE 19
co co co
Q)
l5
,.!&'. co
(J
CIS co
l5
co r-I co Q)
r-I
Q)
rz.
Q)
co rz. .c.J
.

Q) Q)
::::
't:S
.

. r-I (J
r-I r-I
Q)
0 r-I
r-I rz. r-I CIS .0

Q)
rz.
,.!&'. 0 (J
rz. rz. 0 'l-I
r-I
(J (J Q)
.d 't:S 0 '1"1
rz.
CIS
Q)
E-t 't:S
Q)
E-t P=l III .c.J

Q)
.d
Q)
e
>. 'z ::::
Q)
C,) co .
CIS
Q) N
g
(J Q)
't:S N be
.
c:: '1"1
(J
'1"1 CIS r-I Q)
0
Q) .c.J r-I r-I

Q)
.
0 .c.J CIS r-I 0 'l-I Q) ,.Q .c.J 'l-I .c.J CIS CIS
(J
'l-I '1"1 .c.J '1"1 ,.Q 'l-I Q)
CIS .c.J CIS . .c.J .c.J
'1"1 .d Q) Q)
'1"1 .c.J
g
0 0
&l
r-I 0 ::s 0 0
rz.
C,)
III 00' E-t P=l l=l
C,) C,)
III P=l III 0- E-t E-t
0-15cm 1a 1 1b 1 1 .5gm .5
15-30cm 2 29 1c 7d 31 25.0 .81
30-45cm 1e 1 23 1f 19 24 5.9 .24
.po
VI
45-60cm
60-75cm
75-90cm
90-105cm
105-120cm
120-135cm
Key - a. Quartz,
b. Plain, 'fine sand tempered
c. Base of triangular point
d. One burnished plain, five plain, and one
check stamped, all fine sand tempered.
e. Limestone
f. 92.6 gm
g. Burnished plain
!!!!!'::7,. !f.,;rt", ',,:;hdfL ABf.,8\ .'''''',fiAh;- "h:XiL ;:;;:Xgh ;:::;',. -,t; ""dc,::.., ,.,};; :::;;; "',;; :c, p:, d:'t:! f';; ,i; 'J,- #,4 ". 4 ,A
;.;:
CD
r..::
I
O'l:l:l
I-tlO
I-'::S
l:l:l CD
1-'-
::s 0'

Ii
H). ::s
1-'- 1-'-
::s l'J)
CD
CD
to p,
l:l:l
::s
P,c+

e+Ii
CD CD
S CD
"0.
CD "0
Ii I-'
CD l:l:l
P, 1-'-
::s

o
::s
CD
(')
o
S
"0
I-'
1-'-
(')
l:l:l
c+
CD
P,
to
.c+

"0
CD
P,
l:l:l
I-'
I-'
H)
1-'-
::s
CD
to
l:l:l
::s
P,
c+
CD
S

Ii
CD
P,
9v
\D -J 0\ Vl
.....
0
I\)
0 0 U1 0 U1 0 U1 I
. .. "'-.
0 U1 I I I I I I
.....
I
I I
.....
\D -J 0\ Vl U1
..... .....
0 0 U1 0 U1 0 (')
Vl I\)
U1 (') (') (') (') (')
S
U1 0 (')
S is S S 51
(') (')
51
S 51

l
.
'1:l
I-'
l
1-"
::s

H)
1-"
::s
(l)
til
l
::s
p.,
c+
(l)
= to
(l)
Ii
(l)
p.,
6i?
1.0 -.J \)l --'0
f\)
0 0 \J'1 0 \J'1 0 Ul I
0 \J'1 I I I I I I
--.
I I
.....
1.0 -.J
\)l \J'1
--. --.
0 0 \J'1 0 \J'1 0
(')
\)l f\) \J'1 (') (') (') (') (')
=
\J'1 0
(')
= = = = = (') (')
=
= =
f\) --.
\)l
f\) --.
f\)
-.J \J'1
f\)
l l
f\) f\)
f\)
CD CD
f\)
f\) \)l. \J'1
. . .
Ul Ul
. . .
\J'1 CD
-.J 1.0
Fire Cracked Rock
Chunks
Prim. Decor.
Secon. Decor.
Tertiary Flakes
Biface Thin. Flakes
Utilized Flakes
Retouched Flakes
Uniface
Core
Cobble Tool
Hafted Biface
Preform
Blank
Pottery
Quartz Pebbles
Total Number Flakes
Total Wt. Fl. Gms
Average Fl. Wt. Gms.
>
::d
8
H
I7.j
>
Q
8

H
t:/)
8
::d
H
to
c:::
8
H
0
2:
t:/)

8
t:1 >
lZJ to
'1:l t"'
8 lZJ
::t::
f\)
I7.j \)l
!:tl
0

>
c:::
Q
lZJ
::d
8
lZJ
tr.I
8

\)l
CD
>
t"'
--.

\J'1
o
8
g
n
8
g
n
g
g
n

9
o
B
g
n.:.>
u
a
o
o
-
VI
o
0-15cm
15-30cm
30-45cm
45-60cm
60-75cm
75-90cm
90-105cm
105-120cm
120-135cm
()
<II
l-I
Co)
Q)
l-I
OM
CI.l

1b
.
l-I
o
()
Q.l
A
.
S
0'"
l-I
III
TABLE 24
CI.l
CI.l
CI.l Q.l
Q.l
CI.l Cll CI.l
CI.l r-l CI.l .Q)
Q)
OJ
CI.l
.
Q.l Q)

l-I .
cd r-l ()
r-I l-I 'r-!
0 r-l r-l
r-! Cll
..a Q.l
()
:E

0 \I-l
..a-

r-l
Q.l
"t:I
0 OM
Q.l
8 "t:I Q.l
8
III Q.l ,.d Q)
e
:>- z OJ
.
Cll OJ til () ()
OJ "t:I
l-I til
00
OM
cJ OM
::s cd r-l Q.l 0 OJ r-l r-! cd
0 cd r-l 0 \I-l Q)
..a
&1

<II cd l-I
()
l-I oM OM
l-I ..a Q.l Q.l
Q.l Q.l OM Q.l
0 0
pg
l-I r-l 0 ::s 0 0
til 8 l=l l=l Co) Co)

III III 0- 8 8
11
6a 11 6.6 .60
10
2c 12 8.7 .72
13
3c 1 14 10.8 .77
1e 1d
Key - a. Four plain, two complicated stamped, all fine sand tempered
b. Quartz
c. Plain, sand tempered
d. Check stamped, sand tempered
e. Quartz cobble fragment
'" "'o/A,"jf; '. ,', !6!It,kL,.. ,0j,.0i..i# .. ti5i,.,ih..,iA.N dll!8...fAf, ,. , """iiL,$).!\'T,.i" e, ."M!.\ ,.2+' h ,I'!!'!:"@,,, . ,,i!.'il1'Iit!f "JMf,b IS ,,,',il'Jt,ffi$'t, AJ#,.,.%i, 0;. h!J,i\f" . ; &Iii .$$!i,I", _"J.,""-W.,,. "!l'/If,L,_. ,A , .'I!!. ,L,,..i$iE f! "l.,J. .\if. "fi@ f.9AJ32 $".,!JiJ ",.,)\!!i'-7\. )6..,:5,", h ,, ,;',iM,@
o
Ii",!",!,
tJ
IJ,'.,',','"
II
D
I
U
o
I
I
I
I
I
o
D
D
o
I
o
triangular point found in auger test 2 also supports an occupation within
this time frame.
Artifactual debris from the site is restricted to a nearly pure sand
mantle which overlies a culturally sterile red clay subsoil. This clay was
reached rather quickly during testing between 45 and 55 cm in depth. Based
on the age of the artifacts, the sand overlying the clay must be only a few
thousand years in time. This is assuming a colluvial origin for the sand.
No other lithic artifacts were found which would indicate preceramic' occu-
pations. The geomorphological processes responsible for this relatively
late deposition of sand on the terrace are not known, but the sand is like
that found on the hillslopes over 300 meters away. A colluvial sheetwash
is a possible explanation for the sand'deposit bearing artifacts. Biotur-
bation is also a possibility here as a means of burying artifacts (Michie
1983). Because of the evidently Woodland occupation, "this site has value
for what it might reveal about chert procurement and utilization in the
last 2,000 years. For example, there is only one core and very few early
stage reduction flakes (Tables 18-24).
38AL161
This site is situatedalong a terrace-like feature at the pediment of
the uplands. The western perimeter is defined by the bluff edge along the
80-footcontour (Fig. 8). The north-south axis is approximately 1,500 feet
and the east-west axis is at least 200 feet. Subsurface testing was only
done along the terrace edge; the eastern site boundary is unknown as it is
obscured by vegetation. The site was discovered by the presence of spora-
dic flakes exposed in the dirtroad maintained by the Sandoz-Martin Works.
The site is interesting from an ecological perspective because it appears
to be related to utilization of the higher, better-drained terrace which
overlooks a swamp.
Three auger tests were dug to gain some knowledge of site depth and
deposi tion (Fig. 8). Auger test 1 was placed on the northern end of the
site at the intersection of dirtroads. After the A horizon (ca. 15 cm),
the soil was a uniform tan sand until about 75 cm where a grey clay with
orange mottles was reached. Artifacts only occurred in the sand. Table 25
presents the artifacts recovered which were exclusively debitage. Numerous
small quartz pebbles were encountered but not saved.
Auger test 2 was placed in a clearing next to the road (Fig. 8). It
presented a different sedimentological picture. Here, the soil was coarse
reddish-brown throughout the profile. Interestingly, artifacts were absent
in the upper 30 cm but continued to a depth of 90-105 cm (Table 26). Like
auger test 1, only debi tage was found. These flakes were typically
weathered and displayed a yellowish color. The association of reddish soil
and highly weathered flakes is again noted. There were numerous small
quartz pebbles in this red soil.
Auger test 3 was dug at another road intersection (Fig. 8). The pro-
file he.re was similar to auger test 2, except the depth only extended to
60-75 cm (Table 27). A single plain sherd was found in the upper 15 cm
level.
51
ZS-
\.0 -..:J 0'1 -"
0
I\,)
0 0 \J1 0 \J1 0 \J1 I
0 \J1 I I I I I I -"
I I -"
\.0 -..:J 0'1 \J1
-" -"
0 0 \J1 0 \J1 0 ()
I\,)
U1 () () () () ()
=
U1
0 ()
S
= = = =
() ()
=
=
S

\.0 -J
"""

0
f\)
0 0 \11 0 \11 0 \11 I
0 \J'1 I I I I I I

I I

\.0 -J 0\
"""
\J'1

0 0 \J'1 0 \J'1 0 ()
f\) \J'1 () () () () ()
S
\J'1 0
()
s s s s s
() ()
s
s s

0 """
/
.....
"""
-J 0 \J'1
.....
.....
0
"""
-J CD
f\)
. . . .
..... .....
-J \.0
I\) -a.
I\)
. . . . .
\J'1 \J'1 -J \.0
I\)
CD -J 0 0
Fire Cracked Rock
Chunks'
Prim. Decor.
Secon. Decor.
Tertiary Flakes
Bifaee Thin. Flakes
Utilized Flakes
Retouched Flakes
Uniface
Core
Cobble Tool
Hafted Biface
Preform
Blank
Pottery
Quartz Pebbles
Total Number Flakes
Total Wt. Fl. Gms.
Average Fl. Wt. Gms
<
t?;l
l:tl
1-3
H
Q
>
t""
t;
H
U2
1-3
l:tl
H
ttl
c:::
1-3
H
0
!2:
0
I"jl
:to-
l:tl
1-3 1-3
H
>
"':I ttl
>
t""
Q t?;l
1-3
U2
I\)

"':I
l:tl
0
:s=
>
c:::
Q
t?;l
l:tl
1-3
t?;l
U2
1-3
f\)
..

CD
>
t""
-a.

0/'.: ....
u
n
R
D
n
I
I
n
I
I
I
g
I
g
I
...
!I
B
B
o
X
en
t<
tTl!'
I l ~
l!' ~
::l ::l
p-en
ta
c-t-ta
o l!'
::l ::l
en p-
---. c-t-
." en
-s
(")'g
o Ii
Ii en
c-t-P-
en
1><'0
~
l!'
~
::l
i]S
\0 -J 0"\ ~ \>l
~
0
I\)
0 0 \}1
0
\}1
0
\}1
I
0
\}1
I I I I I I
~
I I
~
\0 -J 0"\ ~ \>l \}1
~ ~
0 0 \}1
0
\}1
0 (")
\>l I\) \}1 (") (") (") (") (")
s
\}1
0 (")
s s s s s
(") ()
s
s s
o
u
o
u
o
o
o
o
u
u
u
u
u
u
u
U
D
U
U
Obviously three eight-inch auger holes in a site as spatially exten-
sive as 38AL161 hardly constitutes very much data with which to generalize
about this site. On terms of its location the site is no closer than 300 m
to the nearest upland chert source (38AL152). Flakes from the tests were
relatively small and cores were not found, which together suggests that
later stage tool manufacture and core reduction were related to the occupa-
tion of the bluff edge.
The mode by which these artifacts became buried, up to a meter in some
places, is enigmatic. It seems unlikely that alluvial deposition occurred
because sites of probable contemporaneity lie between 38AL161 and the
Savannah River. Site 38AL135 (Fig.8), for example, lies on an alluvial
terrace dominated by clay and silt. Heavy sands should drop out closer to
the river source rather than farther away. The sands represented. at
38AL161 are like those on the hillsides 300 to 400 meters east. This
terrace seems to have been formed by colluvium and sheetwash of sands from
upland sources. The slope , however, is only 1%(10' 1000') which would
seem to preclude the deposition of artifacts by colluvium. This site is
located in a situation similar to that of 38AL145 and 38AL146 (Fig. 8).
Bioturbation, therefore, might be a better explanation for their artifact
burial, an idea presented before for stratified upland sites on the Coastal
Plain (Michie 1983). More testing along the terrace edge might prove
valuable toward discovering sites other than quarries in the study area.
38AL143 Big Pine Tree Site
This site was discovered by the presence of flakes exposed in a dirt
road and by flakes eroding from the bank of Smiths Lake. The site is named
"Big Pine Tree" because of a large pine, nearly a meter in diameter at
breast height, located in the middle of the road which serves as a turn-
around. The western boundary of the site is Smiths Lake where it begins at
its upper reaches, resembling as it does B.ll old cutoff of the Savannah
River (Fig. 8). The northern boundary of the site is also Smiths Lake
Creek where it turns eastward (Fig. 8). The remaining borders of the site
have not been determined by testing, but it is probable that the main con-
centration of artifacts follows the higher portion of the floodplain as it
exists along the margin of Smiths Lake Creek and Smiths Lake. An estimate
of site length following the creek would be ca. 60 meters north and south.
The east-west dimension is not known, but auger tests revealed artifacts at
least 16 meters away from the creek. The site is a chert processing loca-
tion where chert cobbles, probably obtained from the creek bed of Smiths
Lake Creek, were reduced and transformed into cores and other tools. The
sediments are purely alluvial in origin, deriving from flood episodes from
the Savannah River and possibly Smiths Lake. Creek. It is clear from ,the
U.S. G.S. topographic maps that the Savannah River has moved eastward at
times and captured Smiths Lake Creek. This also explains the "lake" fea-
ture called Smiths Lake which has the shape of an old cutoff.
When first discovered in August 1983, the site was heavily overgrown
and very little artifactual material was visible. In January 1984, many
flakes could be seen in the turnaround area of the road. Of particular
interest were quantities of flakes eroding out of the bank where they were
washed out of the floodplain by fluctuations of the Savannah River which
55
regularly backs up into Smiths Lake Creek Valley. of the flakes
exposed on the bank, it was decided to test for the presence of artifacts
buried in the floodplain away from the bank.
Auger test 1 (Fig. 11) was dug quickly without screening to check for
. buried artifacts. A large number of flakes came up in the auger, as well
asa biface fragment (Fig. 12d) between 30 and 60 cm, and an endscraper
(Fig. 12g), which was found at approximately 80 cm surface. At about
this depth artifacts stopped. Based on the results of this test, it was
decided to systematically auger test the floodplain for buried artifacts.
A transit was set up over a permanent brass marker set in a concrete
post. This marker was erected by the Atomic Energy Commission and marked
as "S.R.O. 233," now the boundary of the Savannah River Plant. It is
dEmoted as "C32" on the Sandoz-Martin Works maps (Fig. 8). Using this
marker as a site datum, four additional auger tests were dug. A plan map
(Fig. 11) was made of three test loci and their relationship to Smiths Lake
Creek and the road turnaround. The results of each auger test are reported
in Tables 28-31.
Few cultural diagnostic artifacts were found in the five auger tests.
A hafted endscraper was found in auger test 1 (Fig. 12g). It is thought to
be either Paleo- Indian or Early Archaic In addition , five plain sherds
were recovered from auger test 5. Stratigraphically, . there is a signifi-
cant clustering of bifaces wi thin the 45 to 90 cm level of the site,
appearing in all auger tests except auger test 2. The three bifaces from
the 45-60 cm zone (Fig. 12a-c) are distinctive in that they are not very
weathered and are brightly colored and glossy, reflecting thermal altera-
tion. Two of these (Fig. 12a, b) are well shaped and thin and may be
broken hafted bifacial tools or nearly finished preforms broken at the last
stage of completion. Technologically, these pieces are similar to Guilford
and Brier Creek lanceolate forms of the Late Middle Archaic on the South
Carolina and Georgia Coastal Plain (Michie 1968; Brockington 1971; Anderson
et al. 1979) because of the rounded and indented basal outlines and the
emphasis on heat treating. The two biface fragments from the 60-75 cm
level are tips which reflect either tips of early stage hafted bifaces or
late stage preforms. One (Fig. 12e) is brightly colored and lustrious,
reflecting heat treatment, while the other (Fig. 12f) comes from a broader
biface and is weathered to a yellowish cream color. The former is similar
to the biface technology of the level above, while the second resembles the
Early Holocene bifaces of Paleo-Indian and Early Archaic types. The biface
from within 30-60 cm level of auger test 1 is heat treated and lightly
faded from weathering (Fig. 12d). It too could be either a finished point
or preform. The biface fragment found from 75-90 cm is the distal portion
of an early stage blank (Fig. 12h). It is markedly weathered. The glossy
surface which survives despite deterioration suggests it too was thermally
altered. Its low position in the site and its weathered condition suggests
an Early Holocene age.
Sedimentologically, there is a general uniformity revealed in the
auger test. The first 15 cm in every hole had a clayish sand probably
referrable to historic floods of the Sayannah River. From about 15 cm to
90 cm below surface, a coarse goldish sand dominated. Between 90 and 100
em, a clay horizon was encountered which possessed red and gray mottles.
56
Figure 11. Site map showing auger test locations at 38AL143, Big Pine Tree
si teo
.
o

0::
I-
0::
o
C\I

57
CI)
a
o
o
3:: .
o
D

B

D
o
D
D.]
I
U
D
D.i..
I
I
n....;l
II
D
U
o
o
e
0.!i'iiiiiiiiiiiiiiiiiiiic=:iIii__==iiiiiii_i!?f IN.
5 eM.
Figure 12. Lithic artifacts recovered from auger testing at 38AL143, Big
Pine Tree site.
58
D
C\I "<t I:- ..-
a8a:ZaAV
C\I 0 I:- <Xl C\I

li
0
D
\0
""
<Xl \0 Lt'\ 0
0 0 0 0 0
D

li
0:Ji'l
Ta=lo.r. tr\ I:-
""
.....
"<t "<t C\I tr\
"<t C\I I:- "<t Lt'\
.:taqmnN Ta:Jo.r. "<t \0 tr\
D
saTqqad ..- ..-
0

tr\
"<t
U
..-
...:I

<
<Xl
tr\
..
m.;oJ:a.Id
D
C\I
8
Cf.l
r:il
aoaHg
8
III
-C'
TOO.r. aTqqo:>
::>
<
g
:::: ,c
Q)
0
a.Io:>
CUl p::::
'Cl
<Xl IZ!
Q)
C\I
::z::
'Cl
r:il 8
aoaHUn Q)
D
...:I Pol
..s:::
P=l r:il
.....
ES
A
0
0
>-t

El
P=l
OJ
I
Cf.l
..s::: :z;
.....
0

M
H
8
::>
.....
D
P=l
"<t <Xl
""
Lt'\ Lt'\
I:l
H
oU"]:q.r. aoaJ:"]:g tr\ Lt'\ C\I Q)
p::::
El
8
CUl
Cf.l
III
H

D
A

8
Q)
0
r-t 0
<
,c
El
IZ!
.IOOaa ouooas <Xl Lt'\ ,c
CUl
H
0
D
8
()
I:- p::::
0
<
()
0
.IOOaa m"]:.Id C\I
M I:-
I:l
III
..
() .....
r-t I:l

0 Q)
>
El
III CUl
C\I ..... III
Q)
D
a.I"]:i
.....
El
tr\

III
0 r-t
I:l
CUl
U

s:::
<
El
El El
() ()
El El El El El
()
0 Lt'\ III ,c
El
() () () () ()
Lt'\ C\I tr\
()
0 Lt'\ 0 Lt'\ 0 0 ..- ..-
D
Lt'\ tr\ "<t \0 I:-
""
..- I I
..- I I I I I I Lt'\ 0 >.
I l!\ 0 l!\ 0 l!\ 0 0 C\J Q)
0 tr\ "<t \0 I:-
""

0
59
TABLE 29
ARTIFACT DISTRIBUTIONS BY DEPTH FROM AUGER TEST 3. 38AL143
(I)
(I) (I) Q)
.Q)
(I) u
ClI (I)
ClI
l5
0 (I)
r-l (I) Q)
r-l
p:;
Q)
r.r. Q)
.
co r.r. 4.1
.
Q)
Q)
l3:
"t:l .
1-4 III .
ClI r-l
U
..-l 1-4 ..-l
Q)
1-4 0 r-l s:l r-l r.r. r-l III
,.0 Q)
r.r.
0 u r.r.
:a
r.r.
0 4-l
,.Q
..-l
U U Q)
'C
0 ..-I Q).
r.r.
III Q)
A >. E-c 't:l Q)
E-c j:Q
Pol 4.1
1-4 A J.j Q)
..c: Q)
e
>. z l3: Q)
u CD .
ClI Q)
N U U Q)
'C
1-4 N l)l)
1
.
s:: 0.-1
U 0.-1
=='
ClI ..-l Q)
0
Q)
4.1 ..-l ..-l

Q)
13 0 4.1 ClI r-l 0 4-l Q) ,.0 4.1 4-l 4.1 J.j
III ClI
J.j 0.-1
U 1-4 4-l 0.-1 4.1
. 0.-1
1-4 ,.0 4-l Q)
III 4.1 ClI 4.1 4.1 Q)
0.-1 J.j Q) Q)
..-I 4.1
:5
0 0
1-4 ..-l 0
=='
0 0
r.r. u Pol tJ:l E-c j:Q

::J u u
Pol j:Q
Pol 0' E-c E-c
0-30cm
1 2
34 1a
36 19.8
.55
30-45cm
1 1 2 26
30 34.9 1.16
0\
45-60cm
1 2
4 1
1b 1c
69 27.2
.39
0
60-75cm
4 5 9 136
1d 2e
2 155 140.3 .90
75-90cm 1 2 1
45
49 35.7 .73
90-105cm
1 11
12 6.2 .52
105-120cm
120-135cm
Key - a. Quartz cobble fragment
b. Basal fragment
c. 284.2 gm.
d. Two distal fragments
", ,,,;it,,,,.,. - .HiE!, .. ,fl. .L, :,,*,. #5'"'., , .' ,""''"f ..Z M% ;, y'" , ... ;:. ,",.Q.". c_,_ ci#J.ttt. _":,,,,!, _.i"" ;M ,..2..,c- -c:' ,'-'c'::,,,!!,:!
TABLE 31
l/)
l/) l/)
C1l
C1l

.. l/) tJ
Cd l/)
Cd
0 l/)
r-l l/)
C1l
r-l
A::
C1l C1l
to .j.J .

C1l
C1l

't:l .
Cd .
.-I
0
.-I .-I
C1l 0 .-I s:: r-l r-l tIS
.0 C1l
'tl
0 (J
oM 0 Il-I
.0
.-I
0
C1l
..c: "C 0 oM
C1l

tIS C1l t=l
t'
E-t "C C1l
E-t !Xl
t.\.l .j.J t=l
C1l ..c: C1l
e
>- z <U
c.J l/) .
tIS C1l N tJ 0 C1l 't:l
N
bO
.
s:: .-1
0 .-1
::J tIS r-l C1l 0 C1l .j.J
.-I .-I tIS
C1l
S 0 .j.J
tIS .-I 0 Il-I C1l ..0 .j.J
'H
.j.J
tIS tIS k
.-1
0 Il-I .-1 .j.J .-1
..0 Il-I C1l .j.J
tIS .j.J .j.J
<U
.-1
C1l C1l '.-1
.j.J
g
0 0
r-l 0 ::l 0 0

c.J t/)
E-t !Xl :;:l
c.J c.J
!Xl 0- E-t E-t
0-15cm
1
1 15 15
15-30cm
1 1
7
1
9 11.6 1.29
0\
30-45cm
5a
N
45-60cm
15
15 5.2 .35
60-75cm
1 26
29 18.7 .64
,
75-90cm 5 3 27
1b
1c
37 62.5 1.69
90-105cm
7
7 31.5 4.5
105-120cm
120-135cm
Key - a. Plain, fine sand tempered
b. 176.9 gm.; river cortex present
c. Distal portion
"',":,Y'!/,j',":'," ".. - - #'_"0'''_ ..tl5;t- ,AiI)2;,., . :-- Jc,,,%,, -.(,I..,. !Jkl5t,. : "}".\.,,H ....."," j.j'dJ&",. !ffi.'.t@,' "<',-''''" ,W!..fl,'-*' 'I'.!- ",* . "",_,., 1",- d,; -,..,.stiA "' "':"f.""f!l
o
D
U
U
U
~
U
U
~
~
lij
II
~
~
i
~
~
IR.i.
II
~
o
No artifacts were found in this clay. The best prediction of artifacts at
this site was the presence of sand. Based on limi ted testing, it appears
that the site is deeper toward the creek bank, probably attributable to the
increase in levee height in that direction.
The Big Pine Tree site has substantial depth for a site in the Coastal
Plain of . South Carolina, up to ameter in places. .The few diagnostics
present indicate Woodland, Middle Archaic to Early Archaic";'Paleo-Indian-
occupations. In a general way, the characteristics of the debi tage also
reflect a stratified pattern. Levels wi thin the upper 45 cm possess
bright, fresh, unweathered flakes. The 45-75 cm zone has somewhat faded or
weathered flakes which are brightly colored from thermal alteration. Pres-
ent in this level, but really predominantly in the lower parts of the site
(75-100 cm), are weathered, cream-colored flakes. The weathered nature of
these flakes matches the condition of Paleo-Indian and Early Archaic points
found made from Coastal Plain chert. No sterile flood-sealed layers were
encountered in the auger tests and it is likely that the entire 12,000
years of human occupation are compressed within a meter of alluvium.
Nevertheless, because of this depth and the density of artifacts present,
and the relationship of the site to quarry activity, further excavations
would be rewarding.
38AL135
This is one of the sites that caused our strong interest in searching
for new quarries and quarry-related sites in the Allendale County area. In
July 1982, this si te was discovered by Tommy Charles and Gary Mahle of
Hampton, South Carolina, while exploring Smiths Lake Creek by boat. At
that time, it was obvious to them that a stratified ceramic and preceramic
si te existed in the floodplain. Smi ths Lake, which is fed by 3miths a ~ e
Creek and was probably formed by a cutoff of the Savannah River (Fig. 8),
has washed away an unknown portion of the eastern bank exposing artifacts
in situ. At the time of their visit, two distinct layers of cuItural
material were recognized (Fig. 13).
The upper zone, 30-80 cm below ground surface (Fig. 13), possesses
dark brown silty clay sediments with pottery and brightly colored heat
treated flakes extruding from the profile. A highly distinct lower stratum
is present below this (Fig. 13) which begins about a meter below ground
surface and about 20 cm in thickness. The sediments of this layer are
coarse sands of a light brown color. Preservation of alluvial deposition
is remarkably good as the microlenses of sand are still evident (Fig. 14).
Clusters of flakes can be seen lying flat, also indicating minimal post-
cuItural disturbances. This lower zone is characterized by weathered
white, yellow and cream-colored flakes which are so densely deposited (Fig.
14) they could be seen from the boat. About 15 cm below the weathered
flake zone is a sterile gray clay with red mottles that can be seen any-
where along the bluff where the exposure is deep enough. It was obvious at
the time that a reasonably deep stratified site was present.
At the time of their visit, the water level of the Savannah River was
down and a beach area was exposed. Several flakes, a few sherds, and other
lithic debris were observable along an approximately 50-meter strip which
63
Figure 13. Profile showing natural stratigraphy from eroded
bank at 38AL135. Trowel points to clusters of flakes lying
in lower most zone of site.
64
o
D
D
IB
I

1[').; .. 1.
II
o
Figure 14. Closeup view of lowermost zone in bank profile
of 38AL135. Note the thin micro1enses of al1uvially
deposited sand and weathered flakes lying flat.
had been washed out of the bank. Very little pottery was present. The
quantity and condition of the chipped stone indicated a high quality chert
resource close by. Of particular significance was the recovery of a
Paleo-Indian point preform. Its highly weathered condition matched the
chert material in the bottom layer of the bank. This piece appears to be a
preform broken during percussion thinning and shaping. A substantial
flute-like basal thinning flake (25 mm) has been removed from one face
(Fig. 15). The reverse face has not been appreciably thinned as it still
retains most of the original surface of the central face of the preform
flake. On technological grounds, this preform could have been intended as
a Clovis, Suwannee or Dalton point.
Based on this information, it was decided to include this site as part
of the chert quarry survey of 1983-1984. In August 1983, Albert C. Good-
year and Tommy Charles visited the site. At that time, the most obvious
archeological manifestation along the bank was the discrete bottommost
layer of weathered flakes. It was clear that a potentially large amount of
the earliest layer may have been washed away by the rising and lowering of
Smiths Lake due to fluctuations of the Savannah River. Subsurface testing
in the floodplain away from the bank was planned for the following winter
survey. Underwater exploration was undertaken in November 1983.
Insti tute underwater archeologists Alan Albright and Ralph Wilbanks
dived in Smiths Lake Creek in front of the site for about two hours. The
maximum depth of the water was only about 4.5 meters and visibility was
poor, less than a meter before silt and organic matter was disturbed. By
65
a
9
o ~ __==_... = = _ ~ 7 IN.
a 5 eM.
Figure 15. Weathered, light colored unifacial and bifacial artifacts
presumed washed out from the lowermost zone on the beach of 38AL135.
66
feeling in the sediments, a number of flakes, cores, and sherds were found
throughout all areas of the creek bottom. From this tactile method of
survey, it was clear that a substantial number of artifacts had been washed
into the creek. Imports.ntly, boulders of rock could be felt toward the
western shore which appears to be the old channel bottom of Smiths Lake
Creek, probably before it was captured by the Savannah River. Some of
these boulders are likely to be chert as asubstanial boulder of chert was
. recovered which exhibited flake scars. Several hand-sized cobbles of
unworked chert were also found in the creek indicating the presence of a
source. This underwater exploration confirms that a chert source for the,
occupants of the floodplain was immediately in front of the site, a fact
that would explain the great densi ty of flakes and chunks, many of which
possess river-smoothed cortex on their exterior surfaces.
In January and February, 1984, the site was subsurface tested by Good-
year and Charles using a bucket auger, a backhoe and hand-excavated two-
meter test pits. Figure 16 shows the location of these tests and the
topography of the site.
Auger tests 1-3 were dug first to determine how far the bottom layer
extended back from the bluff. The stratigraphic pictures revealed in auger
test 1 was similar to the profile exposed on the beach. It was also
located nearest the bluff (Fig. 16). The upper 30 cm were sterile consist-
ing of dark brown mottled clay. From 30-45 cm, three sherds were found
(Table 32), two of which appear to be curvilinear complicated stamped,
although the decoration is faint. It is suggested that this zone is the
late prehistoric ground surface prior to historic floods which are known to
have covered Mississippian sites along the Savannah River (De Baillou
1965). The next artifact-bearing level was the 60-75 cm zone which pro-
duced three sherds, two of which are fiber tempered (Table 32). At the 85-
95 cm level, light brown sand was encountered along with two weathered
flakes. Above this zone all sediments were dark in color and rich in clay.
This zone and the levels below until the depth of 120 em (Table 32) had
coarse light brown sandy sediments and weathered flakes paralleling the
bottom layer observed on the beach profile.
Auger test 2 and the other auger tests had very little in them and no
tables are presented. A single brightly colored heat treated flake was
found between 75-85 cm in auger test 2 and a nearly whole ovoid biface
blank (Fig. 17a) was found at 94 cm below surface. Two weathered flakes
were found in the 85-95 cm level. The heat treated flake was found in
clayish sediments, but the biface, which is light in color from weathering,
and the two weathered flakes were found in the light brown sandy clay.
Again, these weathered pieces, in terms of depth and the type of sediment
in which they were found, correspond to the artifacts found in the bottom
layer visible on the beach.
Auger test 3 was dug about 25 meters back from the bluff in an old
logging roadbed (Fig. 16). It was dug to find the bottom layer. We began
sifting at about 75 cm below surface in order to be somewhat above the
expected zone. The upper 75 cm consisted of silty clay sediments like
auger tests 1 and 2. Between 75-90 cm sandy soil was encountered along
wi th two yellow weathered flakes. This sand continued through 125 cm. A
single bright glossy white flake was also found similar to the chips found
67
.....------------
...

... --

./
/ -----
_..
L A
,+ D

/
OATUM A
N200 I I I REBAR .
g: b 1.22m length
- 0
II: '"
II:
T H '5
___ BACKHOE

5 M
eAT7
-_.. ------
o _..:==1? M.,... '-""" -------------.J
-----.. A. C. GOODYEAR a 1. CHARLES'
e AUGER TEST
..,---------
I:Ej
.....
Otl
n-S::
(J)
Ii
UJ (J)
n-
.....

II>
n-1-3
0
'-'Pd
0>0

'->l'tj
Vl::r'
. .....
()
0\
0>
'tj
P-
I-'
0
()
II>
n-
.....
0
l:l
UJ
0
1-1)
UJ
s::
a-
UJ
s::
Ii
1-1).
II>
()
(J)
eAT5
"-"""',"i'.,,'i!ii."i';'f' ----,-- ", ".""...,.
D
D
D
U
D
~
~
~
~
I ~ .. i...
IlJ
I ~ .. '
ow
I
~
Irnl
II
~
lI.mij
WJ
~
D
o
Figure 17. Weathered, light colored unifacial and bifacial artifacts
recovered from lower most zone of 38AL135 by excavation.
69
TABLE 32
VERTICAL DISTRIBUTION OF ARTIFACTS FROM AUGER 1, 38AL135
Key - a. Curvilinear complicated stamped (?)(2); fine sand tempered plain (1)
b. Fiber tempered (2); fine sand tempered plain (1).
c. Found in 85-95cm zone
d. Found in 95-105cm zone
e. Found in 105-120cm zone
I 'NXS&,iiii'1"AYil.."\4M"',,:,"!t,,W}!tp;:, ,:::R.L.Uc ,::,#pW;;S'&2"',,;:;;J;:"'J'pY})fE .' ,)531% Pitt@!,1F:\' iM'Lf:, i"!!M." * iL', ,P j!,'t'!'t"- ;",1 '''f!j''J'''iI',;;;"'Iffii Si\ljiL ifsKf':'j,:t- 3btiUfr,_" 1,i,!J,' "'!""fI? ".H
rii.t ..
i
.
U
1nI ..... i.
ilill
in the ZOne above the oldest layer. Nothing was found again until the
100-110 cm level where a single weathered flake was found. In the final
level (110-125 cm), six small weathered flakes were found in very coarse
sand.
Based on these auger tests it was clear that the basic stratigraphic
picture displayed on the beach profile was present wi thin a substantial
portion of the floodplain. The excavation of two-meter squares was then
planned in order to gather more culturally diagnostic data from a larger
area with tighter vertical control.
Using a transit, the site was placed in a metric grid system. Datum A
was a stick of iron reinforcing bar, one-half inch in diameter and 48
inches in length, driven nearly its full length into the ground. Nails
with heads one-half inch in diameter were driven in oak trees along a line
running grid north and south which intercepts datum A (Fig. 16). From this
baseline survey stakes were shot in using a transit and metric tape.
The excavation design called for three two-meter squares to be exca-
vated inside a larger hole with at least a meter removed on its perimeter
to eliminate contamination from higher levels (Fig. 18). This larger hole
was excavated to a depth of 75 cm by a backhoe. The Sandoz-Martin works
graciously loaned us the use of this machine and an operator for a day to
prepare the pits for excavation. The bucket of the backhoe had teeth which
left ragged cuts on the sides and bottom of the excavation pits, but, none-
Figure 18. Test pit N205, R197 at 38AL135. Photograph shows
placement of two meter square within backhoe pit. Bottom of
square is 97 cm below surface.
71
theless, this procedure was the one thing that allowed us to hand excavate'
at the lower levels of interest in the short amount of time and .labor. The
three two-meter pits were excavated on a Friday and Saturday using volun-
teer help.
Squares N205, Rl97 and N21 3, Rl97 (Fig. 16) were begun first. Of
these, only squareN205, R197 proved profitable. Square N209, R197 was dug
on the second day and finished.
Square N213, Rl97 was initially excavated by the backhoe to a depth of
80-90 em below ground surface which was a bit deeper than we would have'
preferred. It was excavated for nearly a full day by Glen Hanson and his
archeological team from the Savannah Hiver Plant. It took that long to
excavate a single 10-cm level because of the high clay content of the soil.
Light brown sand was present at this depth in patches like the sand of a
comparable depth on the bank profile, but generally the soil was highly
disturbed. A large pine tree stump was removed by the backhoe over this
square whose tap root went deep into the ground. This kind of disturbance
was no doubt constant over the floodplain for thousands of years. A;single
thin biface fragment was found between 80-90 em below surface that is pink
from heat treating but highly faded from weathering. In terms of depth, it
matches the lowest level of the site. Technologically, it is undiagnostic
except to note the faded condition from weathering.
Square N205, R197 yielded the most stratigraphic information. The
.backhoe removed the upper portion of this square to a depth of 60-65 cm
below surface. At this point a 10-cm level in the southwest quad was exca-
vated and screened through 1/4-inch wire. A small triangular point was
found in this level (Fig. 19a) as well as a fragment of a drill. The
flakes were dark red and brown, many thermally altered, and they presented
a fresh, unweathered appearance. . The remaining quads at this level were
shovel skimmed and not sifted because of the high clay content and lack of
time. The character of the flakes was identical to the southwest quad. In
the northwest quadrant several pieces of a fire exploded red and blue core
were found. A glossy, pink-purple biface preform midsection was found in
situ at 75 em below surface.
The next 10-cm level was dug as a single two-meter unit. It was
shovel skimmed and completely sifted. The flakes from this level were
identical in color and thermal alteration to that above. Dignostic arti-
facts include four small thin sand tempered plain sherds (intrusive?); a
piece of volcanic argillite which may be debitage; two pieces of a bifacial
blank, highly vitreous and colored purple from thermal alteration; the base
of a heat treated preform which is Guilford-like (Fig. 19c), found 82 em
below surface; another base of a heat treated Guilford-like preform found
83 cm below surface (Fig. 19d); another base ofa Guilford-like preform,
not apparently thermally altered, at 84 em (Fig. 1ge). At the beginning of
the next 10-cm level, an edge fragment of a bright red, thermally altered
biface blank was found, precisely at 88 cm below surface. A significant
stratigraphic clustering of bifaces can be noted in this square, a total of
five which occurred in a 13-cm vertical band. With one possible exception,
these bifaces all share the trait of obvious thermal alteration. Three of
these (Fig. 19c-e) have a GUilford-like appearance with rounded bases.
72
73
~ iiiiiiiiiiii_===:iiiiiiiiiiiiiii.-C:::::=iiiiiiiiiiiiiiiiiiiii?f IN.
a 5 eM.
Lithic artifacts from excavation of N205, R197 at 38AL135.
a
Figure 19.
In the next 10-cm level, several flakes were recovered by shovel skim-
ming and screened through 1/4-inch mesh, but no diagnostic artifacts were
recovered. A single retouched flake with a serrated edge made by half-moon
fractures was found, but no other tools were found in the 71 flakes recov-
ered. At least half were noticeably weathered while the remaining flakes
were fresh and dark colored like those in above levels.
In the next 10-cm level, 95":'105 cmbelow surface,' the practice of
shovel skimming and screening was continued. A number of flakes (N';'81)
were found in this manner, most of which were noticeably weathered. A
single chipped stone tool, a graver, and a winged bannerstone fragment made'
of steatite were all the diagnotic artifacts found. The atlatl weight
fragment (Fig. 19f) was found at 104 cmbelow surface and is clearly intru-
sive at this depth. This style is common during the Late Archaic period
between 3,000 and 4,000 years ago (cf. Claflin 1931: 29). The graver
spurred tool (Fig. 199), typical of Paleo-Indian and Early Archaic tool
ki ts of North America, is unusually massive. The spur, formed by the
placement of two unifacially retouched notches, is quite thick, being 6.7
rom at the thickest point. It is discolored white from weathering, but the
chert still retains its silica.
In the 105-115 cm level, two highly weathered artifacts were found.
The first is a utilized flake (Fig. 19h) with one edge heavily damaged from
contact with a hard surface. The second artifact is an edge of a biface
blank or preform (Fig. 19i). Both are discolored, having a creamy yellow
exterior but a chalky white interior. Both have a mild accumulation of
iron material like the biface from the bottom level of site 38AL23 (Fig.
26k). Only 14 flakes were found in this level, some of which were brightly
colored and fresh looking, indicating intrusions from higher levels. Most
of the colorful flakes came from the southeastern quadrant indicating a
localized disturbance. The bottom half of this level was not screened
because practically nothing was found by shovel skimming or sifting.
Square N209, Rl97 (Fig. 16) was dug on the second day, Saturday, using
volunteer labor. It was 'bElgun at about 85 cm below surface in a yellow
sandy clay. The yellow sandy soil matched that of the lowest cultural
layer on the beach profile. A unit 15 cm in thickness was first excavated
wi th all soil screened. This level yielded 47 flakes and a quartz cobble
that appears to have been used as a hammer. The flakes were all moderately
weathered. Some exhibited red coloring from thermal alteration, although
the red was strongly faded from weathering. The remaining 10 cm of this
northern half of the square were shovel skimmed down to a depth of 118 cm
below surface. Only two small flakes were found. Auger test 2 was located
on the northern edge of this square. Interestingly, the only chipped stone
artifact found in this pi t was obtained by the auger test where an ovoid
biface (Fig. 17a) was found 94 em below surface.
In order to determine more about the lateral extent of the bottom
layer, six more auger tests (AT 4-9) were excavated (Fig. 16). The sedi-
ments from these were not sifted until about 60 cm in order to save time
and excavate more holes. All of these produced weathered flakes in the
lower levels and the general stratigraphic profile seen on the bank and in
the previous auger tests was replicated. Auger tests 6-8, located close to
the bluff, indicate that the site may be deeper here as flakes were found
74
between 130 and 140 cm. The presence of coarse tan or light brown sand was
diagnostic of a high presence of weathered flakes. The terrace rises up as
the creek is approached (Fig. 16), which is natural since this is a levee
created by overbank flood deposits.
The overall picture for the stratigraphic condition of site 38AL135
seems fairly clear. There is a substantial amount of disturbance from
trees' that became obvious from the excavations. Clay soil as well as
brightly colored flakes was often mixed with sand at the lower levels.
Most of the larger artifacts were not lying flat. The density of artifacts;
from the bottom layer is greater by far in the zone exposed along the bank
profile. Nowhere in the auger or two-meter tests was such a density en-
countered. Given the richness of material in Smiths Lake Creek as found by
the divers, it seems probable that much of the Early Holocene occupations
have been washed into the creek. The low density of artifacts in the two-
meter squares was somewhat disappointing. Nonetheless, there is clear
stratigraphy displayed in the soil changes as the darker, silty-clay sedi-
ments existed from the ground surface to about 80 or 90 cm and sandier
sediments were typically found below that.. The artifacts found in each of
these two soil types were also distinctive. Brightly colored, fresh look-
ing flakes along with heated treated thick bifaces were found between 75
and 90 cm. Below that flakes and tools were usually faded from weathering
to cream and yellow shades. This discoloration and weathering is often.
found on Paleo-Indian lanceolates and Early Archaic notched points in
Gerogia and South Carolina made from Coastal Plain chert (Kelly 1938, 1954;
Michie 1977).
Using data gathered in situ from the beach profile and excavations,
coupled with surface collected pieces from the beach, at least three
temporally and technologically distinct lithic expressions can be recon-
structed for this site.
The first and oldest of these is represented by the weathered chert
material in the lowest layer. Eight worked specimens were recovered from
this zone (Fig. 17a-h). The ovoid biface found in auger test 2 (Fig. 17a)
and a crudely made side scraper found between 100 and 110 cm below surface
in auger test 6 (Fig. 17b) are the only two excavated. The remaining items
were observed in situ and removed from the beach profile. All of these
artifacts exhibit the faded discolored condition related to weathering.
The bifaces (Fig. 17c-f) are thick ovoid-shaped pieces with steep edge
angles on their perimeter. These scars serve more to shape than thin the
biface. A simple unifacial (Fig. 17g) tool and a utilized flake (Fig. 17h)
complete the list of worked material recovered from the bottom layer.
Materials collected from the beach which can be argued to have washed
out of the lowest zone are illustrated in Figure 15. These pieces all
exhibit the characteristic weathering and also are similar in .. terms of
flaking patterns to those artifacts found in the bottom layer. The Paleo-
Indian point prefonn (Fig. 15) has already been discussed. Three well:"
executed unifacially retouched tools were also found. These include two
endscrapers (Fig.15b, c) and a pointed oval uniface (Fig. 15d). The thick,
steeply chipped, but well-made unifaces are diagnostic of Paleo-Indian and
Early Archaic assemblages near chert quarries on the southeastern U. S.
Coastal Plain (Kelly 1938: 6-8; cf. Daniel and Wisenbaker 1983: Fig. 6).
75
The other examples of these unifaces were found in the lower levels of
38AL143 and 38AL23. They are relatively common in private collections in
the Allendale-Hampton County areas as well. The bifaces illustrated in
Figure 15 (e-h) appear to be thick and oval shaped. The rounded base and
relatively thick nature of these bifaces reminds one of stout Palmer
corner-notched points or Edgefield scrapers, although this is simply
speculation based on impressions. It can be noted that they are not as
weathered as the lanceolate preform (Fig. 15a) which is chalky white. The-
matter of thermal alteration for these pieces is not easily settled because
of their faded color due to weathering.
The second distinct technological manifestation refers to the cluster
ofmarkedly heat treated bifaces from the dark silty clay zone above in the
bottom layer. These bifaces were discussed when describing material from
N205, R197 and are in Figure 19 (b-e). Three'more of these
were found in situ in the beach profile in the silty clay zone (Fig.
20e-g). One of these has a rounded Guilford-like base (Fig. 20f). Four
were found on the beach (Fig. 20i-l), two of which have a square, base. All
specimens (Fig. 20e-l), including the base of a side notched point (Fig.
20h), are heat treated. They are only lightly faded from weathering,
except k which is fairly faded. The square and rounded bases coupled with
the pronounced thermal alteration argue for a post Morrow Mountain-Middle
Archaic placement. Savannah River points made from Coastal Plain chert are
usually not heat treated (Anderson, Lee, and Parler 1979: 53-54), at least'
as we can detect macroscopically.
Based on basal shapes, this technological cluster appears to represent
bifaces intended to be Guilford and Brier Creeklanceolate forms (Michie
1968; Brockington 1971). It is worth noting that the highest incidence for
heat treated hafted bifaces in the Anderson, Lee and Parler (1979: Table
10) study was for Brier Creek (85.7%) and Guilford lanceolate (80.8%). The
problematic issue of a late Middle Archaic side notched point occurring in
the southern Coastal piain of S. C. arises with the recovery of the basal
element pictured in Figure 20(h). This piece was found in situ in the
darker silty clay sediments 77 cm below surface. It is heat treated, faded
whi te but still glossy. It is the stem of a broadly side notched point
with a relatively thick stem (6 mm). The notches are sharp and the base is
dulled from fine battering. It is like the thermally altered "Type L"
described by Charles (1981: 31) on the southern Coastal Plain of South
Carolina. This specimen is virtually identical to the side notched point
found at 58 em at 38AL23 (Fig. 26h). The condition of thermal al teration
and its depth of 77 cm would suggest that more than just Guilford lance-
lates and Brier Creek stemmed points make up the late Middle Archaic clus-
ter. It is also important to note that the third highest incidence of
thermal alteration found by Anderson Lee and Parlor (1979: Table 10) was
their category "Kirk stemmed" (75.8%). We suggest that many of the heat
treated notched points of the Coastal Plain typed as "Kirk", are in fact a
late Middle Archaic side notched point which may be related to the Halifax
side notched (South 1959; Coe 1964).
Before leaVing this technological cluster it should be mentioned that
much could be learned from an expanded excavation and technological analy-
sis of this horizon. Careful excavation beginning at a higher level could
verify the temporal position of the point with broad side notches. All of
76
Specimen Figure 20. Lithic artifacts recovered from the beach (38AL135).
h found in situ in the bank at 77 cm below surface.
o
o
U

10
10
ID
lu
IJ
I]
I]
11
(]
D
[]
[j
[j
[I
[]
77
the post Morrow Mountain phase-preSavannah River phase hafted biface types
made of Coastal Plain chert seem to be characterized by an emphasis on
thermal alteration (cf. Anderson, Lee and Parler 1979: Table 10). Some of
the broken blanks or preforms found at 38AL135 reveal that large flakes
were thermally altered and then bifacially thinned and shaped. Some of the
ventral faces are dull and exist as islands surrounded by glossy, vitreous
flake scars (cf. Crabtree and Butler 1964). More data need to be gathered
~ t quarry and lithic processing sites such as 38AL135 and 38AL23to further
our understanding of technological systems where heat treatment was
apparently so critical.
The third technological pattern is illustrated by the first four
bifaces in Figure 20 (a-d). The first three are well-thinned and well-
shaped bifaces that appear to be preforms for Savannah River stemmed on the
smaller Early Woodland stemmed forms (Fig. 20d) variously called Otarre,
"Stallings Type 3" and Gary (Goodyear, House, and Ackerly, 1979). The
chert is not weathered and they do not seem to be heat treated as the
coloring is a jasper brown typical of chert cobbles found in the rivers.
They are somewhat glossy, but not as much as obViously heated examples
i g ~ 20e-l). These pieces, including the stemmed point (Fig. 20d) , were
found on the beach and probably are derived from the upper 50 em of the
siteo It is interesting that two of the three preforms (Fig. 20a, b) do
not appear heat treated and the remaining pieces (Fig. 20c, d) are ques..:
tionable as to thermal alteration. Anderson, Lee and Parler (1979: 54)
have noted a decrease in thermal alteration of Late Archaic hafted bifaces.
The dark brown color of specimen a-c (Fig. 20) is similar to the chert
found in river cobbles. ..
Summary
In summary, based on various subsurface testing activities, the site
is clearly stratified. Although obviously disturbed in the upper meter by
trees and animals, a basic artifact stratigraphy appears to have survived.
The suggested late Middle Archaic occupation would merit more study. The
lowest occupational horizon, most evident on the bank profile, is, however,
lightly represented throughout the areas of the floodplain tested. It is
denser and better preserved near the bank as evidenced by the thick dis-
crete layer of flakes lying flat and the thin depositional lenses of sand
which have survived natural disturbances over 10,000 years. As previously
stated, it may be that the bulk of the Late Pleistocene-Early Holocene
deposit has washed into Smiths Lake. As demonstrated by the underwater
archeological reconnaissance, the chert source for the si te is the creek
bed itself. Recovery of early man artifacts from the terrace would be a
next logical step at 38AL135 in order to generate a substantial collection
of stone tools and waste products for further study. The site could help
establish a "typological datum" representing lithic reduction and
production behavior by early groups in the vicinity of the raw material
source, against which the variability in lithic technologies away from the
source could be compared.
78
D
D
38AL142 White Woman's Point
This is an outcrop of chert boulders eroding out of both sides of a
dirt roadbed. Flakes and chunks of chert can be found along the road (Fig.
21) for about 50 meters leading down to the Savannah River to a boat land-
ing. An artesian well which has been tapped with a pipe and a cement
catchment is present near the landing. The quality of lithic raw material
represented here for the most part is poor. Some limestone boulders that
we broke open were barely silicified (Fig. 22). Relative to some of the
other quarries, this site has less debris from extracting the raw material.
Flakes and cores of high quality Flint River chert can be found along the
road, however, especially on the floodplain, suggesting that good quality
chert was available here or nearby. The terrain around the road is rather
steep (Fig. 23), promoting erosion, a situation commonplace among other
quarries to the north along the bluff system. No diagnostic artifacts were
observed from this site and no testing was undertaken because of the low
density of chippage and the eroded nature of the terrain.
Figure 21. Bricks and chert
debris in the roadbed at
38AL142, Whi te Woman's
Point site.
o
79
Figure 22. Large chert boulder at 38AL142, White Woman's
Point site.
A sample of a better grade of chert was analyzed by Upchurch (Appendix
A). He classifies it as Category 7d, "light brown, silicified grainstone."
It is an example of "Allendale chert," and this quarry is a member of the
Allendale Chert Quarry Cluster as formulated by Upchurch (Appendix A).
38AL139 The Topper Quarry
This si te was originally shown to us by David Topper, of Fairfax,
South Carolina, in June 1980 and is one of two Coastal Plain chert quarries
alluded to by Charles (1981: 14). At that time we were impressed with the
size of the quarry and the quality and abundance of the lithic material.
Chert boulders, cores and numerous flakes could be seen on a dirtroad that
runs on the north side of a small unnamed creek (Fig. 23). The creek runs
noisily over chert and limestone outcrops and no doubt was a primary expo-
sure from which aboriginals could gather chert (Fig. 24). .
This quarry was revisited by our surveying team in July 1983 and
February 1984. During the winter it was possible to see more clearly the
true extent of the chert outcrop. Basically, boulders, split chunks of
chert and cortex, flakes and cores can be found on both sides of the creek
over an area about 150 meters wide. The length of quarry activity as it
follows the creek is approximately 250 meters (Fig. 23).
No testing was done within the qurry. It was felt that there was too
much rock to learn profitably about core reduction patterns through time in
the midst of the source. The site immediately below here, however, 38AL23
80
oI
------
o
D
o
o
o
RIVeR
Figure 23.
o

fI':"0 ' ----
I

.._ .......
"
,", -,
, \ "
I \ ,\
\ \ '\
\ \ "
, \. \ \
,. I "
\ , .,
I .... \ " . ,
, .... \ I \
, .... \'-4 , \
'( \ \ , \
, \ \ \ ,
I " \ \
\ '\ \ \
\ \ \ \ \
\ \ \ \ \
, ,\ \ \
\ \' , \
, ,\ \ \
\ \ \ \ \
\ \ \ \ \
\ \, \ \
'-' \ .\ \
.\, \ \ \
I \ \ \ \
I \ \
\ \,........... -
... __.l.-- - 291i)===;=.1
\
\
\
I
\
I
;' I
I I
I \
: I
V
I _--:-.--------1,
/ ---- ,
;' . ,
,\
I \
I \
,
o 500 1000 FEET
i , , "
o 100 200 300 METERS
Locator map for sites 38AL142, 38AL139, 38AL23 and 38AL141.
81
(Fig. 23), has excellent potential to inform about quarry utilization
through time. The chert from 38AL139 is unusual compared to other Flint
River formation outcrops. It tends to be white, translucent, gray or amber
when freshly broken. It is of an extremely fine-grained quality which
explains the extensive evidence of its exploitation.
Figure 24. Small unnamed creek filled with chert boulders and
debris running through 38AL139, the Topper Quarry.
Rock samples were collected for petrologic analysis. Upchurch (Appen-
dix A) classifies a sample from here as category 7d, "light brown, silici-
fied grainstone." This material, like all of the other categories of
silicified grainstone, are examples of the Flint River Formation chert or
what has been called "Allendale Chert" (Upchurch, Appendix A).
38AL23
This site was originally recorded by Dr. Don Sutherland in 1973. At
that time he reported in the Institute site files that it was a "habita-
tion" site on a "sandy bluff overlooking Can) oxbow lake remnant of the
Savannah River channel" (Fig. 23). He observed flakes and a single undes-
cribed sherd in the dirtroad leading to the bluff. No collection was made.
Immediately adjacent to and east of this bluff, which appears to be an
alluvial terrace, is 38AL139, the Topper Quarry (Fig. 23). The distinction
between sites 38AL23 and 38AL139 is based on physiography and probable site
function, although in reality, artifacts are continuously distributed
between the two. Si te 38AL139 refers to the upland hill area which con-
82
o
o
D
U
U
U
D
U
U
II
II
D
D
o
D
D
D
o
tains exposures of chert exhibi ting extensive evidence of extraction by
aboriginal groups. An apparent alluvial terrace (38AL23) has been formed
on the river side of the hill which manifests considerable occupation by
ceramic and preceramic period groups.
An uncontrolled surface collection was made from the dirtroad and a
small field where peas were formerly planted. -Only temporally diagnostic
sherdsand bifaces were recovered. These artifacts indicate occupation
over the past 3,000 years. Decorated ceramics include Deptford linear check
stamped (Fig. 25a), Deptford check stamped (Fig. 25d), Thom's Creek simple
stamped (dowel impressed) (Phelps 1968: 21) (25 d), and an unknown sherd of
early curvilinear complicated stamped (Fig. 25b). Late hafted bifaces
found include a triangular arrowpoint (Fig. 26a) and a probable late Wood-
land stemmed point (Fig. 26b).
It was clear to us, given the closeness of the secondary river channel
and the sandy composition of the bluff, that a great potential for overbank
deposi tion was present, hence, a buried site. An area of the bluff over-
looking the secondary channel was chosen for auger testing and excavating" a
test square (Fig. 23).
Auger test 1 was dug to a depth between 1. 5 and 1.7 meters. Flakes
ceased to occur at about 90 cm, as lower levels only produced a few flakes
(Table 33). At about 1.5 meters- brown linear stains (lamellae) appeared
followed by wet gray clay. Given- the relatively deep nature of what we
presume to be alluvial deposition and nearly a meter of artifactual depos-
it, a 1 x 2 meter square was excavated adjacent to auger test 1 (Fig. 23).
The test pit was excavated over a two-day period with the help of
Frank and Andee Steen, members of the Archeological Society of South Caro-
lina. The soil was nearly pure sand of a coarse nature that sifted easily
through the quarter-inch screen. No features were seen. It is likely
that, if present, they would have been leached out in the porous soil. Some
vertical changes in soil color were noted (Fig. 27) as well as patterns in
the vertical distribution of artifacts (Table 34). The soil color changes
appear to be related to B horizon formation of sandy humid soils. No dis-
cernible differences in texture were noted during excavation or in profile
indicative of discrete alluvial deposits.
From 0 to 15 cm the soil was strongly organic related to contemporary
root systems. Sherds, flakes, and quartz cobble fragments were present
-(Table 34). - Three sherds were decorated including two Refuge simple
stamped (Fig. 25g) and one rectilinear complicated stamped (Fig. 25c). Two
ceramic fragments of what may be ei ther a pipe or small bowl were also
present (Fig. 25f). These, coupled with the complicated stamped sherd and
the small triangular arrowpoint (Fig. 26a) found on the surface, indicate a
late prehistoric occupation of the bluff. At about 10 cm, the dark organic
soil gave way to a tan color. This tan, loose sand had a substantial Early
Woodland Refuge phase component which was present to about 30 cm (Table
34). Of the 57 sherds found in the 15 to 30 cm level, 31 were Refuge sim-
ple stamped (Table 34; Fig. 25h, i). There was also a relatively large
number of bifaces and interesting flake tools found in this level, includ-
ing retouched and sidescrapers (Fig.26e, f). Although no
hafted bifaces were recovered, two examples of what may be triangular point
83
e
f
IN.
o 5 eM.
Figure 25. Pot sherds from surface and excavations at 38AL 23.
84
85
~ __=:::::i_'-iic::::=__i?f IN.
o 5 eM.
Lithic artifacts from surface and excavations at 38AL23.
a
Figure 26.
0'\ 0'\ t'- 0 0 t<'\ 0
T.i a8a.1aAV
t'- t<'\ t'- ll'\ 0 '0 t<\ 0 0'\
":JM 0 0 0 0 0 0 0 0
C\J
0 0
0'\ \0 ll'\ C\J 0 0 0 0 0 0'\
0 0 0 . . . . . . ..
o slIlf)
"T.i :JM Ta:J0.1 t'- 0 ll'\ t'- ll'\ C\J
t'- t<'\
0 t'- 0'\ C\J \0 ll'\ C\J t<'\ C\J .....
..1aqmnN t<\ 0'\ \0
salqqad z:J.1ano
C\J
t<'\
A.1a:J:Jd
C\J
H
<
CO
t<\

8
t1Ula.1d C/)
I'il
8
P::
aoal1=H pa:JlaH
I'il
C'
::J
<
::E:
T00.1 aTqqo::>
0
P::
I%<
t<\
C/)
a.1o::>
/
t<\ 8
0
I'il
<
H I%<

P=l H
<
8
8 P::
<
I%<

0
z
paz1=T1=:Jfi ........
0
.... H
'-' 8
::J
U"\ \0 \0 ll'\ C\J t<'\ ..... ..... ........ (1)
P=l
U1=q.r. t'- t<'\ 0'\ U"\
..... s:: H
'-' 0 P::
+" 8
+" to C/)
Ua1=:J.1a.1
H
..... ..... .....
(1) s::
A
.s:::: IU
o to
H
<
0
.100aa ouooas C\J ................
H
..........
8
'-''-' P::
I'il
>< >< >< :>
.100aa "m1=.1d
C\J ..... t<\ t<\ (1) (1) (1)
+"+"+"

000
000

IU .c
0 0
C\J C\J
t<\ t<'\ ........................
". ". ",
'-''-''-'
a.11=.i (1) (1) (1)
s:: s:: s::
000
+"+"+"
to to to
(1) (1) (1)
fa E! E!
'M'M'M
HHH
E! E! E! E!
E! 0 0 0 0 0 .
E! E! E! E! 0 0 ll'\ 0 ll'\ 1U.c 0
a 0 0 0 0 ll'\ C\J t<'\ ll'\ \0
0 ll'\ 0 ll'\ ll'\ 0
..... ..... ..... ..... I
0 \0 t'- 0'\ ..... I I I I
t<\ I I I I I ll'\ '0 ll'\ 0
I 0 ll'\ 0 ll'\ . ll'\ 0 C\J t<'\ ll'\ (1)
0 t<'\ "
\0 t'- 0'\ ....

86
[=:J ii::J e"J I::l EJ EJ IE::i E]. I t:1 ] i",",-) EJ .....".1 r::=J
TABLE 34
VERTICAL DISTRIBUTION OF ARTIFACTS FROM 1 X 2 M. TEST EXCAVATION, 38AL23
Ul Ul ell
Ql

Ul
() lIS
(I)
l5 0 (I) M Ul Ql r-l
p:; Ql r:t-l Ql
Ul r:t-l +J
.

Ql Ql
tit
"C
.
J.l lIS . M () M J.l M
Ql J.l 0 M M r:t-l M lIS .0 Ql r:t-l
0 ()
r:t-l 'r! r:t-l 0 ll-I .0
r-f
() () <U ...c: "C 0 'r! <U
f>:f
lIS <U 1:1 >. E-l "C
E-l j:Cl P-l

J.l 1:1 J.l <U <U
e
>. z <I.l
U Ul .
lIS Ql N
g
() <I.l 't:I ... N bO
]
.
or! ()
-r! lIS M <I.l 0
<I.l +J M M lIS
<U S 0 +J lIS M 0 tH <I.l ,.0 .j,J tH .j,J
J.l lIS lIS . ...
J.l -r!
()
tl
tH or!
.j,J
-r! J.l .0 tH <U .j,J
lIS +J .j,J
<I.l
'r! J.l <U .r! .j,J <I.l
8
0 0
J.l M 0 ::3 0 0
r:t-l U P-l
tJ)
E-l j:Q
P
p:; u u P-l
j:Q
P-l 0- E-l E-l
0-15cm 11 a 8 15 11 137 lb 3c 23d 4 171 181.3 1.06
15-30cm 15e 23 53 18 535 3 2 (2)f (2) 4 57g 17 629 614.1 .98
30-45cm 17h 5 16 200 (1)i
4 221 197.2 .89
CD
45-60cm 1(1)1 38 . 1,016 1.36
.......
70j 7 30 44 922 2 lk 5 1385.0
60-80cm 16m 16 58n 20 603. 10 1p lq 1 29 701 861.6 1.23
80-90cm 5r 5 27 2 240 18 1t 19 274 217.0 .79
90-100cm 15u 1 1 47
12 49 15.6 .32
100-110cm 10u 3 8 3 .5 .17
120-135cm
.... """"', ". ,., '''.''''''''''Af'''l ,... ".&:-: -, . ","'''*'M","" .,Q e'"",""":';::"?_ "." ,._._......,;" p;'e:!If''':!''''"' - ......""?""_.",'", "'''--'.,.' . ,., . "5" i""":"". ., .. "e_""",!:,,'::",?,," """ ._
= = . ~ = = = ....... = = ~ = = = = ~
KEY TO TABLE 34
<Xl
<Xl
Key - () - Fragments
a. Pieces of limestone (?) cortex. One piece of possible metavolcanic.
b. 34.4 gms.
c. Quartz cobbles
d. 18 plain fine sand tempered; 2 simple stamped; 1 rectilinear compo stamped; 2 pipe (?) bowl rims
e. Sandstone cortex (1); limestone (?); cortex (14)
f. Tips
g. 31 refuge simple stamped; 26 plain sand tempered
h. Limestone (?) (11) cortex; sandstone (6)
i. Tapered stem
j. Limestone (?) cortex (57); chert (13)
k. 127.7 gm.,heat treated
1. One whole shallow side notched point, thermally. altered; one tip
m. Limestone (?) cortex (12); chert (4) .
n. 48 have river cobble cortex
o. Drill tip on flake
p. Endscraper on chunk
q. 613.7 gm.
r. Limestone (?) cortex
s. 134.3 gm.
t. Highly weathered; iron stained
u. Limestone (?) cortex
o
preforms were found (Fig. 26c, d). This Early Woodland deposit seems
fairly distinct and restricted to this 15-em level since below 30 cm, no
more pottery of any sort was found (Table 34).
Figure 27. Profile of north wall of 1 x 2 meter pit at 38AL23.
The 30-45 cm level had a greatly reduced number of artifacts (Table
34). The soil was still colored a light tan from leaching of the humic A
horizon. This was also discerned while excavating the level. The only
potentially diagnostic artifact found was the stem of a heat-treated con-
tracting stemmed point (Fig. 26g). The sharply pointed character of the
stem reminds one of a Morrow Mountain, but the thickness (6.9 mm) is more
like the Early Woodland, gary-like, stemmed points found with fiber tem-
pered and Thoms Creek pottery at Stallings Island (Bullen &Green 1970) and
at the Mack site, 380R67, (A. Robert Parlor personal communication). This
piece is heated treated, however, a trait not typical of Early Woodland
contracting stemmed points.
The 45-60 cm level was the densist layer of the test (Table 34). Over
1,000 flakes, most of which were brightly colored from thermal alteration,
were recovered as well as five bifacial blanks, a core, and two utilized
flakes. The tip of a hafted biface was found as well as a heat treated
broadly side notched point (Fig. 26h). This latter specimen is bright pink
in color and only lightly faded from weathering. The piece is relatively
thick at the shoulders (9.0 mm) and not basally thinned. The base and
notches are not ground or polished but are dull from fine battering. Typo-
logically, we are not certain of this specimen's temporal position, but on
technological grounds it is more like the Halifax side notched (Coe 1964:
108-109) than the Kirk corner notched. It was found at exactly 58 cm below
89
surface which puts it somewhat below the midpoint in the stratigraphic
column of this 'test pit (Table 34). On technological and stratigraphic
grounds the piece has a probable late Middle Archaic position. As noted,
the stem of a nearly identical point was found in a late Middle Archaic
context at 38AL135. Numerous other specimens like this one have been seen
in the collections of private collectors in this region of the state. They
are distinctive in that they are brightly colored from thermal alteration
and only mildly faded from weath"ering.lt appears to be similar to Charles'
(1981: 31) "Type L" which he posted on technological ground to be of Early
to Middle Archaic in age. Heretofore, these heat treated side notched
points probably often have been called "Kirk" for a lack of better typol-
ogical controls on hafted bifaces on the Coastal Plain. At 60 cm, the
light tan stain resulting from humic leaching ceased. From here downward,
the profile was characterized by an off-white sand.
The next arbitrary level, 60-80 cm, possessed the second largest num-
ber of flakes (701) (Table 34), although it is a 20-cm level and with some
interesting flake tools. One of these is a drill or perforator (Fig. 26i)
made on a thin flake (3.8 mm). Macroscopically, this tool is not weathered
in the least and is, in this respect, unlike most of the debris from this
level and other parts of the site. A steeply retouched endscraper was
also found (Fig. 26j), made on a chunk of fairly weathered chert. The
thickness of this endscraper (12.5 mm) is reminiscent of other Early Holo-
cene uniface tools on the southeastern Coastal Plain. Anot.her natural
layer of debitage was observed, this time between 70-80 cm. These flakes,
weathered to a white color, were lying flat, in the soft white, easily
sifted sand.
The top of the 80-90 cm level exhibited a concentration of large white
flakes which gave way to smaller flakes. The sediments were a white,
easily sifted sand. The only unusual artifact from this depth (82 cm) was
a highly weathered and stained biface abandoned or broken in the early
stage of reduction (Fig 26k). The stain is a result of iron leaching
through the soil. Sand grains have also been cemented to one face while the
opposing face has a dark, smooth ferric accumulation. The low p o ~ t o of
this piece and its highly weathered condition are suggestive of Paleo-
Indian and Early Archaic bifaces. This staining is frequently observed on
Paleo-Indian and Early Archaic points found in the sandy soils of the
Coastal Plain.
The remaining levels (90-110 cm) have Iittle data beyond debi tage.
The deposit seems to terminate abruptly about 100 cm below surface (Table
34). Flake size, as estimated by mean flake weight drops to .32 and .17 gm
respectively (Table 34) which suggests very light small flakes that perhaps
may have been worked to these depths from above by bioturbation.
Auger tests 2 and 3 were dug nearly side by side in an area where the
terrace met the hillside (Fig. 27). It was thought that there might be
greater site depth here because of colluvial deposits originating from the
hillside combining with alluvium from the river. Other than debitage a few
artifacts were found in these holes (Tables 35-36). A thick, steeply
chipped endscraper made on a heat treated cortical flake (Fig. 261) was
found between 45-60 cm in auger test 2, and a thick, crude bifaee blank
(not heated) was in the 30-45 em layer in auger test 3. A single sherd was
90R7'iT'ET7---- --=zm:=m
[ i E:5
Ej
e:::i Ei E:::i Ei Ei I:::J Ei c:::J c::J r:::::l
E:::]
r::::J c::J
TABLE 35
VERTICAL DISTRIBUTION OF ARTIFACTS FROM AUGER TEST 2, 38AL23
t/)
Q)
t/)
t/)
Q)
l3
CJ (I) t/) t/)
0 t/)
r-I co Q) Cll
l3 CIJ IJ:l Q)
r-I
.

CIJ
co IJ:l ol.J
"d .
1-4 (I) . (I) r-I
CJ
Q)

Q)
1-4 0 r-f s:: r-f IJ:l r-f Cll
r-f 1-4 r-I
0 ()
IJ:l
:a
IJ:l
0 'l-I
,0 Q)
IJ:l
CJ CJ Q)
"Cl 0 ..-l
.g
J
r-I
(I) Q.I
A
1J:l. :>. E-l "Cl Q)

E-l jX:I
Pol ol.J 1-4 A Q)
Q.I
(.) t/) .
Cll Q.I til CJ () Q.I
"Cl e
l>.
Q)
.
1=1 orf ()
oM ::I (I)
til

Q.I
S 0 ol.J (I)
r-I Q.I
0
Q.I
ol.J r-I r-I (I)
r-f 0 'l-I Q.I ,0 +J 'l-I ol.J 1-4
.(1)
Cll
$-I orf () $-I 'l-I ..-l ol.J ..-l 1-4 .0 4-l Q.I (I) ol.J
1-4
orf
$-I Q.I Q.I orf ol.J
:S
ol.J ol.J Q)
0 0
pg
1-4 r-I 0 0 IJ:l (.)
Pol CI)
E-l jX:I
P C,) (.)
0
Pol jX:I
Pol CY E-l E-l
0-15cm
9 2 1 2
14 24.5 1.75
15-30cm
3 2 10
la 1 15 9.7 .65
30-45cm
3b
3 18 lc
22 42.5 1.93
\0 45-60cm
14 11
44 1 ld
69
.....
78.7 1.14
60-75cm
5e 7 15
43 lf 19
2 65 54.2 .83
75-90cm
5 5 1 17
1 28 28.2 1.01
90-105cm
6
7
1 13 10.7 .82
105-115cm
2
1 2 .5 .25
115-130cm
3e 1
1
2 1 .5 .50
130-145cm
le 1
4
3 5 1.5 .30
145-160cm
,
2e
3 1
4 22.7 5.66
160-175cm
2e
4
5
1
9 1.1 .12
175-190cm
4e
1
Key - a. Fine sand tempered plain
b. Limestone (?) cortex
c. Two utilized edges
d. Endscraper on cortical flake
e. Limestone (?) cortex
f. 144.0 gm; thermally altered
g. Quartz cobble fragment
.,.,.", ' ...p;* 'T--?&.?::+1;;,:h "'?9"fiU? 4fd n, ,?0&!"''" : .. L ' :w,J ... . T __,=.;t, "x_ , ..", .r,;"","::" ,..."."" "." ._ . *
TABLE 36
VERTICAL DISTRIBUTION OF ARTIFACTS FROM AUGER TEST 3, 38AL23
co
co co
4)
4)
(S
co
tJ
co (S
0
co .-I co 4)
r-i
p:;
4)
rx..
4)
co rx..
0

4) 4)

'0
.. $-I
0 r-i
() r-I $-I .-I
4) $-I 0 r-i s:: r-i rx..
.-I lIS
,..c 4)

0 tJ rx..
0,-1
rx..
0
.-I
() 4) .c "tJ
0 OM

l\S
4)
A :>.. E-i "tJ
E-i j:Q P-i

$-I A
$-I
4) 4)
e
>. z
4)
t..) en
0 l\S
4) t.;j () () 4) "C $-I t.;j bO
0
s::
or-! tJ or-! ::s lIS .-I
4) 0
4) r-i .-I lIS
4)

0 l\S .-I 0 4-1
Q) ,..c +J 4-1
+J $-I. lIS lIS $-I"

$-I
() $-I 4-1 or-! +J or-! $-I ,..c 4-1 4) lIS . +J
4)
OM ..c $-I
4) 4) or-! +J
s:: 0 0 lIS $-I r-i 0 ::l 0 0
t..) Pol CIl E-i

j:Q :::> :::> u u p:I P-i
j:Q P-i c:r E-i E-i
0-15cm 2a 4 5
2 9 10.4 1. .15
15-30cm 1 9
10 2.5 .25
30-45cm
1a 5 5 1 13
1b 1 24 9.8 .41
\0
45-60cm
N 7a 15 8 55
3 78 99.5 1.27
60-75cm 6a 24 5 1 46
5 76 68.5 .90
75-90cm 9 5
18
,
32 18.8 .59
90-105cm 2a 1 1 7
2 9 32 .35
105-120cm 80a 14 10
16 24 6.1 .25
120-135cm 79a 8 2
8 10 2.2 .22
135-150cm 85a 3
2 3
5 1.4 .28
150-165cm 4
1
5
1.0 .20
Key - a. Limestone (?) cortex
b. Whole; ovoid; not thermally altered; 114.5 gm
,,, ,," ,fli"@,.wy"":,:'.':'Ii!ilSi"if Y. 'ii,"". "1 ,i.lf.'l'ii\,' ;; .. ""'M't!',Iji"!,,,I' "';'-"""'"'1"'."'c':::":"7;i)''i'i''i''''''''''' .... .,.,,"",.....' 'i",,,,"""""" #3,i\kl'U_,"'''"'' <;a; , ""\AI. ,
found in auger test 2 (15-30 em), further indicating a high stratigraphic
position for pottery in this site.
The unusual thing about both of these tests is the great depth to
which flakes were found. Impressive numbers' of flakes were found well
below a meter in each test (Tables 35-36). In auger test 2, fo r example,
nine flakes were found between 160-175 em; in auger test 3, 10 flakes were
found between 120-135 cm,and another 10 were foUnd between 135-165 em. In
both tests, beginning at about 110 em a marked occurrence of limestone (?)
cortex chunks appears. These are small fragments and may indicate down-
slope colluvial movement of soil. This cortical material matches the'
boulders of chert up the hill in the quarry area (38AL139) Thus, the
northeastern end of the site appears to have substantial depth with the
promise of producing valuable stratigraphic information.
Auger test 4 was dug about 60 meters south of auger tests 2 and 3in a
small clearing (Fig. 23). It was dug to gather more information about the
lateral extent and depth of the site. As can be seen in Table 37, the
artifact density is very sparse. A 1ittle pottery was found in the upper
levels. The impression given from this test is that the site in this area
is about a meter deep.
Auger test 5 was dug on the extreme southern end of the terrace. The
terrace here is fairly low, with an elevation between the 70 and 80 foot
contour (Fig. 23). The soil tliroughout was a sticky, moist, silty clay
related to Savannah River floods. The upper 30 em was less clayish, but
still very difficult to screen. The 30-45 em zone was so clayish that the
soil was discarded. We auger tested to 75 em through red mottled clay,
chopping through the spoil for artifacts. None were seen (Table 38).
Cultural occupations on this portion of the terrace was probably light and
late in time.
Summary
In summary, site 38AL23, because of its integrity and deep strati-
graphic condition, exhibits the best potential for further scientific
research of all the quarry and quarry-related sites discovered in the sur-
vey. The natural conditons of high quality cryptocrystalline chert in
abundant and accessible quantities, coupled with an excellent freshwater
source which runs through the chert outcrop, no doubt made this an attrac-
tive place for Indian groups throughout time to visit. The well-drained
soft sandy soil of the terrace in front of the quarry, facing as it does a
segment of the Savannah River, no doubt further contributed to the desira-
bility of this location for working, habitation and gathering subsistence
resources. Although not verified by the present test excavations, there is
reason to believe the cultural sequence here is virtually complete from
Paleo-Indian through late prehistoric periods.
93
TABLE 37
VERTICAL DISTRIBUTION OF FROM AUGER TEST 4, 38AL23
(J)
(J) (J)
Q)
Q)

,!:o'.
,!:o'.
,.!lol (J) ()
ttl (J)
ttl
0 (J)
r-l (J) Q)
.-l
p:: Q)
Q)
(J)
. ,.!lol
Q)
Q)
:l: '0 .
J.I ttl .
r-l ()
r-l J.I .-l
Q)
I-l 0 .-l s:: r-l r-l ttl ,.a Q)
,.!lol 0 ()
oM 0 lj.j
.g
r-l
() () Q)
..c: '0 0 -.-I
ttl Q)
A E-t
Q)
E-t F'l
Pol J.I A J.I
..c:
Q)
e
z :l:
Q) u (J) .
ttl Q) to! () () Q)
'0 J.I to!
00
.
s:: -.-I () oM ::3 ttl r-l Q)
0
Q)
.-l r-l ttl
Q)
a 0 ttl .-l 0 lj.j Q) ,.a lj.j
I-l ttl ttl k J.I ::3 -.-I ()
I-l lj.j -.-I -.-I J.I ..0 lj.j Q)
til
Q) -.-I ..c: I-l Q) Q)

-.-I
s:: 0 0 ttl I-l r-I 0 0 0
U Pol C/)
E-t F'l P P U U 17: Pol F'l Pol 0- .E-t
,E-t
0-15cm
4
1a
4
5 .12
15-30em 1b 1
9
1 10 20.9 2.09
30-45em
1 10
2c 11 4.8 .44
\0
45-60cm
6, 1 2
4
13 7.5 .58
60-75em
1 8
9 15.0 1.67
75-90em 1b
3 1 7 1d
2 11 21.0 1.91
90-105 1b 1 1
3 2 .5 .25
105-120cm
120-135em
Key - a. Plain fine sand tempered
b. Limestone (1) cortex
c. Simple stamped; plain fine sand tempered
d. Fragment, 81.7 gm
";)t"') , '/"'.' .. " ,.1,'i'1\!"" , '" ,w""""'.'! .",;i'/!if' !JAIhI"iI',1!"W"" 4 , "",,",'*"""""'" '," "":;;;;;;;', ':"' ."""':, "!'E*",:A"' ...,*,,,",,,.",,",,, .,.. " . .o:"5\\li\1q::!if'f'::-"' . ""ffI'I",f ,.. "",""""..,,,," - m=;\"nM - O)JiitiJ ' \Ii Iif" }"IM % ##
38AL141 Johnsons Landing
This site is located on a steep hillside approximately 100 meters
south of Johnsons Landing, a public boat landing on the Savannah River
(Fig. 23). Chert boulders have been exposed by gradual erosion of the
bluff. A flowing artesian well which pours constantly into a cement trough
is in front ofa homesite on the bluff. The chert exposed in the bluff
(Fig. 28) does not exhibit evidence of exploitation, but flakes and sherds
can be found on the road which runs over the bluff. It is highly probable
that much of the chert quarried from the outcrop was procured from the
river at times of low water. Boulders of chert can be seen at the water's
edge. Rock samples were taken for petrologic analysis.
Figure 28. View of chert outcrop in bank of Savannah River
at 38AL141, Johnsons Landing.
In his study of Geology of the Coastal Plain of South Carolina (1936),
C. Wythe Cooke described the chert outcrop as follows:
"On the bluff on the Savannah River at Johnsons Landing,
almost due west of Allendale, great masses of chert lie loose or
are embedded in coarse reddish pebbly argillaceous sand. In the
fields north of the landing many small chunks of chert contain
Lepidocyclina ocalana, Bryozoa, and other fossils that indi-
cate the Ocala (Jackson) age of the rock" (p. 91).
We walked field roads north of the landing as described by Cooke and
found no chert outcrops or quarry evidence, but we did find numerous
96
o
ru
nJ .. .i
IJJ
D
D
D
D
U
R.. ..
IJ
D
D
D
E
D
D
~
~
o
flakes. It is like that Cooke mistook prehistoric Indian lithic sites for
in situ geological outcrops.
Upchurch p p e n d ~ x A) analyzed a sample of chert from this outcrop and
classified it as category 7a, silicified grainstone. This material is the
classic "Allendale Chert," and part of the Flint River Formation. The out-
crop at 38AL141 isa member of the Allendale Chert Quarry Clusterformu;..
lated by Upchurch (Appendix A). ..
Other Lithic Outcrops and Quarries
In Allendale County, South Carolina
38AL149 County Road 66 Outcrop
This is an outcrop of a highly siliceous brown sandstone. The outcrop
consists of massive boulders (Fig. 29) that occur on both sides of county
road 66 on the east side of Lower Three Runs Creek (Fig. 30). During two
visits to the outcrop, one in the summer and one in the winter, no evidence
of prehistoric utilization could be observed. A cultivated field was walked
on a hillslope immediately overlooking the stream bottom displaying the
rock. With the exception of one possible flake, no flakes of the sandstone
were found but over 30 flakes of typical Allendale chert were observed.
This outcrop ,is similar to the siliceous sandstones at Little Hell Landing
and the Millett Recreation Center in that an apparently knappable material
was ignored. We broke off the edge of a large boulder which revealed a
dark brown siliceous sandstone with thin veins of jasper-like chert in
places. These thin veins were of fairly good quality. Their minimal size
and embedment within large boulders probably precluded their use. A sample
of this rock was analyzed by Upchurch (Appendix A).
38AL150 Millett Recreation Center
Charles Dunbar, an artifact collector and employee of Sandoz-Martin
Works, informed Tommy Charles of an outcrop of stone near the recreation
center of Millett on Falls Mill Creek (Fig. 30). The silicified sandstone
occurs in large boulders (Fig. 31) protruding from a prominent hill. The
material is a white, hard siliceous sandstone. Like the other outcrops of
silicified sandstone observed in the Allendale County area, there was no
evidence of it being quarried by the Indians for chipped stone tools.
Material of this quality found to the east of the South Carolina Coastal
Plain would have been utilized for stone tools. Lithic material from the
outcrop was not included in the petrological analysis by Upchurch, but
similar material from the Coastal Plain was analyzed and described
(Appendix A).
38AL1 Little Hell Landing
This is a silicified sandstone outcrop immediately downriver from the
habitation site of 38AL1 (Fig. 30). This. stone outcrop was noted during
the period 1957-1959 by Roy J. Lyons, a collector from Aiken, South Caro-
97
Figure 29. Silicified sandstone boulders at 38AL149, where County
Road 66 crosses Lower Three Runs Creek.
lina, who was active in this area. He made a sketch map of the "Village
Site" (38AL1) and noted what he described as a "rock causeway" leading out
to an island in the Savannah River (LA.A. Site Files). Tommy Charles
visited the rock outcrop as part of the present survey and photographed an
example of the boulders (Fig. 32). No evidence that the silicified sand-
stone was used for chipped stone tools was found. A sample of the material
was analyzed by Upchurch (Appendix A) who indicated that it is possible to
knap it.
38AL151 Fastwater Outcrop
This is an imposing exposure of sandstone boulders jutting out from
the South Carolina side of the Savannah River (Fig. 30). It is located
along a sharp bend in the river where the current is rapid. The rock
exposure appears to be collapsing into the river (Fig. 33), no doubt a
result of floods. No roads lead to the outcrop and it must be reached by
boat. The sandstone material is of a very poor quality for knapping.
Intermingled among the large slabs of sandstone, however, are numerous
cobbles and boulders of chert. The chert cobbles possess a thin reddish-
brown cortical surface with considerable water polish. The quality of
chert within the cobbles is excellent. Numerous decortication flakes from
aboriginal quarrying lie wi thin a narrow terrace on the South Carolina
bank, indicating lithic extraction. Little worked material can be found
98
~ o
L __ J
l'::'J r:::J r:::J
I::::]
r=1 r:::l L ~ J r:::l ~ r:::'l Ii ~
r:::']
Ec,j I ! I _J Lt L_.__ l
~
1-'.
(JQ
os::
o 11
s:: (\)
~
n-\>l
~ o
en
o
s::
n-t-'
p-'O
()
01
I n-
11 0
o 11
I-'
~ m
,.l"O
NIS- ER;P
10 Kilo.
_ Survey
- - - Boundary
5 Miles
5 6
o
Ci:l
(\)
o
11 1-'.
(JQ ~
1-'.
~ >
I-'
I-'
(\)
[
I
I-'
(\)
I ~
[11
trl
o
s:: n-
11 p-'
:>r(\)
(\) 11
I
~ 0
A-s::
n-
en()
() 11
11 0
(\)"0
~ UJ
(\)
~ ~
g A-
s::
~ g
1-'. I
(\) 11
UJ 11
1-'.
(\)
UJ
\0
\0
Figure 31. Silicified sandstone boulders at 38AL150, near Millet,
South Carolina.
Figure 32.
Landing.
Silicified sandstone boulders at 38AL1, Little Hell
100
o
amongst the sandstone boulder. It may be that floods have washed away
smaller lithic artifacts from the outcrop. It may also have been the case
that the Indians took cobbles onto the beach and terrace for reduction.
Similar cherts to those found here can be found upstream, such as those at
Stony Bluff (Fig. 30). It is possible that the Savannah River has washed
chert boulders downstream where they have become lodged wi thin the sand-
stone outcrop. Quality chert such as this can be found within outcrops of
sandstone, however, as in the case at 38AL152. Upchurch (Appendix A: Cate,..
gory 8) describes one sample of chert from 38AL151 as "chalcedonic mud-
stone." It is from the Flint River Formation. This is the only example of
chalcedonic mudstone encountered by Upchurch, and accordingly, it might'
prove to be a geographically sensitive raw material type.
38AL148 Kerwin Outcrop
This is a chert outcrop on the west side of the Coosawatchie River
near the community of Barton about five miles south of Allendale (Fig. 30).
The chert was exposed by the dredging of a canal through a swamp behind the
home of Bert and Becky Kerwin out to the river. The chert exists in heav-
ily corticated boulders (Fig. 34) but is of excellent quality for flaking.
There was little evidence of exploitation of this chert source since only
occasional flakes with cortex were found. It may be that very little of
the chert was exposed to Indians, occurring as it does, down in low, wet
ground. A sample of chert was analyzed by Upchurch (Appendix A) who des-
cribes it as a "silicified grainstone," a variety of chert in the Allendale
Chert Quarry Cluster. The Kerwin Outcrop is significant for it is 'the
easternmost occurrence of Flint River Formation chert that we know of. All
other outcrops in South Carolina are restricted to the Savannah River or
its immediate tributaries.
38AL14 The Rice Quarry ("Red Bluff Flint Quarries")
Prior to the discovery and the recording of quarries on the Sandoz-
Martin Works property, the only other quarry known to produce high quality
Flint River Formation chert in South Carolina was that of 38AL14, recorded
in the National Register as "Red Bluff Flint Quarries," alias "The Rice
Quarry." This site was nominated to the National Register in 1971 by E.
Thomas Hemmings. This is an impressive chert outcrop and quarry which,
counting all loci, is over 20 acres in extent. The Rice Quarry, as it is
mOTe commonly known, is located about 10 miles south of the Sandoz-Martin
Works quarries (Fig. 30). The Rice Quarry was originally surveyed and des-
cribed by James B. Stoltman in 1964 as part of the Harvard expedition to
Groton plantation. He referred to it as "GR-5: Rice" (Stoltman 1974: 175).
Based on the present 'field survey, no other quarries are known to exist
between Johnsons Landing (38AL141) and the Rice Quarry (38AL14).
The quarry was tested in 1971 by James L. Michie, Sammy T. Lee and A.
Robert Parler. They were interested' in testing the site because of the
large quantities of chipped stone present and the highly weathered condi-
tion of the chert (Michie 1984). This extreme weathering, or "patination"
as it is often erroneously referred to in the archeological literature of
the Southeast (see Upchurch, Appendix A), has often been noted as charac-
101
Figure 33. Sandstone and chert boulders at 38AL151, Fastwater Outcrop
on the Savannah River.
Figure 34. Chert and limestone boulders at 38AL148, the Kerwin Outcrop
near Barton, South Carolina.
102
o
I
o
I
o
terizing Paleo-Indian and Early Archaic hafted bifaces and tools in South
Carolina (e.g., Michie 1977).
These investigators opened up three test pits in the cultivated field
on a low rise immediately south of King Creek (Michie 1984). Test pit 1
was located on the hill near the wood line. It was a 5 x 5 foot square dug
in 5-inch levels and screened. The plow zone was from 5 to 6 inches deep
and cultural material ceased about 24 inches below the plow zone. In
total, lithic material extended about 30 inches in depth (Fig. 35). No
natural soil layers were discerned nor were any diagnostic artifacts found
with the exception of a single thick steeply retouched uniface (Fig. 36e)
made of a highly siliceous tuff. This uniface was found in the lowest
5-inch level, between 25-30 inches (James L. Michie, personal communica-
tion).
Figure 35. Profile of 5 x 5 foot test pit at 38AL14, the Rice Quarry.
(Photo by James L. Michie, 1971).
Test pit 2 was a 10 x 10 foot square placed about 20 feet north of pit
1. After removing the plow zone, it was excavated in 6-inch levels and
screened. Surprisingly, sterile red clay subsoil was reached at 18 inches
below surface (Michie 1984). In the first 6-inch level below the plow
zone, a weathered chert preform of a Clovis, Suwannee or Dalton point was
found (Fig. 36a). A complete Dalton point was found on the surface of this
field by A. Robert Parler in 1971 while testing the site. The third test
pit, a 5 x 5 foot square, was dug on the hillcrest at the extreme southern
end of the field. It was very shallow and abandoned (James L. Michie,
personal communication).
103
Figure 36. Selected lithic artifacts from test excavations at 38AL14, the
Rice Quarry.
104
D
n
u
n....!
iii
o
The Rice Quarry, where tested by Michie, Lee and Parler, appears to
represent a substantial Late Pleistocene and Early Holocene chert process-
ing location. In terms of technology and weathering, the flakes, cores,
unifaces and biface fragments (Fig. 36) recovered here are virtually iden-
tical to the bottom layer of sites 38AL135, 38AL143, and 38AL23. It
appears that chert was procured from boulders down in King and Watch Call
creeks and their swamps and worked into cores, preforms, and tools on the
adjacent higher and drier ground. Chert boulders more thana meter in
diameter are present in the swamp of King Creek (Fig. 37).
The soil bearing artifacts on the hill tested by Michie and others
appears fairly homogeneous (Fig. 35). The artifacts are contained in a
brown sandy clay resting upon culturally sterile red clay subsoil. The
depth from ground surface to the red clay seems quite variable over the
hilltop. In test pit 1, however, artifacts were continuously found until
30 inches, indicating that parts of the hill can provide stratigraphic
information (Fig. 35).
Figure 37. Chert boulders from outcrop in swamp of Watch Call Creek,
38AL14, the Rice Quarry.
Two early man artifacts made of Piedmont rocks have been found on this
hill (Fig. 36b, e) and are significant from the standpoint of studying
Early Holocene hunter-gatherer settlement systems. The uniface which was
excavated by Michie and others appears to be a fine-grained tuff. The Kirk
stemmed biface (Fig. 36b), found on the surface in 1980 by Sammy Lee is
made of flow banded rhyolite. Both of these raw materials are more common
in the eastern portions of South Carolina and western North Carolina. Fol-
lowing Binford's (1979) argument that artifacts made of materials exotic to
105
a region are potential indicators of mobility scale among hunter-gatherers,
these two pieces indicate substantial movement from the Piedmont to the
Middle Coastal Plain. It is interesting to note that quartz Palmer points
were found at the Theriault site (38BK2) (Fig. 30), indicating Piedmont
relationships (Brockington 1971: Fig. 11h, k). Middle and ~ t Archaic
points from there were also made from quartz andrhyoli te (Brockington
1971: Fig. 12d, h; 1). Conversely, the finding of. Allendale _chert Paleo-
Indian lanceolatesand Early Archaic notched points in the western Piedmont
of South Carolina is commonplace and well documented (Goodyear, House, and
Ackerly 1979: 181-184; Charles 1981: 17-18; Anderson and Schuldenrein 1983:
203-205).
The chert from the Rice Quarry has received some limited petrologic
analysis (Baum 1979: 30-38; Anderson et al. 1982: 126). Using primarily
fossil criteria, it has proven to be reasonably distinguishable from other
Coastal Plain chert forms in the Black Mingo Formation. Anderson (1979a)
has also performed thermal alteration studies using chert from the Rice
Quarry in an effort to derive macroscopically visible correlates of heat
treatment. In the present study, raw material from unmodified boulders was
analyzed and described by lTpchurch (Appendix A). . The Rice Quarry chert
represents a type of silicified grainstone shared with sources on the
Sandoz-Martin Works property and Brier Creek, Georgia. Based on our cur-
rent survey knowledge the Rice Quarry, 38AL14, is the largest single mani-
festation of Flint River Formation chert in South Carolina and displays the
greatest evidence of prehistoric utilization
. Other Outcrops and Quarries in
Burke and Screven Counties, Georgia
U.S. 301, Allied Chemical Sylvania Terminal
Screven County, Georgia
This is a natural source of chert exposed by digging a small indus-
trial pond on the east side of the Allied Chemical Sylvania terminal plant.
It is located .25 miles west of the public boat landing on the downriver
side of U.S. 301 in Georgia (Fig. 30). A paved road runs west from the
boat landing back to U.S. 301. The chert boulders were seen while in route
to the boat landing. A few very large boulders (greater than 1.0 m) were
seen from the paved road. This material was sampled and analyzed by
Upchurch (Appendix A) who describes it as a "light brown, silicified grain-
stone." It is a member of the Allendale Chert Quarry Cluster. There was
no evidence that this material was utilized prehistorically. The boulders
appear to have been dragged up during the construction of the pond. The
surrounding ground is very low and wet which may have precluded aboriginal
detection and/or utilization. The value of this outcrop for archeological
purposes would be that it provides another datum point in the geographic
study of the distribution of the high quality Allendale Chert Quarry
Cluster.
106
-
D
EJl ....'....
U
U
I
D
D
I
i
o
D
D
I
i
B
I.. ...
I
Q
o
Wade Plantation Quarry No.1, Screven County, Georgia
This is a chert processing site located on a small cultivated hill
cqvering approximately three to four acres. The hill rises about 8 to 10
feet above the surrounding floodplain of McDaniel Creek which lies imme-
diately south. The Savannah River lies 220 meters to the east (Fig. 30).
There isa substantial amount of worked chert on the site including flakes,
cortical chunks, and . broken bifacial blanks . Thermally altered material
was evident. Natural chert was present on the hillsides in the form of
boulders and cobbles. It is not clear if the hill was the chert source or
if lithic material was gathered from McDaniel Creek. No check was made of
the densely vegetated bottomland. At any rate, given the abundance of
cortical material and boulder-sized nodules, a source is no doubt present
within a few hundred meters of the site. Upchurch (Appendix A) analyzed a
piece of chert from the site which was unique (Category 6). He refers to
it as "coralgal grainstone"because of the presence of coralline algal
remains. Since coralline algae is absent in the Allendale Chert Quarry
Cluster of the Flint River Formation, it is thought that this material is
from the Barnwell or Sandersville Limestone Formation. Upchurch (Appendix
A) reports that this material should lend itself to flintknappingas it is
well cemented and homogeneous. A survey of McDaniel' Creek on the Wade
plantation would likely yield major chert processing sites and quarries.
9BUR6 Stony Bluff
This site was visited by Clarence B. Moore toward the end of the nine-
teenth century (Moore 1898:172). It is an outcrop and workshop located on
the bank of the Savannah River south of Little Sweetwater Creek (Fig. 38).
Moore (1898: 172) described the site as where ..... the solid ground comes to
the water's edge with masses of rock above the surface." A contemporary
photograph (Fig. 38) reveals an identical situation. Today, pastureland
and homesi tes obscure much of what Moore was able to see in terms of
sprawling lithic debris from Indian chert processing. There is no doubt
that Stony Bluff was a major chert source and processing area for this area
of Georgia. Chert samples were not taken during the present survey.
9BUR59
This site was first recorded during the present survey. It is located
on the north side of Stony Bluff Road about 450 meters east of the juncture
of Rock Fence Crossroads and Stony Bluff Road and about 900 meters south of
Little Sweetwater Creek (Fig. 30). The site is a quarry and a source.
Considerable quantities of lithic debris are present, almost carpeting the
ground surface (Fig. 39) over a two-acre area. The site had been recently
providing excellent Visibility. The density is of the
Rice Quarry, 38AL14, in South Carolina. A chert sample was analyzed by
Upchurch (Appendix A) who describes it as a silicified grainstone, a member
of the Allendale Chert Quarry Cluster.
107
Figure 38. The outcrop and quarry at Stony Bluff, 9BUR6 on the
Savannah River, Georgia.
Figure 39.
Chert debris on recently cleared land at 9BUR59 , Georgia.
108
9BUR11
9BUR60 Roy Hillis Road Quarry
9BUR61 Ellison Bridge Road
Ellison Road (Fig. 42), exposing
cores and flakes. A sample of
This sample was unique among all
109
Earth moving has taken place next to
many cobbles of chert as well as worked
chert was taken for petrologic analysis.
During the present survey the site was revisited and chert samples
were taken and submitted for petrologic analysis. Upchurch (Appendix A)
describes this as a silicified grainstone with coralline algal fragments
present. This quarry is a member of the Allendale Chert Quarry Cluster.
The existence and location of this site was recorded by George S.
Lewis in 1975 in conjunction wi th his reporting of site 9BUR11. A site
number was assigned during the present survey. During a visit in 1983, the
si te seemed to be mainly represented by natural boulders and cobbles of
good quality chert. Evidence of extraction and core reduction, while not
seen in the roadbed, is no doubt immediately nearby. The chert is exposed
primarily where Roy Hillis Road , a dirtroad, crosses Little Sweetwater
Creek (Fig. 30). Upchurch (Appendix A) analyzed raw material from here and
describes it as silicified grainstone, from the Flint River Formation, a
member of the Allendale Chert Quarry Cluster.
This is a tremendous chert exposure and quarry recorded in the Georgia
site files in 1975 by George S. Lewis (Fig. 30). Lewis reported that this
site and others north of Little Sweetwater Creek have been heavily col-
lected by rockhounds who have paid mining fees to landowners. One man was
reported to have hauled off 15.tons from just east of River Road.
This is an extensive outcrop and quarry located on the north side of
Brier Creek on the west side of Ellison Road about five' miles south of
Girard, Georgia (Fig. 30). It was mentioned and its location noted in 1975
by George S. Lewis in conjunction with reporting the Jenkins Peanut Field
site (9BUR10) in the Georgia site files. The Jenkins site is located about
a mile west of 9BUR61 on the south side of the creek. Both 9BUR10 and
9BUR61 are members of the extensive quarry-workshop manifestations along
the famous Brier Creek locality of which the Theriault site, 9BUR2 (Fig.
30), is perhaps the best known (Brockington 1971).
The site is exposed along the dirtroadca11ed River Road and extends
an unknown distance east and west in the woods. Nodules, cores, flakes,
and cortical debris extend for nearly 600 meters along this road (Fig. 40).
The main exposure begins in a small valley of a tributary of Little Sweet-
water Creek and continues northward along River Road until the apex of a
hill. In June 1980, Goodyear and Charles were shown this site by David
Topper of Fairfax, South Carolina. A deep layer of sand was observable on
the western side of River Road in a road cut (Fig. 41) which exhibited
lithic artifacts up to 90 cm below surface. This situation suggests a good
potential for data recovery by excavation.
D

D
I
I
I
D

0...,..'..
I
I
D
D
D
D
D

D
D
Figure 40.
Chert debris in roadbed of 9BUR11 , Georgia.
Figure 41. Profile in road cut of 9BUR11 showing depth of artifacts
in sand on west side of road.
lW
o
o
o
Figure 42. Chert debris at Quarry 9BUR61 on Ellison Bridge Road,
Georgia.
the samples submitted to Upchurch from South Carolina and Georgia. Up-
church (Appendix A) describes Category 5 as "opaline, foraminiferal wacke-
stone/grainstone." He relates this chert to the Barnwell Formation and
tentatively classes it as "opalized Sandersville Limestone." This may be a
distinct lithic raw material type but one which is not very abundant. The
piece he analyzed appears to have been flaked by man and thermally altered
and otherwise not ideal for petrologic analysis (Appendix A). The Brier
Creek locali ty should be sampled and analyzed more extensively than has
been possible in the present study.
111
112
-
CHAPTER III
CONCLUSIONS
The Flint River Formation cherts of Allendale County are significant
in relationship to the Southeast and South Carolina. As discussed in the
introduction, the tertiary age cherts of the southern Coastal Plain in the
Southeast formed the primary source of chipped stone raw material for the
aboriginal inhabitants in this broad area. This material, especially when'
thermally altered, provided an excellent lithic resource for their tech-
nologies. The Paleo-Indian and preceramic Archaic occupation of the
Coastal Plain is known almost exclusively by stone tools manufactured from
these cherts.
Through geographic coincidence, the extreme northernmost distribution
of this broad chert manifestation lies in western Allendale County, South
Carolina. Following Cooke ( 1936), the Allendale chert, as it has been
referred to by archeologists working in South Carolina, has been thought to
be derived exclusively from the Flint River Formation of the Oligocene in
South Carolina and the Burke County area of Georgia. The results of this
survey confirm previous impressions of the nature, sources and geographic
distribution of this distinctive high quality cryptocrystalline chert.
Cherts from South Carolina quarries that would be macroscopically classed
as "Allendale" by amateur and professional _archeologists are all derived
from the Flint River Formation and all but one source, 38AL151, Fastwater
Outcrop, are examples of silicified grainstone. The latter, while macro-
scopically similar to the silicified grainstones, under microscopic exami-
nation is a chalcedonic mUdstone.
In Georgia, some of the cherts which have been called Allendale, or
sometimes Brier Creek, are from the Flint River Formation, but in two cases
they appear to be from different fOI1Dations and petrologically distinct.
The samples taken from the three sites along Little Sweetwater Creek
(9BUR59, 9BUR11, 9BUR60), which flows near Stony Bluff, and the outcrop
behind the Allied Chemical Terminal are all silicified grainstones of the
Flint River Formation. They unite with Allendale County chert sources in
South Carolina to form the Allendale Chert Quarry Cluster (Appendix A).
The chert from 9BUR61, where Ellison Bridge Road crosses Brier Creek, is
tentatively classified as opalized Sandersville Limestone derived from the
Barnwell Formation. Because of apparent thermal alteration, Upchurch found
this sample to be less than ideal for analysis (Appendix A). The sample
from Wade plantation in Screven County, Georgia, could have been derived
from either the Barnwell or Flint River Formation. It too is possibly from
the Sandersville Limestone. Both of these samples are homogeneous and
brittle and would have flaking characteristics like classic Allendale
chert. More samples need to be analyzed from these quarries plus many more
sources which no doubt exist along Brier Creek to fully document the exis-
tence and geographic extent of high quality cherts of extreme eastern
Georgia that are not part of the Allendale Chert Quarry Cluster.
The results of this survey tend to confirm the impression held by
artifact collectors and professional archeologists that western Allendale
113
county is the primary source area for Allendale chert originating in South
Carolina. Although other areas of Allendale County were not surveyed,
collectors have not reported other chert outcrops to the east, north or
south of the area between Martin and the Rice Quarry. No chert quarries
have been located on the Savananah River Plant to the north in Barnwell and
Aiken counties after five years of near continual survey (Glen Hanson,
personal communication). The exception is 38AL148, the Kerwin Outcrop near ..
Barton which is the furthest distance east of the Savannah River where
Allendale chert has been reported. It was scarcely exploited, however.
Nine new, previously unrecorded quarries were documented by this sur-
vey on or adjacent to the Sandoz-Martin Works property, all of which are
close to the Savannah River. These nine plus five related sites were
proposed as an archeological district for the National Register of Historic
Places in 1984.
While it is possible some additional minor terrestrial sources will be
discovered in the future in Allendale County, we believe this survey has
pinned down the basic inventory of Allendale chert quarries in South Car-
olina. Additional outcrops of chert available and exploited by prehistoric
populations within the Savannah River itself is, however, quite likely.
Furthermore, it is probable that many more sources and quarries of Allen-
dale chert of the Flint River Formation exist in eastern Georgia and that
these contributed numerous artifacts that have been recovered in South
Carolina.
There is evidence of other minor high quality chert sources that may
have been available in the extreme southern Coastai Plain of South Carolina
in Hampton, Jasper and Beaufort counties. In his extensive travels with
the private collector survey (1979-1984), plus 15 years of personal col-
lec ting experience, Tommy Charles has observed evidence that Allendale
quality chert may occur naturally at sources closer than the Rice Quarry.
He has observed cobbles of chert bearing cortex in the yards of collectors
and on sites along the Coosawhatchee and Salkahatchee rivers in Allendale
and Hampton counties. The quality of chert ranges from excellent to un-
knappable. The large size of the cobbles still retaining cortex argues for
a source closer than the Rice Quarry which is over 25 miles away. The
Kerwin Outcrop on the Coosawhatchee River (38AL148) near Barton would be
only 15 miles distant, but it was scarcely used. Sites along the Coosa-
whatchee show more evidence of chert sources being nearby than sites along
the Salkahatchee. Despite the evidence of nearby chert sources, the Kerwin
Outcrop is' the closest source Charles has ever seen. It may be that minor
sources of Allendale chert exist in this area but like the Kerwin Outcrop,
they are situated down in swampy floodplains, precluding all but minor
exploitation or infrequent use during very different hydrological condi-
tions over the past 12,000 years.
Charles also has the impression that the cobbles and pebbles of high
quality Allendale-like chert were available at least in limited quantities
in areas that today are near tidal rivers and creeks in Beaufort and Jasper
counties. These materials have water-worn cortex revealing a stream
source. Most of. these examples are stained blue, gray and black from their
submergence in watery, organic environments. He has also observed an unu-
sual high quality tranlucent orange chert in cobble form on the Coosaw
114
o
U
IC1. J
U
o
[1 ....'...
u
D
D
D
D
D
D
o
o
o
River. This material appears to have been limited in availability as only
occasionally are artifacts found made from it. Another good quality chert
has been observed by Charles exposed in the bank of the Savannah River at
Purreysburg, South Carolina. It is a gray river cobble chert found in a
low-lying stratum only visible at low tide. There was no evidence of
aboriginal exploitation here. Another material resembling Allendale chert
has been observed on an old terrace of .the.Savannah River in South Carolina
revealed by a borrow pit next to Interstate 95. It is olive gray in color.
Chunks of it were found in this borrow pit by a collector. Charles has
seen occasional chipped stone artifacts made from this material.
In his survey of the Broad River estuary, James L. Michie (1980: 76)
discovered a site of prehistoric utilization of cobble chert. Michie
. observed occasional cobbles of chert bearing river-worn cortex on the sandy
beaches of the Broad River. On one site, 38BU285, he found a concentration
of worked cobbles 10 to 20 em in diameter used to produce cores and debi-
tage. Some of these are split in such a way as to suggest they may have
been reduced by bipolar flaking. This work is important because it firmly
establishes a series of sites where local river cobble chert was utilized.
In terms of quality and color ,these various cherts resemble the
Allendale chert of the Flint River Formation. The Flint River Formation,
however, is restricted to a narrow strip in Allendale County (Cooke 1936:
98). Hawthorn Formation cherts occur in Beaufort County (Appendix A), but
these are decidely inferior to 'the Flint River Formation cherts. The
Coosawhatchee River, known from the Kerwin outcrop example, must have some
natural supplies of Allendale chert. It may be that the cherts reported by
Charles and Michie in the counties south of Allendale to the coast are
stream transported. The Coosawhatchee, for example, drains into the Broad
River. Extensive petrologic analysis needs to be done to establish the
nature and origin of these cherts. One piece of chert from Hampton County
(38HA74) was submitted to Upchurch for classification. In his analysis,
Upchurch (Appendix A) found it to be "coralgal boundstone," suspected to
have derived from the Oligocene Flint River Formation. Regardless of their
source, they appear to have provided a small proportion of chipped stone
raw material to inhabitants of the South Carolina Coastal Plain, being
by the extensive quarries of Allendale County and probably neigh-
boring Georgia counties.
In addition to chert samples which were thought to be Flint River
Formation in origin, other specimens were analyzed by Upchurch (Appendix A)
to determine their petrologic character.
Certain chert-like artifacts are' sporadically found in the Piedmont
which often approach the cryptocrystalline quality of the Allendale chert
and sometimes resemble the latter in color and texture. Two of these are
described as sheared siltstone and fractured phyllite (Appendix A).
Another has been generally described as Jasper (NoVick 1978; Goodyear ,
House and Ackerly 1979: 198). Jasper found in the northwest Piedmont can
resemble unweathered dark brown Allendale chert.
The category of opaline coquina A) has been grossly des-
cribed as "fossiliferous chert" by South Carolina archeologists. One form
of this material has been labelled "Buyck's chert" from its outcrop at
115
Buyck's Bluff (38CL17) on the Congaree River. A sample taken from this
si te and one from a quarry in Sumter County were easily distinguishable
from Allendale chert by Upchurch. The highly fossiliferous and thus coarse
structure of this chert is due to the well preserved mollusc fabric.
The sandstone (Appendix A), as classified by Upchurch,
is basically what has been .traditionally called orthoquartzite by archeolo-
gists in South Carolina. It is interesting that this usualiy knappable'
. material, which was extensively used in the center of the South Carolina
Coastal Plain, was essentially ignored in the area of Allendale County.
Huge boulders of this material can be found wi thin five miles of the
Sandoz-Martin Works quarries. This suggests that the techno-economic util-
i ty' of Allendale chert was much higher than that of the silicified sand-
stone which was ignored as long as good chert was nearby.
A piece of waxy chert-like material from the bayshore at Beaufort was
identified as silicified siltstone. It is typical of the Hawthorn Forma-
tion (Appendix A). This material looks softer and waxier than Allendale
chert and is also highly prone to weathering. This material may be what
appears after extensive decomposition as soft, eroded, mud-like flakes on
sites in the southern Coastal Plain.
The last significant category of material investigated is what turns
out to be flint imported as ship ballast from Western Europe (Appendix A).
For some time divers have brought up small cobbles and pebbles of high
quali ty flint from lower Coastal Plain rivers. Some pieces resemble the
Allendale chert in terms 'of translucency and orange and yellow colors. It
was considered possible that Flint River Formation cherts or other occa-
sional high quality Santee Limestone cherts might be producing chert peb-
'bles transported down Coastal Plain streams. Their condition has been
enigmatic, however, in that they were never utilized aboriginally. The
fact that it is historic ballast neatly accounts for that.
Finally, a word can be said about the relationship between the study
of chert quarries and things of interest to anthropologists. There are
certain research directions our studies can take at this point now that
chert sources have been located and an individual stone tool can be
potentially traced 'to its approximate point of origin.
Greater refinement of knOWledge is needed of the content and variation
of high quality cherts from eastern Georiga. It is clear that cherts not
part of the Allendale Chert Quarry Cluster exist in eastern Georgia and
were probably used by aboriginal groups who also inhabited South Carolina.
Given the tremendous geographical expanse covered by known ethnographic
groups in their settlement space, not just within a year but over a life-
time of an individual (Binford 1980), data on sources and dispersions of
li thic tool kits will be needed that completely cover what are now states
wi thin the Southeast. Eventually, the quarry clusters as defined for
Florida by Upchurch and others (1981) will need to be discovered and mapped
for the entire state of Georgia ultimately connecting with the South Caro-
lina clusters. This will allow analysis of settlement strategies through
time and space over this broad 500 mile long area, using a physical factor
common to all occupants, namely the use and dispersion of Coastal Plain
cherts.
116
Data wi thchronological controls also need to be recovered from key
sites where the processing of raw chert into transported cores and tool
kits took place. Ideally, this is a stratified site with good separation
between occupational horizons. Sites 38AL135 and 38AL23 are just such
sites. Assemblages from sites such as these need to be analyzed in order
to provide an analytical datum point against which tools and assemblages
found away from the source can be compared. Theoretically speaking,
economic utility of. a tool kit is highest at or near the raw material
source because all reduction episodes designed for a tool kit, core, or
tool have yet to occur. As time (partially measurable by geographic dis-
tance from the source of a lithic artifact) unfolds, variability begins to.
manifest itself due to reduction episodes (always resulting in a loss of
mass) related to use, rejuvenation, recycling, loss and caching. Strate-
gies of tool design and construction also need to be considered in conjunc-
tion with mobility strategies because they are ultimately interrelated. As
mobility levels decline and change due to population growth, changes in
technological strategies can be expected as well. In the Institute of
Archeology and Anthropology, extensive survey collections gathered from
nearly every county in South Carolina are available for studying variabil-
ity in lithic assemblages over space.
117
118
REFERENCES
119
1980 Willow Smoke and Dogs' Tails: hunter-gatherer settlement sys-
tems and archaeological site formation. American Antiquity
45: 4-20.
curated
35(3) :
E. Donahue,
Report No. 2311, Common-
The Mattassee Lake Sites,
the Lower Santee River in
formation processes: looking at
Journal of Anthropological Research
the Coastal Plain of South Carolina.
wealth Associates,
Excavations at Four Fall Line Sites: the Southeastern Columbia
Beltway Project. Commonwealth Associates, Inc., Report No. i
R-2008. Jackson, Michigan.
R.
Lithic and thin section analysis results. In Excavations at
Four Fall Line Sites: The Southeastern Columbia Beltway Pro-
ject, Commonwealth Associates, Inc., Report No. R-2008, pp.
30-38. Jackson,
David G. and Joseph Schuldenrein .
Early Archaic settlement on the southeastern Atlantic . slope:
a view from the Rucker's Bottom Site, Elbert County, Georgia.
North American Archaeologist 4(3): 177-210.
David G., Sammy T. Lee, and A. Robert Parler
Cal Smoak: archaeological investigations along the Edisto
River in the Coastal Plain of South Carolina. Occasional
Papers 1, Archeological Society of South Carolina, Inc. Uni-
versity of South Carolina, Columbia, South Carolina.
Anderson, David G.
1979a Prehistoric selection for intentional thermal alteration:
tests of a model employing southeastern archaeological mate-
rials . Mid-Continential Journal of Archaeology 2(2): 221-254.
Baum, Gerald
1979
Anderson,
1979
1979b
Anderson,
1983
Bartram, John
. 1942 Diary of a journey through the Carolinas, Georgia, and Flor-
ida, from July 1, 1765 to April 10, 1766, annotated by Francis
Harper. Transactions of the American Philosophical Society
New Series Volume 32, Part 1. Phildadelphia.
Binford, Lewis R.
1979 Organization and
technologies.
255-273.
Anderson, David G., A. Lee Novick, William A. Lovis, Randolph
and Gerald R. Baum
1982 Laboratory analysis procedures. In
archeological investigations along
Brockington,
1971
Paul
A preliminary investigation of an early knapping site in
southeastern Georgia. Universit of South Carolina, Institute
of Archeology and Anthropology, Notebook III 2 : 23-46.
Bullen, Ripley P. and H. Bruce Greene
1970 Stratigraphic tests at Stalling's Island, Georgia.
Anthropologist 23: 8-23.
Florida
Caldwell, Joseph R.
1958 Trend and tradition in the prehistory of the eastern United
States. American Anthropological Association, Memoir 88.
Charles, Tommy
1981 Dwindling resources: an overture to the future of South Caro-
lina's archeological resources. University of South Carolina,
Institute of Archeology and Anthropology, Notebook 13: 1-85.
Claflin, William H.
1931 The Stalling's Island Mound, Columbia County, Georgia. Papers
of the Peabody Museum of American Archaeology and Ethnology,
Harvard University XIV(1).
Claggett,
1982
Stephen R. and John S. Cable
The Haw River Sites: archeological investigation at two
stratified sites in the North Carolina Piedmont. Commonwealth
Associates Inc., Report No. 2386. Jackson, Michigan.
Coe, Joffre L.
1964 The formative cultures of the Carolina Piedmont. Transactions
of the American Philosophical Society, n.s., 54.
Cooke, C. Wythe
1936 Geology of the Coastal Plain of South Carolina. United States
Geological Survey, Bulletin 867.
Crabtree,
1972
Don E.
An introduction to flint working: Part I: an introduction to
the technology of stone tools. Occasional Papers of the Idaho
State University Museum 28. Pocatello.
Crabtree, Don E. and B. Robert Butler
1964 Notes on experiments in flintknapping. Part 1: heat treat-
ment of silica minerals. Tebiwa 7:1-6.
Daniel, Randy and Michael Wisenbaker
1983 A preliminary report on the excavations at Harney flats,
Hillsborough County. The Florida Anthropologist 36(1-2): 67-
80.
f
1984 Salvage excavations at Harney Flats: a Paleo-Indian base camp
in Hillsborough County, Florida. Manuscript on file at Divi-
sion of Archives, History and Records Management, Florida
Department of State, Tallahassee, Florida.
120
u
Goodyear, Albert C., Sam B. Upchurch, Mark J. Brooks, and Nancy N. Goodyear
1983 Paleo-Indian manifestations in the Tampa Bay Region, Florida.
The Florida Anthropologist 36(1-2): 40-66 .
Gregg, Michael L. and Richard J.Grybush
1976 Thermally altered stone from prehistoric contexts. Intentional
versus unintentional alteration. American Antiquity 41: 189-
192.
Gardner, William M.
1977 The Flint Run Paleo-Indian complex and its implications for
eastern North American prehistory. In Amerinds and their
paleo-environments in northestern North America. Annals of
the New York A9ademy of Sciences 288: 257-263.
Hanson, Glen
1981
Site: a technofunctional study of a Dalton Site in
Arkansas. Arkansas Archeological Survey Research
Ie
Chert resources in Georgia, archaeological and geological
perspectives. Wallace Reservoir Project Contribution No.3,
University of Georgia Laboratory of Archaeology Series, Report
Number 21.
T., Jr., Richard D. Brooks, and John W. White
The human occupation along the Steel Creek floodplain: results
of an intensive archeology survey for the L Area Reactivation
Project, Savannah River Plant, Barnwell County, South Caro-
lina. University of South Carolina, Institute of Archeology
and Anthropology, Research Manuscript Series 173.
121
H. and Ronald W. Wogaman
Windy Ridge, a prehistoric site in the interriverine Piedmont
in South Carolina. University of South Carolina, Institute of
Archeology and Anthropology, Anthropological Studies Series 3.
Clemens
A test excavation of the Hollywood mound (9RII), Georgia.
Southern Indian Studies, Vol. 17, pp. 3-11. University of
North Carolina, Chapel Hill.
Albert C.
The Brand
Northeast
Series 7.
Albert C., John H. House and Neal W. Ackerly
Laurens-Anderson: an archeological study of the interriverine
Piedmont. University of South Carolina, Institute of Arche-
ology and Anthropology, Anthropological Studies 4.
A hypothesis for the use of cryptocrystalline raw materials
among Paleo-Indian groups of North America. . University of
South Carolina, Institute of Archeology and Anthropology,
Research Manuscript Series 156.
DeBaillou,
1965
1979
Goad, Sharon
1979
Goodyear,
1974
Goodyear,
1979
House, John
1978
ID
D
D
D
D
10
10
D
10
10
10
10
D
D
10
D
U
D
122
1954 Mid-coastal Archaic in Georgia. Southern Indian Studies Vol.
V, pp. 14-27. Chapel Hill.
Michie, James L.
1968 The Brier Creek lanceolate. The Chesopiean 8: 7.
An early flint industry in southwest Georgia. Short contribu-
tions to the Geology, Geography and Archeology of Georgia.
Georgia Academy of Sciences Bulletin 56: 146-153.
Primitive manufacture of spear and arrow points along the line
of the Savannah River. Annual Report of the Board of Regents
of the Smithsonian Institution for the Year 1879, pp. 376-382.
A preliminary report on archaeological exploration at Macon,
Georgia. Anthropological Papers, Bureau of American Ethnol-
ogy, Bulletin 199, Washington, DC.
1977 The Late Pleistocene human occupation of South Carolina.
Manuscript on file at the Institute of Archeology and Anthro-
pology, University of South Carolina.
Moore, Clarence B.
1898 Certain aboriginal mounds of the Savannah River. Journal of
the Academy of Natural Sciences of Philadelphia, Second
Series, Vol. 2, Part 2, pp. 162-172. Philadelphia.
1984 Notes on the 1971 test of the Rice Site, 38AL14. Manuscript
on file at the Institute of Archeology and Anthropology, Uni-
versity of South Carolina.
1980 An intensive shoreline survey of archeological sites in Port
Royal Sound and the Broad River Estuary, Beaufort County,
South Carolina. University of South Carolina, Institute of
Archeology and Anthropology, Research Manuscript Series 167.
1983 Bioturbation, gravity and site formation processes in the
Coastal Plain of South Carolina. Paper presented at the 40th
Annual Southeastern Archaeological Conference, Columbia, South
Carolina.
Johnson, Jay K.
1979 Archaic biface manufacture: production failures, a chronicle
of the misbegotten. Lithic Technology 8:25-35.
Jones, C. C.
1880
Kelly, A. R. and V. J. Hurst
1956 Patination and age relationships in South Georgia flint.
American Antiquity 22: 193-194.
Keeley, Lawrence H.
1980 Experimental determination of stone tool uses: a microwear
analysis. University of Chicago Press, Chicago.
1950
Kelly, A. R.
1938
123
Semenov, S. A.
1964 Prehistoric technology. Adams and Dart, Bath, Somerset, Eng-
land.
Wormington, H. M.
1957 Ancient man in North America, 4th Edition, Denver Museum of
Natural History. Popular Series 4. Denver.
an archaeological'
Lee
Prehistoric Iithic material sources and types in South Caro-
lina: a preliminary statement. South Carolina Antiquities
10(1): 422-437.
Sam B., Richard N. Strom, and Mark .G. Nuckels
Methods of provenance determination of Florida cherts. Report
submitted to the Florida Bureau of Historic Sites and Proper-
ties, Department of State, Tallassee, Florida.
Kenneth E., Jr.
Middle and Late Archaic settlement in the South Carolina Pied-
mont. M.A. Thesis, Department of Anthropology, University of
South Carolina.
James B.
Groton Plantation: an archeological study of a South Carolina
locality. Monographs of the Peabody Museum, Harvard Univer-
sity No.1.
Novick, A.
1978
Wilmsen, Edwin N.
1970 Lithic analysis and cultural inference: a Paleo-Indian case.
Anthropological Papers of the University of Arizona 6. Tucson.
South, Stanley
1959 A study of the prehistory of the Roanoke Rapids Basin. Unpub-
lished M. A. Thesis, Department of Anthropology, University of
North Carolina, Chapel Hill.
Purdy, Barbara Ann and H. K. Brooks
1971 Thermal alteration of silica materials:
approach. Science 173: 322-325.
Stoltman,
1974
Phelps, David S.
1968 Thomt s Creek ceramics in the' Central Savannah River locality.
The Florida AnthropolOgist 21(1): 17-30.
Sassaman,
1983
Upchurch,
1981
124
D
B
IQ
I
10. ......
LJ -"

i
ImI ... .. .
W

D
I
m
I
I
I
I
1m
iii
U
I
o
APPENDIX A
PETROLOGY OF SELECTED LITHIC MATERIALS
FROM THE SOUTH CAROLINA COASTAL PLAIN
Sam B. Upchurch
Department of Geology
University of South Florida
Tampa, Florida 33620
April 8, 1984
Submitted to the
Institute of Archeology and Anthropology
Columbia, South Carolina
.125
126
o
u
o
u
u
~
II..,..,.
I
~
i
I
I
I
I
I
~
I
g
I
o
Introduction
This report is a preliminary description of lithic materials provided
by the Institute of Archeology and Anthropology (IAA) for petrographic
description and categorization. The materials were collected by Dr. A. C.
Goodyear and Mr. T. Charles of the IAA and delivered tome at ~ p a These
rock samples were taken from known or suspected Indian quarry sites..The
purpose of this report is to describe the material and develop criteria
that would allow preliminary determinations of provenance of Iithic arti-
facts from the archeological record.
Upchurch etal. (1981) have developed a series of criteria for Iithic
material provenance determinations based on rock fabric, fossil content,
and inclusions. These criteria have been successfully applied to Iithic
materials in Florida (Upchurch et a!. 1981; Goodyear et a1. 1983). While
the sample set prOVided by the IAA is not large enough to completely define
a similar set of criteria for South Carolina and nearby portions of
Georgia, it does indicate that this is possible and that there is a wide
diversity of materials available for aboriginal tool manufacture.
In addition, Upchurch et a1. (1981) developed the concept of quarry
clusters to categorize areas within which similar lithic materials could
have been available to early man. These areas include known or suspected
quarry sites from which rock with, similar chemical, lithologic, and fossil
characteristics could have been obtained. The premise in attributing
lithic materials toa quarry cluster, rather than a known quarry site, is
the certainty that our knowledge of quarry sites available to early man is
flawed. Most quarry sites used by early man are not known. They have been
lost by later weathering; sedimentation; inUndation by rivers, lakes,
wetlands, and the sea; overgrowth by vegetation; or remoteness to modern
man.
The samples ~ r o m the South Carolina and Georgia Coastal Plain clearly
indicate that quarry clusters can be identified. Further work is needed to
clearly define them, but the present sample set can be categorized in such
a way as to indicate some of the possible quarry cluster locations.
Methods
In order to initiate quarry cluster definition and make attribution of
lithic materials to quarry clusters possible for archeologists, the first
step is to develop categories of rock materials that are easily recogniz-
able in the field. The sample set was treated in this way. All hand sam-
ples were labeled wi th a numbering scheme independent of that provided by
IAA. Then, wi thout bias as to origin of the rock material, the samples
were placed into piles that contained samples with similar composition,
fabric, color, fossils, inclusions, etc. The rocks were thin sectioned to
determine the true fabric, composition, and fossil and inclusion content
and the samples were again categorized. Finally, the locations of origin
of the samples were compared and related to known geology of the area
(Cooke 1936; and other sources).
127
Thirty-three samples were treated in this manner and they were grouped
into twelve different categories, ten of which may ultimately be included
wi thin distinctive quarry clusters. The categories are shown in Table 1.
The descriptions are grouped according to these twelve categories and sub-
divided into hand-sample descriptions, which are most u s ~ l to field
archeologists and the basis of the classification, and thin-section des-
cription, which is most useful- for detailed attribution to quarry clusters
and material evaluation for tool manufacture. -
The term chert is used in many ways by sedimentologists and archeolo-
gists. In this report it is used to include any sedimentary rock in which
the bulk of the material is one of the silica polymorphs (quartz, opal-CT).
There is a great variety of different mineral combinations and fabrics
(microarchitectures) included wi thin the rock type known as chert. These
different mineral combinations and fabrics give cherts considerable vari-
ability in utility for tool manufacture and in weathering characteristics.
In the discussions that follow the textural terms used were introduced and
defined by Upchurch et ale (1982). The reader is referred to this text for
details of the various fabrics, inclusions, and fossils contained in chert.
Upchurch et ale (1982) also discussed the role of chert mineralogy and
fabric in weathering. They used the British terminology for weathering,
which is contrary to common usage in the Southeast. They made a distinc-
tion between patination, which is a secondary silicification phenomenon on
the surface of cherts, and cortication, which is secondary development of
porosi ty and recrystallization wi thin the body of the chert specimen.
Therefore, cortex, as used by Upchurch et ale (1982), is the same as patina
in the common local useage. This is an important distinction because pati-
nation is common and yields glossy surfaces which may be easily mistaken
for having been caused by thermal alteration. It is also commonly respon-
sible for the preservation of surface details (flake scars, use wear) on
otherwise thoroughly corticated artifacts. Although soil chemistry, _age,
and other factors are important, the composition and fabric of chert pre-
determines to some extent the nature of weathering of artifacts. Where
appropriate, predictions as to the nature of weathering of the cherts des-
cribed herein will be -included with the discussions of tool manufacturing
characteristics.
Distribution and Modes of Silicification
There is a large number of strata in the Coastal Plain of the South-
eastern United States that contain silicified sediments, including replace-
ments of fossils, replacements of limestone, camented shales _and sand-
stones, and primary deposits of opaline or quartzose sediments. Many of
these materials were used by early man. The sample set provided includes a
diversity of these materials from several different strata.
Cooke (1936, 1943), Heron et al. (1965), Goad (1979), Carver (1980),
and Upchurch et ale (1982) have described many of the occurrences of chert
in the Coastal Plain and recognized many of the different types of silici-
fied sediments. In addition, there are dozens of minor publications that
describe specific localities where chert has been found. These many publi-
cations make it possible to state that certain strata are prone to silici-
128
TABLE 1
LITHIC CATEGORIES RECOGNIZED IN THIS STUDY
(These may correspond to quarry clusters with additional study.
Categories are in stratigraphic order.)
d. Light brown, silici- Allendale County, SC
3
fied grainstone Screven County, GA 1
8. Chalcedonic mudstone Allendale County, SC
9. Flint Berkeley County, SC 1
Beaufort County, SC 2
Georgetown County, SC 1
10. Silicified siltstone Beaufort County, SC
11- Coralgal boundstone Hampton County, SC
Barnwell/Flint River
Barnwell
Barnwell
Barnwell
Black Mingo
Black Mingo
British
Cretaceous
??exotic (poss.
Barnwell or
Flint River)
Flint River
Flint River
Hawthorn
Flint River
Flint River
Flint River
Flint River
Flint River
Flint River
Indeterminate
(Piedmont rock) ,
(Piedmont rock)
(Piedmont rock)
1
4
1
1
2
1
4
1
Number
Samples Formation
Abbeville County, SC
Locations
Where. o ~
129
Lancaster County, SC
Burke County, GA
Lancaster County, SC
Allendale County, SC
Orangeburg County, SC
Burke County, GA
Screven County, GA
Calhoun County, SC
Sumter County, SC
Burke County, GA
Aiken County, SC
c. Spotted silicified
grainstone
Fractured phyllite
b. Banded, silicified
grainstone
1. Sheared siltstone
2. Lake Secession Jasper
5. Opaline, foraminiferal
wackestone/packstone
3. Opaline "coquina"
7. a. Silicified grain-
stone
4. Silica-cemented
sandstone
6. Coralgal grainstone
Lithic Category
12. Moldic, sandy
limestone
fication in very specific locations. Table 2 lists the major formations of
the Coastal Plain of Georgia, Florida, and South Carolina and indicates
which strata commonly contain chert.
Carver (1980), Strom et al. (1981), Upchurch et al. (1982), and others
have described the mechanisms of formation of the various chert types. The
mechanisms of formation are" distinctive and when superimposed on thedif-
ferent chert-bearing strata, lead to reco"gnizaole" concentrations of chert
Wi th "diagnostic characteristics. Thus, while there is not sufficient data
to characterize quarry clusters, there is enough information to indicate
with certainty that it is possible to develop criteria that will enable'
recognition of quarry clusters.
The stratigraphy of the Coastal Plain of Georgia and South Carolina
has been most extensively described by Cooke (1936, 1943). Cooke's work
has" been the basis of terminology for decades and, while many authors have
suggested significant revisions of the stratigraphy (e.g., Cooke and
MacNeil 1952; Baum et al. 1980), Cooke's terminology remains appropriate
for chert provenance studies and is used herein. More recent stratigraphic
studies have refined Cooke's work, but the basic framework remains. Refer
to Cooke's (1936) map of the distribution of Cretaceous and Tertiary depos-
its of South Carolina for very generalized locations where the deposits
discussed below crop! out.
Chert-Bearing Strata of South Carolina and Eastern Georgia
The basic stratigraphic framework of chert-bearing strata in South
Carolina and adjacent portions of Georgia is described below. Units with
minor or no known chert are omitted from discussion, but are included in
Table 2.
Black Creek Formation (Upper Cretaceous)
The Black Creek Formation appears to be the oldest Coastal Plain
deposit in South Carolina that contains any indication of silicification.
The formation consists of very dark grey, laminated clay and micaceous
sand. No significant chert is known to exist in the Black Creek, but at
Mars Bluff on the Pee Dee River silicified shells are reported (Cooke
1936) This could indicate minor " silicification of clays and/or sand-
stones. Chalcedony casts of mollusk shells are certainly present, but are
of little use for tool manufacture. In all of my examination of Coastal
Plain chert tools, I have only seen one tool made out of this void-filling,
" long fibre chalcedony.
Pee Dee Formation (Upper Cretaceous)
The Pee Dee Formation is highly variable in lithology, but generally
consists of sandy marl with scattered beds of hard marl. Chert is very rare
in the Pee Dee, but is reported in two forms. Chert pebbles are found in
scattered localities, and silicified marl and/or limestone has been
130
o
TABLE 2
o
CORRELATION OF COASTAL PLAIN STRATA OF GEORGIA )ND SOUTH CAROLINA
(Chert-bearing formations are highlighted. Table is
modified from Marsalis and Friddell [1975].)
o
o
u
10
..... -
- ...........
........ -
\
I
\
\
\
\
\
---
... -
-
-
-
.... -
'" -
U
:>
c; CI
--
N
...
...- I
<.>
CI

i
'"
......--
I
...
:&
=<.>
MAW
--
-
--
o
131
reported at three minor localities (Sparrow Swamp, Florence Co.; Black
River, Williamsburg Co.; (?)Bucksville, Waccamaw Co.). The Pee Dee should
not be considered a significant chert source.
Cretaceous Exotics
Cretaceous flints have been reported at many localities along the
Atlantic coast (Cooke 1936; Emery et'ale 1968; Emery 1979). Emery et ale
(1968) and Emery (1979) have attributed the origins of these nodular flints
to ballast stones, gun flints, and other exotic materials imported from the'
Cretaceous chalk deposits of eastern Britain, western France and Belgium.
These flints are diagnostic and are discussed below in detail.
Clayton Formation (Paleocene)
The Clayton is the basal Paleocene stratum in Georgia (Cooke 1943)
It consists of two beds, an upper, calcareous, light gray clay and a lower,
sandy, white limestone or calcareous sand. The Clayton limestone weathers
to produce lumps of chert. The chert ranges from highly weathered tripoli
'to homogeneous, lustrous micromosaic quartz. The chert is yellowish in
color and contains molds of molluscs. It can be found in residuum from the
Clayton in Quitman, Randolph, Stewart, Webster, Marion, Schley, and Sumter
counties. This chert is Widespread and abundant, so it can be expected to
represent a resource to early man.
Black Mingo Formation (Eocene)
The Black Mingo Formation is the lowest Tertiary formation exposed in
the Coastal Plain of South Carolina. It is also one of the most interest-
ing with respect to silicification. The Black Mingo consists of inter-
bedded sands and mudstones. The sands contain lenses of opalized coarse
molluscan grainstone ("shell hash" or "coquina") and the clays may be
opalized.
Locali ties that contain silicified sediments have been described by
Cooke (1936), Heron et ale (1965), Colquhoun et ale (1969), Heron (1969),
and Carver (1980). The great diversity of localities that contain opaline.
sediments in association with zeolites, magnesian clays (palygorskite,
montmorillonite), and carbonate lenses suggests that the environments of
deposi tion or of diagenesis that resulted in silicification were wide-
spread. Thus, until further investigation can prove otherwise, one should
assume that the outcrop belt of the Black Mingo includes a potential chert
quarry cluster.
The cherty materials derived from the Black Mingo are not generally of
the sort considered desirable for lithic tool manufacture. Opaline sedi-
ments are soft and brittle, and opal-cemented clays or sands are aniso-
tropic and inhomogeneous. Upchurch et ale (1981) have noted, however, that
opaline rock (porcelani te) similar to that found in the Black Mingo was
used by prehistoric Indians in central Florida. Chert that sounds as if it
132
-
might be appropriate for tool manufacture is reported by Cooke (1936) from
the Black Mingo near Rimini and Spring Grove Creek, Clarendon County, SC.
In western Georgia the equivalent of the Black Mingo is the Wilcox
Group. Wilcox cherts are rarely found in Clay County (Cooke 1943).
Congaree Formation (Eocene)
The Congaree Formation, and related Warley Hill Formation, crops out
in Calhoun County, SC. The Congaree consists of a clayey siltstone with.
interbedded fine sand and silty clay. According to Colquhoun et al.
(1969), theCongaree, Warley Hill, McBean (see below) and Barnwell (see
below) are laterally equivalent to the Santee Limestone.
The Congaree and Warley Hill Formations contain opaline sediments,
especially opal-cemented clays, which do not appear to constitute a sig-
nificant chert source for early man. These opaline sediments have been
described by Heron et al. (1965), Colquhoun et al. (1969), and Carver
(1980).
McBean Formation (Eocene)
The McBean Formation consists of interbedded fine- to medium-grained
sand; thin beds of greenish, glauconitic marl; laminated, flaky clay; and
silicified limestone. Silicified sediments in the McBean indicate that
some of the environments of deposition during McBean time produced primary
silica deposition (Houser 1982). Upchurch et al. (1982) have shown that
similar systems in Florida produced significant deposits of high-quality
chert when juxtaposed with limestones. A similar situation can be pre-
dicted for the McBean. .
Exposures of silicified material from the McBean have been described
by Coo ke ( 1936) , Heron et al. ( 1965), Colquhoun et al. (1969), Houser
(1982), and others. Many of the exposures described are silicified clays
or marls and probably do not contain material sui table for tool manufac-
ture. The silicified limestones probably are suitable in some areas, how-
ever, and these exposures should be examined for evidence of use by early
man. Exposures of silicified limestone are reported by the above authors
in Aiken, Orangeburg, and Calhoun counties. Silicified shells can be found
in Washington County, Georgia (Cooke 1943).
Jackson Stage (Eocene) and Oligocene Deposits
Before discussing. the various from the Jackson stage of the
Eocene and from the Oligocene it is important to note the confusion in
stratigraphic nomenclature present in the literature. Jackson stage strata
include, in ascending order, the Clinchfield Sand, Twiggs Clay, Irwinton
Sand, Sandersville Limestone, and, in Georgia, the Cooper Marl. The latter
three units are commonly mapped as the Barnwell Formation. The Congaree,
Warley Hill, McBean, and Barnwell appear to be laterally equivalent to the
Santee (Colquhoun et al. 1969). The Black Mingo may also be equivalent to
133
portions of the Santee. In Georgia the name Cooper Marl includes Jackson
age (Eocene) strata. In South Carolina, the Cooper Marl is Oligocene and
is equivalent to the Flint River and Suwannee Limestone of Georgia (Table
2) (Ward et ale 1979). Fossil evidence is poor in many localities and the
stratigraphic positions of various units as well as their ages remain un-
clear. Therefore, assignment of chert materials to a particular formation
is risky and subject to revision.
Santee Limestone (Eocene)
The Santee Limestone is laterally equivalent to the Congaree and
McBean Formations. The Santee is a white to creamy yellow, bryozoan pack-
stone wi th significant moldic porosity, Pecten and other pelecypods are
common (Baum et ale 1980; Powell and Baum 1981, 1982). There is no mention
of chert or other silicification in the Santee, but equivalent strata in
Florida are silicified.
Clinchfield Sand (Eocene)
The Clinchfield Sand is the basal Jackson unit in central and eastern
Georgia. It does not seem to contain significant chert, but is important
because of the silicification i ~ certain clay units. In Richmond County
the Clinchfield includes a spiculi tic clay with a thick layer of opal-
cemented spiculite, sandy spiculite, and sand (Carver 1980). Exposures of
this material occur along Brier Creek in southern Richmond County, Georgia.
Barnwell Formation (Eocene)
The Barnwell consists of red, fine to coarse , massive clayey sand
(Colquhoun et ale 1969) with interbedded clays and limestones. Within the
Barnwell there is a major clay stratum called the Twiggs Clay Member by
Cooke (1943). More recent stratigraphers have elevated the Barnwell to
group status or they have dropped the term altogether. The Barnwell
"Group" includes the Twiggs Clay, Irwinton Sand, Sandersville Limestone,
and, possibly, the Cooper Marl. The Twiggs is an important fuller's earth
deposit and is known to contain significant silicification. The Barnwell
and Twiggs are best developed in Georgia. The Barnwell crops out in west-
ern South Carolina and in a large part of eastern Georgia, where differen-
tiation into the different members or formations is d i f f i ~ l t
Carver (1980) has described silicified rocks from the Twiggs in
Georgia. These rocks are opaline and consists of opal-cemented clays and
opalized pelecypod grainstones ("coquinas") up to 2 feet thick. Cooke
(1936) described silicified clays and limestones from the Barnwell in Aiken
and Lexington counties, SC. "Buhr sands" or silicified (opalized) sand-
stones are noted in the Barnwell by Cooke (1936). There is significant
chert derived from the top of the Sandersville Limestone.
Cooke (1936) reports "chert" (replaced limestone) from the Barnwell in
Allendale and Barnwell counties, SC. These cherts contain Lepidocyclina
and other microfauna typical of the Ocala Limestones to the south in
134
o
D
U
o
U

U

G

Ill.
II

Ig
I
Q
D
Florida. Cooke (1943) attributes the cherts of Burke County, GA to the
Sandersville Limestone. Some may occur in the Eocene Cooper Marl. Similar
replacements of limestones occur in Jefferson County. Undoubtedly, the
silicified limestones in the Barnwell are thin tongues of Cooper/Santee-
like limestones.
This needs further examination because the Flint River cherts (see
below) occur in the same area. If the Barnwell cherts containubiqui tous
Lepidocyclina sp., as do the equivalent Ocala Limestone cherts, then it
will be possible to define rather closely chert provenance within the major
chert-source area of South Carolina. The Flint River is Oligocene in age, ,
so fossil content may make it possible to differentiate the two different
cherts. It is also possible that, given the poor quality of exposures in
Allendale and Burke counties, Cooke incorrectly differentiated the chert.
Cooper Marl (Oligocene)
The Cooper Marl overlies the Santee and Barnwell Formations in South
Carolina. It is a greyish-green marl with up to about 75% calcium carbon-
ate. The Cooper Marl contains traces of phosphate. Like the Santee, the
Cooper Marl is not known to contain chert, but chert may occur. Care
should be taken in evaluating the literature as the Cooper Marl of South
Carolina is apparently not equivalent to the unit of the same name in
Georgia. Thus, confusion in 'stratigraphic nomenclature has arisen in
southwestern South Carolina and southeastern Georgia.
Flint River Formation (Oligocene)
The Flint River is a foraminiferal grainstone to packstone. It is
equivalent to the Suwannee Limestone of Florida and south Georgia. The
Flint River is thought to be exposed in South Carolina only in a limited
area near the Savannah River in Allendale County, SC. It is also exposed
in adjacent Screven County, GA and at other localities in western Georgia.
The Flint River Formation in Allendale County is known to contain a
characteristic, bright yellow, vitreous chert, which is known as "Allendale
chert" (Novick 1978; Elliot and Doyon 1981; Anderson and Schuldenrein
1983). The Allendale County exposures seem to have been significant sources
of chert to early man (Goodyear, personal communication) Review of the
literature for similar chert types suggests that some of the cherts attrib-
uted by Cooke (1936) to the Barnwell have the same. appearance. Similar
cherts are common in western Georgia. Therefore, a detailed study of
tI Allendale-type" cherts is needed prior to final acceptance of the concept
that all yellow, aphanitic, vitreous cherts found in archeological sites in
the Piedmont and Coastal Plain of South Carolina are from Allendale County.
In Georgia the Flint River is especially well developed along the
flint River near Bainbridge. It is important to this study to note that
the Flint River contains calcareous algae at its type section near Bain-
bridge. Other counties in which the Flint River is known to contain chert
are Seminole, Decatur, Grady, Mitchell, Worth, Randolph, Sumter, Pulaski,
135
The Hawthorn Formation consists of a complex mixture of phosphatic
sands, clays and dolostones and limestones.. Silicified, especially
opalized, sediments are common throughout.
Wilcox, Laurens, and Screven.
minifera are reported.
Hawthorn Formation (Miocene)
In all of these localities orbi toid fora-
The Hawthorn crops out throughout Florida, Georgia, and South Caro-'
lina. It is a source of commercially important fuller's earth in Florida
and Georgia and is being mined for phosphate in Florida. The best expo-
sures are along the Georgia banks of the Savannah River in Screven County
and downstream. Silicified clay is also well developed in Beaufort County
(Cooke 1936). Carver (1980) notes the presence of opaline chert on the
Altamaha River, just below the confluence of the Oconee and Ocmulgee
rivers. Cooke (1943) mentions silicified clay in Lowndes and Screven coun-
ties.
Silicification is common in the Hawthorn and includes opaline cements
in clays and sandstones, replacements of carbonates, and isolated chalce-
dony casts of voids. Modes of silicification in the Hawthorn have been
discussed by Upchurch et ale (1982). Carver (1980) has described the modes
of opalization in the Al tamaha Grit, a sandy horizon on the Hawthorn, in
Georgia. None of the silicified Hawthorn materials known from Georgia or
South Carolina are optimal for tool manufacture. Upchurch et aI" (1981)
have described tools made from opaline replacements of dolostone from the
Hawthorn in Florida, however. '
Materials Descriptions
Category 1. Sheared Siltstone and Phyllite (Figs. 1a and 1b)
Sample Numbers and Locations - Two samples were included in the sample set
that are from the Piedmont province. These samples show obvious evidence
of having been involved in deformation. Sample 16 (USF SC-24a) is from the
southeastern section of Lancaster County, SC, near the Lynches River. Sam-
ple 18 (USF SC-26a) is from 38LA108, Lancaster County, SC.
Formation and Distribution - Both samples are from the Slate Belt of the
Piedmont Province (Overstreet and Bell 1965). These are typical Slate Belt
rocks and similar materials are probably Widespread in the Piedmont.
Hand Sample Description- Sample 18 is a sheared, greenish, fine metasilt-
stone or argillite. It contains obvious shear zones and a very faint
foliation. Sample 16 is a gray-brown, homogeneous microcrystalline meta-
morphic rock and would probably be called an argillite. It has thin ten-
sion gashes or f r t ~ r e s that are filled with whitish material.
Thin Section Description - In thin section Sample 18 is foliated with mus-
covite and/or sericite and fine quartz. The quartz shows a clastic texture
136
-
and indicates a sedimentary for the protolith. The matrix is micro-
mosaic quartz. The tension gashes/fractures are filled with the classic
quartz void-filling sequence (chalcedony-megaquartz). Sample 16 shows
more-or-Iess randomly oriented micas (sericite?) in a siltstone fabric.
Quartz is present as microgranular cement.
Predicted Weathering Characteristics - Both samples contain microcrystal...
line quartz and, therefore,. undergo cortication. Hand samples show white;
"tripolitic" of cortex.
137
Category 2. Lake Secession Jasper (Figs. 2a and 2b)
Utility for Tool Manufacture - The sample studied is extremely homogeneous
and individual quartz grains are well interlocked. Therefore, this mate-
rial should be ideal for tool manufacture. It should work and wear well.
One sample (Sample 17; USF SC-25a) was
It is from Lake Secession, Anderson/Abbe-
Thin Section Description - In thin section the rock appears iron rich. The
iron is present as filaments and clumps of goethite needles. The matrix
ranges from micromosaic to fine microgranular quartz. There is too much
iron present for fabric recognition, but relict microfossils do appear to
be present. If so, this in an anomalous rock for the Charlotte Belt and/or
Inner Piedmont. There are many small, hexagonal platelets which are too
small for identification with a petrographic microscope. These could be
either apatite or cordierite.
Predicted Weathering Characteristics - The sample studies did not have a
weathering rind on it. However, micromosaic quartz will develop cortex and
in acid soils the iron might be removed.
Utility for Tool Manufacture ... These Piedmont rocks would suffice for'
"expedient" tools only. The foliations are weak and the fractures well
cemented, so the material is relatively strong. The foliation and frac-
tures would, however, cause the material to have a "grain" which would make
knapping difficult. The- mica flakes are soft and have a good cleavage, so
the material would wear quickly with use.
Hand Sample Description - In hand sample the Lake Secession jasper has a
waxy lustre. It consists of a clear quartz "matrix" with lacy filaments
and clumps of yellow iron oxides. The iron oxides are dense enough to
prevent the rock from transmitting light along thin edges. The rock has
cracks in it and there appear to be spherical clots of material.
Sample Numbers and Locations -
labelled Lake Secession jasper.
ville counties, SC.
Formation and Distribution - Lake Secession jasper is a Piedmont rock and
its extent as a characteristic material is Jasper and other forms
of siliceous, fine grained rocks are common in the Piedmont (cf. Novick
1980). Lake Secession is near the Charlotte Belt/Inner Piedmont contact.
IrJ .....
01
D
IrnJ .. :.:.
i
-Id
I
Irl.'
I
11lI.:
WI
I
I
Q
o
Category 3. Opaline "Coquina" (Figs. 3a and 3b)
sam
1
le Numbers and Locations - Two samples fell into this category. Samples
31 USF SC-39a) and 21 (USF SC-29b). Thirty-one is from 38CL17, Calhoun
County, SC, and 21 is from 38SU42, Sumter County, SC.
Formation and Distribution- Both samples are derived from the Black Mingo
"Formation. The material represented is not uncommon in the Black Mingo and
many localities of "opalized shell" are reported by Cooke (1936; 1943).
Other, better chert materials are also found associated with this material.
Hand Sample Description - In hand sample Sample 21 is a coarse pelecypod
"coquina" or grainstone. The pelecypod fragments" have been replaced by
waxy appearing quartz. Coarse quartz sand is included in the rock. The
rock-is vuggy. Sample 31 is similar in appearance with a coarse pelecypod
grainstone fabric. The sample is transparent brown in color. It has a
waxy lustre and quartz sand inclusions. The mollusk fragments are imbri-
cated. The rock is well cemented and breaks across grains.
Thin Section Description - In thin section these samples are remarkable in
the detail of pseudomorphic replacement of structure in the mollusk frag-
ments. Replacement must have taken place very shortly after deposition.
Sample 21 contains quartz and opal in a pelecypod grainstone. It is very
moldic. Relict micrite envelopes preserve the original shapes of mullusc
fragments. Rice grain particles of quartz indicate that the quartz was
once opal and has inverted. Opal-CT lepispheres occur in the cement of the
rock. Voids are filled with chalcedony and micromosaic quartz (not mega-
quartz). Sample 31 shows considerable pseudomorphic replacement of the
mollusc shells by microcrystalline quartz. Relict and opal lepispheres
line voids, which have a normal infilling sequence (chalcedony, mega-
quartz).
Predicted Weathering Characteristics - The opal and micromosaic quartz
present in these rocks would predispose them to rapid cortication.
Utility for Tool Manufacture - The coarse, imbricated mollusk fabric should
render this material unsuitable for tools. If, however, the material is
solid and minimal vuggy porosity existed, it could be used. This material
is very diagnostic and should be easily recognized in Indian lithic arti-
facts.
Category 4. Silica-Cemented Sandstone (Figs. 4a and 4b)
Sample Numbers and Locations - Five samples were examined that belong in
this category. These samples are: (1) 8 (USF SC-10d), 38AL149, Allendale
Co., SC; (2) 9 (USF SC-18a), east side of Edisto River and north side US
301, Orangeburg Co., SC; (3) 7 (USF SC-9c), Sandoz #7, Allendale Co., SCi
(4) 6 (USF SC-16e), Sandoz #6, Allendale Co., SCi and (5) 24 (USF SC-32a),
38AL1, Little Hell Landing, Savannah River, Allendale County, SC.
Formation and Distribution - Assignment of these samples to a specific
formation is difficult. Wi th the exception of sample 9, all samples are
either Flint River or Barnwell (IrWinton Sand?). In some areas of Georgia
138
and Florida the Flint River (or its equivalent Suwannee Limestone) contains
quartz sand. Cooke (1936) also suggests that the area from which the sam-
ples were derived is characterized by Flint River. Barring field confirma-
tion, one can only speculate that these sands are not Flint River, but are
Barnwell. Similar sands have been described by Carver (1980) from the
Clinchfield. Sample 9 is slightly different petrographically (see below)
and contains coralline algae clasts mixed with the quartz sand. Cooke
(1936) places the area from which the sample was derived in the McBean. No
coralline algae is reported from the McBean and, while opal cemented sand
is found there, this provenance is questionable. It is more likely that
the opaline sand is Barnwell or Flint River residuum or an erosional out-
lier. The "Flint River" Allendale cherts (see below) contain coralline
algae in abundance.
Hand Sample Description - All of the samples feel "gritty." They contain
coarse, angular, poorly sorted quartz sand with a microcrystalline quartz
cement. There are rock fragments, feldspars, and black, opaque minerals
present as well. The sand grains break across the grains, indicating a
firm, solid cement. Sample 6 contains silicified fossils, including
bryozoa, and is moderately sorted.
Thin Section Description - The samples are angular ,moderately to poorly
sorted sandstones, predominantly composed of quartz. The cement is chal-
cedony in sample 24, micromosaic quartz in 9, opal-CT and micromosaic
quartz in 8. Cement in samples.. 6 and 9 was once opaline as indicated by
the shrinkage cracks and lepispheres in voids. Sample 6 has rims of
length-slow chalcedony and/or fibrous opal-CT. Quartz sand clasts have
micromosaic rims in 9 and 6. Sample 9 includes sponge spicules (the source
of the silica for the opal-CT), replaced mollusc fragments, and possible
coralline algal fragments. Peloids and clotty fabrics occur in8 and 9.
Predicted Weathering Characteristics - The quartz sand in these rocks. is
relatively inert and should not weather. The cement is a different matter
and should weather to cortex rapidly. The result would be disintegration
of the rock.
Utility for Tool Manufacture - This material is well cemented in many cases
and might oe knapped. The contrast in hardness of the quartz and opal/
chalcedony/micromosaic quartz cement would hinder fine flaking. This mate-
rial is ideal as an abrasive and is used as a buhrstone in some areas.
Category 5. Opaline, foraminiferal wackestone/grainstone (Fig. 5)
Sample Numbers and Locations - Only one sample fits this category. That
sample, 22 (USF SC-30a) is from site 9BUR61, near Girard, GA., on Brier
Creek and Ellison Bridge, Burke County, Georgia.
Formation and Distribution - The location and opaline nature of the rock
indicates that it was derived from the Barnwell. It is tentatively opal-
ized Sandersville Limestone. The extent of this material is undetermined.
A literature survey fails to discuss it, so one can only assume that it is
volumetrically unimportant.
139
Hand Sample Description - This rock is not a good sample with which to work
owing to its condition. The hand sample is crazed, oxidized, and glossy,
which indicates thermal alteration. It is vuggy, with stalactitic chal-
cedony void fillings. It contains quartz sand inclusions, which suggests
affinity with the Barnwell opalized sandstones. Much of the rock appears
to be a foraminiferal grainstone with patches of packstone and/or wacke-
stone. The cement is mostly opaque and appears to be micromosaicquartz.
Thin Section Description - This is a replaced foraminiferal grainstone.
The matrix is a mixture of opal lepispheres and micromosaic quartz. "Rice
grain" quartz and shrinkage cracks further indicate that this material has
a opal precurser.
Predicted Weathering Characteristics - The sample has two cycles of corti-
cation. There is an.old rim of cortex, which is probably natural
ing material. The fresher surfaces are just developing a thin, new patina.
The weathering pattern suggests that this has been worked by man.
Utility for Tool Manufacture - The opal in this rock is brittle and rela-
tively soft. It is homogeneous, however, so it should work well. It is
somewhat surprising that thermal al is evident and one can only
surmize that it was accidental. The water content of opal would cause
crazing with thermal alteration and the brittle nature of opal should
obviate the need for thermal treatment.
Category 6. Coralgal Grainstone (Fig. 6)
Sample Numbers
algal remains.
Quarry No.1,
Formation and Distribution - _The sample appears to have been derived from
the Barnwell or Flint River. The presence of coralline algae, which has
been attributed to the Barnwell in other materials and is absent in the
"Allendale chert," suggests that this is Barnwell, perhaps Cooper Marl
(Georgia useage) or Sandersville Limestone. The extent of the material is
indeterminate.
Hand Sample Description - The most obvious material in this unit is the
upright and crustose coralline algae. The rock is a coralgal boundstone to
grainstone. Matrix is opaque, microcrystalline quartz. The rock is-color-
ful, with green, red, and white. Iron oxides are present and stain parts
of the rock. The reddish color may suggest thermal alteration. Gloss is
present in scattered patches on the rock. Color and gloss are insufficient
criteria to attribute thermal alteration, however.
Thin Section Description - In thin section the rock is a poorly sorted,
coralgal and molluscan boundstone to grainstone. It is replaced by micro-
crystalline quartz and chalcedony.
Predicted Weathering Characteristics The fine grained quartz should
weather moderately fast and produce cortex ("tripoli").
140

An Archeological Survey of Chert Quarries in Western Allendale Co

Caricato da

Informazioni sul documento

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

An Archeological Survey of Chert Quarries in Western Allendale Co

Caricato da

Copyright:

Formati disponibili

University of South Carolina

r-I () () QJ ..c: '0 0 'r!

QJ 01-1 r-I r-I ClI QJ

+J I-l ell ell I-l

,.!l4 ,.!l4 Cf)

I'Ll U tI.l E-t D P:l D U u ::Il ct .E-t E-t

A III ,.c::. III

III .j,J r-I .-l III

III I-l I-l -r-! (J

(/) CIl CIl

CIl r-l CIl Q) r-l

a 0 .j.I ClI r-l 0 ll-l OJ ..0

,.!.:\ 0 tJ 'rl 0 lJ.l ,.0

1-1 r-I 0 ::1 0 a

r-I rz. r-I CIS .0

Q) .c.J r-I r-I

CI.l Cll CI.l

CI.l r-l CI.l .Q)

til 8 l=l l=l Co) Co)

r.r. u Pol tJ:l E-c j:Q

E-l j:Cl P-l

1J:l. :>. E-l "Cl Q)

E-i j:Q P-i

+J $-I. lIS lIS $-I"

s:: 0 0 lIS $-I r-i 0 ::l 0 0

t..) Pol CIl E-i

Q) -.-I ..c: I-l Q) Q)

s:: 0 0 ttl I-l r-I 0 0 0

Potrebbero piacerti anche