Archaic Shell Rings
Archaic Shell Rings
Archaic Shell Rings
C. Form Prepared By
Michael Russo, Ph.D., Archaeologist April 2006
Southeast Archeological Center, National Park Service Tel: 850-580-3011, ext. 344
2035 E. Paul Dirac Drive, Johnson Building, Suite 120 Fax: 850-580-2884
Tallahassee, FL 32310 [email protected]
D. Certification
As the designated authority under the National Historic Preservation Act of 1966, as amended, I hereby certify that this documenta-
tion form meets the National Register documentation standards and sets forth requirements for the listing of related properties
consistent with the National Register criteria. This submission meets the procedural and professional requirements set forth in 36
CFR Part 60 and the Secretary of the Interior's Standards and Guidelines for Archeology and Historic Preservation. (See continua-
tion sheet for additional comments.)
__________________________________________________________ ____________________________________
Signature and title of certifying official Date
__________________________________________________________ _____________________________________
Signature of Federal Agency or Tribal Government Date
_________________________________________________________________________________________________________
I hereby certify that this multiple property documentation form has been approved by the National Register as a basis for evaluating
related properties for listing in the National Register.
__________________________________________________________ ____________________________________
Signature of Federal Agency or Tribal Government Date
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Contents
E. Statement of Historic Contexts ............................................................................................................8
Introduction.............................................................................................................................................8
Context 1 Archaic Shell Rings as Early Large-Scale Architecture......................................................8
Context 2 Mounds and Other Architecture Associated with Shell Rings..........................................27
Context 3 America’s First Potteries, 5000–3000 B.P. ........................................................................31
Context 4 Development of New Technologies: Shell & Bone Tool Kits .........................................35
Context 5 Expansion of Exchange Networks ....................................................................................44
Context 6 Establishment of Sedentary Cultures ................................................................................45
Context 7 Changing Mortuary Practices............................................................................................46
Context 8 Establishment of Tribal Identities .....................................................................................51
Context 9 Settlement and Environment .............................................................................................59
Context 10 Sociopolitical Development and Shell Rings....................................................................60
Tables
1. Radiocarbon Dates from Shell-Ring Sites.............................................................................. 11–16
2. Shell-Ring Metrics ................................................................................................................. 25–26
3. Pottery .......................................................................................................................................... 37
4. Projectile Points and Other Lithics............................................................................................... 38
5. Shell Tools.................................................................................................................................... 41
6. Bone Pins and Other Artifacts...................................................................................................... 43
7. Human Remains and Burials from Shell Rings and Nearby Sites ............................................... 48
8. NHL Shell-Ring Historic Contexts and NPS Thematic Framework Themes .............................. 63
9. NRHP Listed Shell Rings and Possible Shell Rings .................................................................. 107
10. Shell Rings Considered Potentially Eligible for NHL Status..................................................... 112
11. NRHP and NHL Status of Sites Identified as Shell Rings by Investigators ...................... 113–114
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Introduction
Since 1998, the National Park Service has been developing a National Historic Landmark Historic Con-
text on Late Archaic (5000–3000 B.P.) shell ring sites in the southeastern United States. The purpose of
this study is to provide the archeological and historical context for nominating nationally significant Late
Archaic shell-ring sites for designation as National Historic Landmarks. Shell rings are circular to horse-
shoe-shaped piles of shell (primarily oyster) ranging in size from 50 to up to 250 meters across and lo-
cated along the coasts of South Carolina, Georgia, Florida, and Mississippi. Hunter-gatherer societies
became increasingly complex during the Late Archaic in the southeastern United States. Monument con-
struction, such as shell rings, and new technologies like pottery are evidence of this complexity. Evi-
dence of the first permanent settlements, the first inter-coastal exchange patterns, the development of
founding new coastal cultural traditions and the increase in social complexity can all be studied through
examination of shell ring sites.
Accompanying this historic context as an example of how evaluations of shell ring sites for National His-
toric Landmark status are carried out is the National Historic Landmark nomination of the Fig Island
shell-ring complex (38Ch42). Two of the Fig Island rings (2 and 3) were listed in the National Register of
Historic Places in 1970. Recent investigations reveal that the site is larger and more complex than was
known in 1970. In fact, the site is the best preserved example of early large-scale monumental architec-
ture on United States coasts. By volume, the complex is larger than any other single Archaic shell-ring
site.
The ten historic contexts identified in Section B are described herein. These establish as nationally sig-
nificant the contexts of architecture, settlement, exchange, technology, ethnic identity, cultural traditions,
and social organization as they relate to Late Archaic shell rings. All are intrinsically tied into the NPS’s
Thematic Framework, which includes the broader applicable themes of Peopling Places, Expressing Cul-
tural Values, Expanding Science and Technology, and Transforming the Environment. Shell rings hold
superlative positions relative to the NPS’s thematic framework because rings represent the first, earliest,
oldest, largest, or most complex examples of architecture, settlement, exchange, technology, ethnic iden-
tity, cultural traditions and social organization of this period in this large region of the United States and
thus constitute national levels of significance.
Shell rings have been identified as nationally significant based on their architectural form and early con-
struction. Preceding the Woodland and Mississippian mound-building periods by thousands of years,
shell rings are among the earliest large-scale architectural features found in the United States. The Rollins
and Oxeye shell rings respectively represent the largest and oldest standing architecture in the National
Park Service (NPS) (Figure 1).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
FLORIDA
SOUTH CAROLINA
3
Skull Lighthouse Horse Chester
Creek Point Island Field 2
Fig 1
Island
Sea Buzzard Stratton
Pines Creek Place
3 Coosaw
4
1 2
Sewee Hankel Auld
Sapelo 2 0 100m
GEORGIA
1
S. Carolina
Georgia Horse Auld
Sewee
Alabama Chester Field
Island
Lighthouse Point
Skull Creek Fig Island
Mississippi Walthour & Oemler Coosaw
Skidaway Rings Barrows & Patent
Sea Pines
West Busch Krick
Cannon's Point Sapelo
Bony Hammock
Rollins
Cedarland Buck Oxeye Guana
& Meig's
Pasture Bayou
Claiborne
Florida
Hill Cottage
Joseph Reed
Bonita Bay
Horr's Island
undated
rings
From South Carolina to Mississippi, Archaic shell rings are found among coastal settings on islands, the
mainland, river banks, and buried beneath marshes (Figure 2). Wherever oysters were abundant 3,000 to
5,000 years ago, Late Archaic people exploited them (along with other resources). In areas where the use
of shell rings was a tradition, ring builders deposited the shells in circular and semi-circular piles ranging
in size from 30 to 250 meters in diameter and 1 to 6 meters in height. These enormous monuments are so
me of the earliest large-scale public works in the United States. There is little to no evidence of significant
human occupation in the coastal zone of the southeastern United States before shell rings appeared. Most
Middle Archaic (8000–5000 B.P.) peoples of the southeastern United States are best known from non-
coastal settings and are thought to have been migratory hunter-gatherer bands that moved among upland
and riverine environments in rhythm with the seasons. By the Late Archaic (5000–3000 B.P.), however,
the first extensive evidence of significant human occupations appear on the coast. Late Archaic coastal
sites vary from isolated finds, small camps, and minor middens to large amorphous shell middens. But the
founding coastal settlements are made most conspicuous by the presence of the large shell rings, which
have yielded over 100 radiocarbon ages (Table 1).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
South Carolina
30'' above charcoal (I-2850) in Calmes 1968:25 (162); Buckley and Willis
I-2849 oyster 0 3120 110 3530
periwinkle layer and 27'' bs 1969:79
level 9, 56–57" bs, bottom half of
I-2850 charcoal -25 3585 115 3585 '' '' ''
shell deposits
Patent (38BU301)
Beta-213397 Base of shell, NE, 30–40 cmbs oyster -1.5 3280 70/80 3660 '' ''
GX-29192 EU1 base, 90–95 cmbs oyster -2 3420 70 3790 Heide 2003:9; Russo and Heide 2003:31
GX-29193 EU2 base, 110–120 cmbs oyster -2.1 3190 70 3560 Heide 2003:11; Russo and Heide 2003:31
GX-29527 EU2 top, 25–30 cmbs oyster -1.8 3230 70 3610 '' '' ''
CAMS 87990 EU2 90–100 cmbd quahog 0 NA 30 3800 Russo and Heide 2003:31
GX-29194** EU3 base, 25–30 cmbs oyster -2.5 3440 70 3810 Heide 2003:13; Russo and Heide 2003:31
Uga 2901 230R60, Level 2 charcoal -25 3190 70 3190 Trinkley 1980:209–210 (191–192)
Uga 2902 230R70, Level 2 charcoal -25 3275 55 3275 '' '' ''
Wk-9746 TU2, 90 cmbs oyster -1.1 3467 46 3861 Saunders 2002:114; Russo and Heide 2003:15
Wk-10103 TU2, top oyster -0.9 3420 54 3816 '' '' ''
Wk-10105 TU1, top oyster -0.5 3550 47 3953 '' '' ''
Wk-9763 TU5, Posthole test oyster -0.6 3627 50 4030 Saunders 2002:114; Russo and Heide 2003:15
Wk-9747 TU2, Feature 1 base oyster -0.8 3594 49 3993 '' '' ''
Wk-10104 TU 1, 23–30 cmbs oyster -0.4 3667 48 4074 '' '' ''
GX-2279 NE Quadrant, C–1, 2' bs oyster 0 3295 110 3675 Trinkley 1980b:5; Russo and Heide 2003:15
GX- 30186 EU1, 33–48 cmbd oyster -1.8 3630 70 4010 Russo and Heide 2003:14–15
GX- 30187 EU1, 150 cmbd, ring base oyster -2.3 3740 70 4120 '' ''
Georgia
Georgia (cont.)
UGA-73** 1 mbs in ring 50 m diam, 2–3 high oyster 0 3430 65 3840 Noakes and Brandau 1974:133
UGA-74** 2 mbs in ring 50 m diam, 2–3 high oyster 0 3430 70 3840 '' ''
UGA-15084 Unit 1, Level 2, 10–20 cmbs sooted sherd -17.04 3480 50 3610 Thompson 2006:183
UGA-15085 Unit 1, Level 2, 10–20 cmbs sooted sherd -18.94 3630 60 3730 '' ''
UGA-15086 Unit 11, Level 4 charcoal -25.57 3740 50 3730 '' ''
UGA-15081 Unit 3, Lev 10, 90–100 cmbs charcoal -26.19 4080 50 4060 Thompson 2006:183
UGA-15087 Unit 2, Lev 9, 80–90 cmbs charcoal -25.03 3070 50 3070 '' ''
Florida
Beta 35347 Mound B, burial, FS 533 human -13 4030 230 4230 Russo 1991:423–424; Russo 1994:90
Beta 40276 Mound B, Stratum G, FS 369 charcoal -25 6070 90 6070 '' '' ''
UM 1918 Mound C, Stratum A whelk 0 4460 105 4860 '' '' ''
Florida (cont.)
Beta 35344 Mound A, Z1, FS 464 human -13 3420 100 3620 '' '' ''
Beta 35345 Mound A, Z5, FS 501 charcoal -25 4760 170 4760 '' '' ''
Beta 35346 Mound A, Z 10, FS 507 charcoal -25 4270 60 4270 '' '' ''
Beta 36466 Mound A, Fire Pit, FS 243 charcoal -25 4140 60 4140 '' '' ''
Beta 36467 Mound A, Z 3, FS 462 charcoal -25 4260 80 4260 '' '' ''
UM 1923 Mound A, Zone 1, Stratum A cockle 0 4335 70 4735 '' '' ''
UM 1924 Mound A, Zone 2, Stratum B oyster 0 4025 75 4425 '' '' ''
UM 1925 Mound A, Zone 4, Stratum A oyster 0 4055 75 4455 '' '' ''
Beta 1274 Test 7, Stratum D oyster 0 4100 110 4500 '' '' ''
Beta 1275 Test 6, Stratum D oyster 0 3885 100 4285 '' '' ''
Beta 1276 Test 11, Stratum D oyster 0 4070 80 4470 '' '' ''
Beta 1277 Test 11, Stratum B oyster 0 4260 90 4660 '' '' ''
Beta 1278 Test 11, Stratum A oyster 0 3790 85 4190 Russo 1991:423–436
Beta 35348 Mound D, submound, FS 587 charcoal -25 4450 190 4450 Russo 1991:423–424; Russo 1994:90
Oxeye (8DU7478)
Beta 119814 ST 1262, 2 mbs oyster -1.8 4500 80 4580 Russo and Heide 2000:57
WK7437 EU5m 10–15 cmbs estuarine shell 0 3990 60 4400 '' ''
Beta 119815 Trench 1, Unit 5, btm of shell oyster -4.1 4230 70 4570 '' ''
Beta 47531 TP3, 60–80 cmbs oyster -1.9 3990 70 4370 Russo 1992:110; Russo 1996:182–183
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Florida (cont.)
WK7433 Unit 3197, 10–20 cmbs, midden oyster 0 2280 60 2690 Russo and Heide 2000:57
Beta 119817 Unit 3197, 80–90 cmbs, midden oyster -0.3 3300 70 3710 '' ''
Beta 50155 4850N,250E, 60–65 cmbs oyster 0 3350 60 3760 Russo 1992:110; Russo and Heide 2000:57
WK 7438 Trench 1, Unit 1, Feat 1, 35 cmbs oyster 0 3230 60 3600 Russo and Heide 2000:57
Trench 1, Feature 11, base, 200 Geochron Laboratory 1999; Alexander Cher-
GX 25750 bulk carbon -25.6 3740 80 3730
cmbs kinsky, Geochron Lab, to R. Saunders 2006
TU 1097, Ringlet I, pit feature (in Alexander Cherkinsky,Geochron Lab, to G.
GX-29516 oyster -3 2100 70 2460
profile) Heide 2002
GX-30379 TU 11, base of shell oyster -3.6 3278 70 3630 '' ''
Beta 90529 Unit 546–547, E550, 10–20 cmbs marine shell 0 3710 70 4120 Houck 1996:31
Beta 90530 Unit 546–547, E550, 100–110 cmbs marine shell 0 3460 70 3870 '' ''
Beta 48533 FS 17, 0–10 cmbs marine shell 0 3850 70 4260 Dickle 1992:161
Beta 48534 FS 18, 100–110 cmbs marine shell 0 3770 70 4180 '' ''
GX-26118 EU 1, feature 2, 122 cmbd charcoal -26.6 2860 130 2850 '' ''
Beta 21253 Trench 2, Feature 3 conch -0.8 3700 80 4100 Curren 1987:71
Beta 21254 Trench 3, Feature 17 conch -0.8 3670 80 4070 '' ''
Beta 21255 Trench 3, Feature 17 conch -0.8 3630 90 4030 '' ''
Dicarb 3295 B Zone 4 marine shell 0 3280 50 3690 '' '' ''
Florida (cont.)
GX-31906 Feature 1, top oyster -2 2362 70 2740 Saunders and Rolland 2006:7
GX-31907 Feature 2/4, top oyster -1.5 3355 70 3740 '' ''
Beta 166869 340N, 440E clam -0.5 3310 60 3720 '' ''
Saunders and Rolland 2006:7; Russo et al.
Beta 154816 340N, 540E oyster -0.2 3450 60 3860
2002:29
GX-29517 469N, 453E oyster -1.3 3430 70 3820 Saunders and Rolland 2006:7
Saunders and Rolland 2006:7; Russo et al.
Beta 154817 469N, 453E oyster -1.2 3210 50 3600
2002:29
Beta 165598 380N, 400E oyster -2.2 3120 60 3490 '' '' ''
Beta 165599 410N, 520E oyster 0.5 3180 70 3590 '' '' ''
G-596 Test A, 1foot deep busycon NA 3350 120 4040* Bullen 1976:13
G-597 Test A, 2–2.5 feet deep venus NA 3225 120 3625* '' ''
G-598 Test A, 4 feet deep busycon NA 3575 120 3975* '' ''
G-599 Test A, 8 feet deep busycon NA 4050 125 4450 Bullen 1976:13; Russo 1996:182–183
G-600 Test A, 11 feet deep busycon NA 4100 125 4500 '' ''
Mississippi
Cedarland (22CH30)
G-561 Top of midden charcoal -25 3200 130 3200 Gagliano and Webb 1970:69
Clairborne (22CH35)
I-3705 Base of midden charcoal -25 3100 110 3100 Gagliano and Webb 1970:69
UGA-1693 few cm to more than 50 cm deep charcoal -25? 3385 140 3385 Bruseth 1991:15, 18
Why shell rings are shaped the way they are and why they were of such large size are questions that have
intrigued archeologists. One theory is that shell rings resulted from the deposition of shell in piles and pits
next to house structures (Thompson 2006; Trinkley 1980a; Waring and Larson 1968). As the incidental
debris of daily meals accumulated underfoot, the ring builders moved their domiciles to the top of the ris-
ing rings of shell (Figure 3) (Edwards 1965; Trinkley 1997). Under these interpretations rings represent
the remains of house and circular village foundations. Encircling the interior plaza, they are integral parts
of the architectural landscape, providing a public stage for ceremony and other community interaction.
Such rings have been identified at Sapelo 3, Lighthouse Point, and Stratton Place.
At other ring sites, archeologists have found that houses and hearths in the early stages of ring occupation
disappear during later stages (Russo 1991, 2004; Saunders and Russo 2002). Large-scale ceremonies are
evident in feast remains found in mounded ring walls consisting primarily of piled shell. The public con-
sumption of huge quantities of shellfish, fish, and other foods results in the refuse being purposely and
rapidly piled in contiguous formation to make the solid ring walls. Shell deposits are placed so quickly,
steeply and high that they preclude domiciliary occupation on their summits. Steep, tall rings are found at
Horr’s Island, Sewee, Sapelo 1, and Fig Island 1 (Figure 4). At these sites, living may not have occurred
on the ring, but nearby. Evidence of associated villages has been found near Horr’s Island, Rollins,
Guana, and Sapelo 1 (Russo 1991, 1992; Russo et al. 2002; Saunders and Russo 2002; Simpkins 1975;
Thompson 2006). Elsewhere, rings seem to lack associated living areas, for example at Sewee (Russo and
Heide 2003), Reed (Russo and Heide 2002), and Hill Cottage (Bullen and Bullen 1976; Sarney 1994)
Some sites are characterized by multiple rings. These vary greatly in shape and size. Two or more rings of
approximately equal size may be connected as a Figure “8,” such as at Skull Creek 1 and 2 and Coosaw 1
and 2 (Figure 5) (Calmes 1967; Heide and Russo 2003). Some sites may have smaller rings attached to a
large, primary ring, such as at Fig Island 1 and Rollins (Figure 6) where four and nine rings, respectively,
are attached to the main rings (Russo 2004; Saunders 2004; Saunders and Russo 2002). Still others may
exist as a grouping of separate rings in close proximity, such as those found at the Fig Island, Coosaw,
and Sapelo shell-ring complexes where three major rings (not counting attached rings) are found at each
site (Heide and Russo 2003; Saunders and Russo 2002; Thompson 2006). The functions of rings in these
complexes—habitation and ceremony—mirror those identified at single ring sites.
Although the name implies a certain symmetrical circularity, shell rings vary greatly in shape. They can
be C-shaped, U-shaped, oval, or circular, while attached ring shapes defy easy description (Figure 1).
Minimally and common to all is a curvilinear ridge of shell encompassing an area with little or no shell,
generally interpreted as a plaza. Because shell rings often lie in marsh settings, many have been impacted
by coastal erosion; distinguishing their current from their original form must be undertaken with subsur-
face, geophysical, or historical analyses. For example, comparison of the current shell-ring configuration
of the Oxeye site, which lies partially buried in the marsh, was aided by historic aerial photos that show a
greater portion of the circular ring above the marsh line (Figure 7) (Russo 2004). Comparison of Moore’s
(1897) sketch map of Sapelo 1 to Simpkins’s (1975) contour map reveals extensive removal and spread of
shell on the south side of the ring, resulting today in an elliptical ring that was formerly more of a sym-
metrical circle (Figure 8). The Lighthouse Point shell ring is known from historical accounts to have once
been a closed circle of shell (Drayton 1802), although by the time it was investigated in the 1970s, it was
too disturbed from erosion and mining for its shape to be identified (Trinkley 1980a).
Understanding the shape of the rings is important because the social organization of ring builders has
been linked to shape. Trinkley (1980a, 1985) suggests that circular shapes indicate an egalitarian social
formation, implying that no status distinctions are apparent in a symmetrical circle (Figure 3).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
960 980 1000 1020 1040 1060 1080 1100 1120 1140
Figure 7. Oxeye shell ring 1943 aerial (above) showing western ring
above marsh, eastern side below; 1999 probe map (below) showing shell
ring distribution above and below marsh.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
-20 0 20 40 60 80
In line with this reasoning, Russo (2004) has suggested that since a number of rings have extremely
asymmetrical shell distributions, which may reflect asymmetries in social relations. That is, the greater
volumes of shell in certain places may be associated with special status. In C- and U-shaped ring forma-
tions, for example, the closed end of the ring typically contains the greatest volume and height of shell.
Comparing these rings to ethnographic examples of similarly shaped villages, the shell distribution mir-
rors the placement of the architecture associated with high-status individuals at the closed end of village
plazas. The accumulation of more shell in certain ring areas may be related to spiritual oblations or offer-
ings to ancestors if the ring served as a monument or a place of ceremony. Or, the larger piling of shell at
the closed end of rings may be the result of efforts by self-aggrandizers common to “big-men” societies.
Such agents seek status in the display of food wealth, the hosting of community feasts, the construction of
edifices as monuments to their attained status, and their location in the community at that point consid-
ered the most prestigious—all efforts to symbolize their greater status in the community (see Figure 9).
Figure 9. South American Shavante village showing C-shape layout with high status symbolized by place-
ment of single house at closed end of C slightly towards center of plaza (Fraser 1968, fig. 73).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The shape and height of shell pilings has been shown to be distinctively asymmetrical at Sewee (Russo
and Heide 2003), Rollins (Russo and Saunders 1999; Saunders 2004), Fig Island 1 (Russo 2002a), and
Horr’s Island (Russo 1991). Combined with other signs of status, such as ceremonial mounds, these ring
sites have been interpreted as among the earliest evidence for the development of distinct non-gender and
non-age status differences among community members in the United States during the precontact period.
Ring size may be related to population. In Georgia and South Carolina, shell rings are usually circular or
C-shaped and average 53 and 64 meters in diameter, respectively. In Florida, shell rings are typically U-
shaped and average 178 meters in length (Figure 1; Table 2). Whether shell rings were used as villages,
places of ceremony, or both, differences in ring sizes indicate that different numbers of people could have
occupied the site at any one time. Since population size may be correlated with complexities involved in
organizing societies, shell-ring size can potentially shed light on the socio-political developments arising
out of the Late Archaic coastal settlements.
Kelly (1995) has shown that mobile hunter-gatherer groups ethnographically average twenty-five mem-
bers, while settled hunter-gatherer groups can reach up to 1,000 permanent inhabitants. These larger
populations may manage to maintain a basic egalitarian form of social organization, but to do so, they
require the adoption of more complex social maintenance strategies. Positions of temporary or earned
status or authority are often granted, large-scale solidarity ceremonies and feasts are held to reinforce so-
cial bonds, and monuments and markers of group identity and social boundaries may be erected to distin-
guish one group from another. Shell rings such as Horr’s Island, Reed, and Fig Island have been shown to
have supported large, settled populations whose maintenance would have required these or other com-
plexes organizing mechanisms to hold the societies together. As such, the sizes of rings and associated
architecture provide a potential tool for gaining an understanding of the development of social organiza-
tion among incipiently complex societies (Russo 2004).
Along with size and shape, the construction techniques behind shell rings may be linked to the specific
kinds of social organization behind the groups that built them. While the initial stages of the smaller rings
seem to represent little more than the incidental discard of food refuse coincidentally placed in the same
planar layout in which the builders lived, larger rings were built rapidly with the shell remains of feasts.
Shell was discarded in large piles, probably by corporate groups rather than individual households. The
largest of the rings may have wholly or partially been constructed through various forms of corporate la-
bor. At Fig Island 2 and 3, the rings were built from primary refuse of feasting episodes. But at Fig Island
1, the upper reaches of the tallest of all known shell rings seem to have been constructed of shell quarried
elsewhere and brought to the site specifically for the purpose of raising the ring higher (Cable 1997;
Russo 2002a; Saunders 2002). That is, this phase of Fig Island 1 ring construction differs from other
rings, which resulted, to variable extents, from the epiphenomena of food consumption. At smaller rings
and early stage of ring construction, the ring is raised through the incidental discard of small quotidian
refuse remains, reflecting largely egalitarian organization. In later stages, larger public construction with
feast remains suggests some form of corporate direction of on site remains. While at the largest of rings,
construction appears to proceed as labor-intensive, public works projects requiring quarrying, transport,
and hierarchical organic solidarity to manage the complex project (Tuzin 2001). Such massive building
projects at shell rings are among the earliest known in the United States (Saunders et al. 1994). While all
shell rings likely served as monuments subsequent to their making, these larger corporate efforts are dis-
tinctive in that their primary use seems to have been as a monument. Determining the methods of con-
struction of shell rings is thus essential in determining their function and the organization of the societies
that constructed them.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Diameter
Ring/Site Site No. Plaza Rise Shell References
Max. – Min.
South Carolina
Sea Pines 38BU7 60 55 42 1.0 0.7 Calmes 1967; Trinkley 1980:39
Skull Creek, Large 38BU8 55 na 24 2.1 1.9 Calmes 1967
Skull Creek, Small 38BU8 43 na 27 2.1 0.6 Calmes 1967
Guerard Point 38BU21 40 na 20 0.7 na Moore 1898:147
Chester Field 38BU29 54 27 30 1.5 1.8 Flannery 1943; Ritter 1933
Barrow's 38BU300 60 40 22 2.0 2.0 Saunders et al. 2006
Patent Point 38BU301 60 45 45 1.0 0.9 Saunders et al. 2006
Bull Island 38CH23 62 na na 0.9 na Bragg 1925; Hemmings 1970a
Coosaw 1 38BU1866 60 55 30 1.4 1.7 Heide and Russo 2003
Coosaw 2 38BU1866 60 55 30 1.4 1.7 Heide and Russo 2003
Coosaw 3 38BU1866 60 55 38 0.3 0.6 Heide and Russo 2003
Coosaw 4 38BU1866 na na na 0.4 na Heide and Russo 2003
Hanckel Mound 38CH7 62 na na 2.4 na Hemmings 1989
Lighthouse Point 38CH12 76 76 37 3.0 na Drayton 1802; Trinkley 1980, 1985
Horse Island 38CH14 61 na na 3.0 na Anonymous 1969; Hemmings 1989
Buzzards Island 38CH23 62 na na 0.9 na Judge and Smith 1991:36
Stratton Place 38CH24 50 40 21 0.6 na Trinkley 1980; 1985
Auld 38CH41 56 50 na 1.8 na Dorroh 1971; Judge and Smith 1991:36
Fig Island 1 38CH42 157 111 30 4.7 5.5 Heide 2002; Russo 2002
Fig Island 2 38CH42 82 77 58 1.4 2.1 Heide 2002; Russo 2002
Fig Island 3 38CH42 49 44 30 1.4 1.9 Heide 2002; Russo 2002
Sewee 38CH45 75 61 31 3.0 3.2 Russo and Heide 2003:31
Crow Island 38CH60 60 na na na na Trinkley 1980:246
Georgia
Oemler 9CH14A 23 na na 1.2 na DePratter 1991; Waring 1968:182
Walthour 9CH16 23 na na 1.2 na Waring 1968:182
Cane Patch 9CH35 na na na 3.0 na DePratter 1976:72, 107–109; 1974
Skidaway 9, Large 9CH63 61 na na 1.5 na Beasley 1970:118-119
Skidaway 9, Small 9CH63 30 na na 1.5 na Beasley 1970:118-119
Skidaway 21 9CH75 na na na na na Beasley 1970:119, 122
Skidaway 9CH77 77 58 35 2.3 na DePratter 1975:17; Howard et al. 1980:251
Odingsell 9CH111 47 37 na 1.5 na DePratter 1975:23-28
Ossabaw 77 9CH203 45 na na 0.9 na DePratter 1974:281
Bony Hammock 9GN53 30 na na 2.1 na DePratter 1976:130
Cannon's Point 9GN57 79 46 35 1.5 1.8 DePratter 1976:131-132; Marrinan 1975:129
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Diameter
Ring/Site Site No. Plaza Rise Shell References
Max. – Min.
Georgia (cont.)
West 9GN76 58 43 43 na 0.7 Marrinan 1975:31, 129
Long Field Crescent 9LI231 na na na na 0.7 Thomas 2006
Sapelo I 9MC65 80 75 55 2.7 2.7 McKinley 1873; Thompson 2006;
Waring and Larson 1968
Sapelo 2 9MC65 75 60 55 0.9 0.5 McKinley 1873; Thompson 2006;
Waring and Larson 1968
Sapelo 3 9MC65 55 40 40 0.9 0.9 McKinley 1873; Thompson 2006;
Waring and Larson 1968
A. Busch Krick 9MC87 40 18 18 1.6 2.4 Crusoe and DePratter 1976
Barbour Island 9MC320 65 25 na 4.0 na Georgia Site File
Florida
Rollins 8DU7510 235 190 75 3.0 3.5 Russo and Saunders 1999; Saunders 2004
Oxeye 8DU7478 160 130 65 2.2 3.0 Russo and Saunders 1999; Russo 2004:51
Horr’s Island 8CR208 160 100 125 3.0 4.5 Russo 1991, 2004:51
Guana 8SJ2554 170 150 140 1.3 1.3 Russo 2004:51; Russo et al. 2003
Joseph Reed 8MT13 250 150 200 1.5 1.7 Russo 2004:51; Russo and Heide 2002
Bonita Bay 8LL717 230 140 210 1.1 1.1 Dickel 1992; Russo 2004:51
Meig’s Pasture 8OK102 77 66 58 0.3 0.9 Curren et al. 1987
Hill Cottage 8SO2 140 120 84 4.2 3.7 Bullen and Bullen 1976
Buck Bayou 8WL90 125 41 41 1.5 na Thomas and Campbell 1991:105, 1993:530
Mississippi
Cedarland 22HC30 165 165 105 4 1.0 Gagliano and Webb 1970:49
Claiborne 22HC35 200 175 130 2 1.5 Bruseth 1991:16
Shell rings are a subset of precontact architectural features known as mounds. Temple, burial, house,
conical, flat-topped, and domiciliary mounds are common features of the Woodland and Mississippian
cultures that followed the ring builders of the Late Archaic. The mounding of shell into large-scale circu-
lar and semi-circular formations (i.e., shell rings), however, was rarely seen after the Late Archaic. While
subsequent coastal cultures definitely organized their settlements in circles and mounded the refuse of
shell into large pilings, the combination of circles and shell pilings never became a tradition after the Late
Archaic (Russo et al. 2006). Late Archaic shell rings represent unique significant structures in the annals
of United States architectural history. Their significance is compounded because of their connection to the
first sedentary settlements found on United States coastlines and the changing social formations that arose
with them.
Shell rings were the first, large-scale architectural features constructed along the United States coasts and
among the first in the entire United States. At a number of shell-ring sites, however, other architectural
features were collaterally constructed. These include causeways, ramps, walls, ridges, and ceremonial
mounds. Ceremonial mound construction was once thought to have been a strictly a Woodland (3000–
1000 B.P.) and more recent phenomenon, appearing some two thousand years after shell rings. But at a
number of shell-ring sites, large mounds and other public works were constructed wholly or partially from
shell.
Four ring sites have been linked to ceremonial mounds. These include four mounds at Horr’s Island
(Russo 1991, 2004) and one mound each at Fig Island 1 (Russo 2002a), Bonita Bay (Dickel 1992), and
the Cedarland/Claiborne shell-ring complex (Bruseth 1991; Gagliano and Webb 1970). Of these, the
Horr’s Island and Fig Island mounds have been most securely associated with shell rings. The Bonita Bay
mound requires more study. The Cedarland/Claiborne mound has been destroyed, and its association with
shell rings may never be positively confirmed.
Horr’s Island Mound A has yielded radiocarbon ages on charcoal and shell between 4760 and 4270 B.P.
Abutting the northern arm of the U-shaped shell ring, it is 30 meters in diameter and rises 6 meters above
the ring plaza and 14 meters above the adjacent Gulf Coast bay, making it the highest point along the
coast for miles. Mound B lies some 100 meters east of the ring abutting a ridge of shell midden on its
western flank. Before its destruction by developers, it stood 20 meters in diameter and 1.5 meters in
height. Two radiocarbon assays on shell from the mound yielded ages of 4645 B.P. and 4615 B.P.
Mound C lies 75 meters southeast of Mound A in what may be either an extension of the southern arm of
the ring or a separate ridge of midden. Before its destruction, it stood 1. 5 meters high and was about 15 to
20 meters in diameter, although its setting within a raised ridge of midden made it difficult to determine
its exact shape and size. Two radiocarbon assays on shell from the mound yielded ages of 4870 B.P. and
4860 B.P. Mound D lies 350 meters east of the ring, is 25 meters in diameter, and stands 4 meters high.
The single age from charcoal at its base was 4450 B.P. Allowing for standard deviations the radiocarbon
ages of all four mounds correlate with the ring’s radiocarbon ages, which range between 4660 and 4015
B.P. (Russo 1991).
The Fig Island 1 mound is 35 meters in diameter and stands 4 meters above the surrounding marsh. As at
Horr’s Island Mound A, the association of the Fig Island 1 mound is physical in that it is directly con-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
nected to Ring A by an elevated causeway of shell. No radiocarbon dates have been obtained for artifacts
from the mound.
At Bonita Bay, an elliptical mound lies 20 meters from the east edge of the shell ring. It measures 30 by
65 meters and rises 2 meters above the surrounding ground surface. Archeologists have not excavated the
mound, but have found human bone and a Woodland sherd near the surface (Dickel 1992), leaving its
cultural affiliation in question. The presence of a single sherd does not preclude an Archaic period of con-
struction. The mounds most persuasive link to the Late Archaic shell ring, however, is its proximity.
Gagliano and Webb (1970:49) identified a conical earth mound 300 meters east of the Claiborne shell
ring. It measures 30 meters in diameter and is 1.2 meters tall. No investigations into the mound were con-
ducted before it was completely bulldozed. Its association with the shell ring is based on its proximity and
a reconnaissance of the area before and after the bulldozer activity, which revealed an absence of artifacts.
The absence of artifacts, particularly pottery, suggests an Archaic rather than a Woodland or Mississip-
pian mound. But confirmation of the mound’s cultural affiliation may never be known.
All of these mounds are built wholly or partially from shell. With the possible exception of Mound C at
Horr’s Island, they are more than simple shell refuse piles. These mounds are identified as ceremonial
based on their shape, size, and internal construction. Mound A at Horr’s Island is the best studied. A se-
ries of trenches placed 15 meters into the mound near its center reveals in profile that the original ground
surface was burned prior to mound construction. A mound of sand was then placed on top of that surface.
The sand was mined from the adjacent bay rather than the dune upon which the mound sits. A layer of
shell more than two meters deep was then placed on the sand mound, followed by alternating layers of
multicolored sands, each layer up to a meter thick, and finally a capping layer of shell over 2 meters thick.
The ritual burning of ground preparatory to mound construction, the careful layering of mound deposits,
the use of distantly quarried materials, and evidence of the basket loading of multicolored sands identify
Mound A as a ceremonial structure.
The presence of an initial mound of sand buried deep beneath capping layers of shell has also been identi-
fied with probes at the Fig Island 1 mound, and in profile at Mounds B and D on Horr’s Island (Russo
1991:454, 482; 2002a), suggesting similarities in construction rituals. Not much is known of the Clai-
borne Mound. It seems to have been constructed mostly of earth, with only a little shell (Gagliano and
Webb 1970:49). At Bonita Bay, the quarry pit adjacent to the mound indicates that a core mound of
sand/earth may lie beneath the shell that caps the mound. The use of both sand and shell for these shell-
ring mound features demonstrates that they were purposeful constructions, and not simply haphazard dis-
cards of food debris.
Debate over the significance of mound construction has been fueled by the recent discovery that Middle
and Late Archaic peoples constructed large-scale ceremonial mounds along the interior rivers and coasts
of the southeast United States (Russo 1994a; Saunders et al.1994). Some researchers believe that migra-
tory hunter-gatherers were capable of constructing mounds without having to alter their migratory and
egalitarian ways of life (J. Saunders 2004). Others see changes in social organization as necessary for the
construction of public works, particularly the large and multiple works common at a number of Archaic
mound complexes and shell rings (Russo 2004; Sasssaman and Heckenberger 2004). Organizing the nec-
essary labor for large-scale public works requires the permanent or situational establishment of social hi-
erarchies to compel and guide construction. As such, the first appearance of mounds on the southeastern
United States landscape during the Archaic is seen as a hallmark in the development of precontact socie-
ties, signifying a switch from egalitarian to more complex forms of social organization.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Coastal Archaic mounds are among the first large structures made from both earth and shell (Russo
1994). Combined with the shell-construction techniques applied to rings, the building techniques em-
ployed by ring builders were famously followed by the shell-works–building cultures of east and south-
west Florida until historic contact. Architecturally, the construction of mounds and rings during the Late
Archaic marks the beginning of a long tradition of shell-works construction.
Because shell rings and mounds are so large, ring-builders found it necessary to construct appurtenances
to gain access to high points, facilitate ingress and egress to the mounds and rings, and drain enclosed
spaces. The Horr’s Island ring, which rises steeply to 9 meters above the surrounding mangrove swamp,
has a shell ramp constructed on its western flank to facilitate access from the swamp to the ring (Russo
1994). The elliptical shape of the mound at Bonita Bay may be an artifact of a shell ramp designed to fa-
cilitate the placement of shell on the uppermost reaches of the mound during construction. Similarly, a
number of the lower-level ring walls attached to the primary Ring A at Fig Island 1 likely served double
duty as a ring wall and a ramp, providing inclined access to the top of Ring A, whose own ring walls were
otherwise too steep to climb. Similar ramps are found at Coosaw 2 and Rollins. At the nearby Fig Islands
2 and 3, distinctive ramps can be seen directly opposite each other. A path of shell constructed between
the rings connects these ramps in a straight line (Heide 2002; Russo 2002a). Between the mound and Ring
A at Fig Island 1, a substantial causeway 20 meters long and 4 meters high was constructed of shell.
With the construction of such large features as rings and mounds, problems arose with drainage. At some
rings sites, the interiors may have been drained by ditches cut through the ring wall (e.g., Fig Island 2,
Reed). At other sites, rings were left open, perhaps, in part, to accommodate drainage (e.g., Fig Island 3;
Fig Island 1, Ring C; Rollins). It is often difficult for archeologists to determine if apparent drainage
sloughs date to the Late Archaic or historic times (Fryman et al. 1980; Russo and Heide 2002), and care
must be taken in assigning ownership of construction.
Other shell features at rings include isolated ridges or walls separate from the ring (e.g., Horr’s Island).
Common to many shell-ring sites, though not the rings themselves, are postmolds, usually of a small size,
10 to 25 centimeters in diameter. These have been found at Horr’s Island (Russo 1991), Lighthouse Point
(Trinkley 1985:113), Stratton Place (Trinkley 1980a:256), Guerard Point (Gantt and Styer 2006:53), and
Skull Creek (Calmes 1967:9). Almost universally, these postmolds have been found beneath shell rings in
the underlying sands/soils and not in the rings themselves. The absence of posts in the shell rings suggests
at a number of alternative explanations. Either ring-builders moved or quit building the structures alto-
gether at ring sites after the shell was piled on to make the ring. Or, as most archeologists believe, post-
molds are simply rarely preserved in the shell matrices.
Interpreting postmolds at ring sites suffers from two opposing conditions—too few, or too many post-
molds. At most sites, excavation units are too small and postmolds are insufficient to identify a structural
footprint. At Horr’s Island, the opposite was true. Here, hundreds of overlapping posts were identified in
plan view. These probably represented numerous domiciliary structures being rebuilt in the same area
over and over through time. The palimpsests of rebuilding, unfortunately, obscured the footprint of any
single structure. By mapping only the deepest of the posts, a circular structure 3 meters in diameter was
tentatively identified, with overlapping portions of other domiciliary huts present (Figure 10) Russo
(1991). At other rings, interpretations of posts include smaller structures, such as lean-tos (Trinkley
1985:113). In any case, the presence of wooden structures at shell rings represents another shell-ring ar-
chitectural feature with the potential to illuminate the settlement patterns of the earliest coastal cultures.
Combined with large-scale corporate architecture found in mounds and the rings themselves, shell-ring
sites have the potential to reveal cultural values in the changing social landscape attendant with the initial
permanent occupations of the coast.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Figure 10. Horr’s island post-molds, pit and hearth features below
shell ring (above); all but the stratigraphically-lowest features re-
moved to reveal circular structures associated with hearths and pits
(below).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The geometrically intriguing rings and the large mounds occasionally associated with them are, of course,
the most phenomenal of the cultural features at shell-ring sites. But they are not the only materials of sig-
nificance. Many shell rings are associated with the regional introduction of ceramic technologies, and
these ring sites often hold the earliest evidence of pottery in their regions. Other shell rings predate the
adoption of pottery and have yielded evidence of baked clay objects or other indirect methods of cooking,
such as cooking stones. The early ceramic production at ring sites make them ideal archeological labora-
tories for investigating the development of new technologies and the causes for their adoption. The his-
toric context, America’s First Potteries, identifies the links between shell rings and the rise of ceramic
technologies.
For archeologists, changing technologies may present material correlates of social transitions. In the
southeastern United States, changes in cooking technology may be correlated with (although not necessar-
ily the cause of) changes in social organization. At ring sites, the most obvious correlates in changing
technologies can be seen between the initial appearance of pottery and the initial appearances of the shell
rings themselves.
Construction of shell rings represent significant changes in social organization, manifested as either the
first occurrences of large-scale ceremonial architecture or the first occurrences of sedentary villages in
their regions. These co-occurrences of initial pottery and initial shell-ring construction can be observed at
the Thoms Creek pottery (Figure 11) producing sites in South Carolina, where Auld, Sewee, and Fig Is-
land shell rings have yielded some of the earliest ages for the type, dating between 4200 and 4100 B.P.
(Russo and Heide 2003:16). For early coastal Stallings/St. Simons wares (Figure 12) in South Carolina
and Georgia, similar age ranges have been obtained from shell rings such as Cannon’s Point and Sapelo 1
(Sassaman 1993:25, 240–241). On the south Florida Atlantic coast, the earliest dates for both spicule- (St.
Johns) and sand-tempered (Glades) wares have been identified at the Reed shell ring. Fiber-tempered Or-
ange wares (Figure 13) were identified directly above preceramic midden strata at the Hill Cottage shell
ring, indicating that ring occupation occurred during the technological transition to pottery, if not during
the initial ring construction (Bullen and Bullen 1976:13; Sarney 1994). In the St. Marys region of north-
east Florida, the Oxeye and Rollins shell rings, less than four miles apart, uniquely span the period of
transition from preceramic (Oxeye) to ceramic (Rollins) production, providing a set of rings ideally suited
for comparative studies on the transition from one ceramic technology (baked clay objects) to another
(pottery).
Fiber-tempered ceramic wares have long been thought to represent the earliest pottery in North America.
Even before the advent of radiocarbon dating, the relatively soft paste, thick walls, irregular surfaces, and
incised designs of these wares led archeologists to conclude that the pottery was somehow connected to
the earliest attempts at a new technology. It was seen as crude, primitive, and so inadequate as to present
doubts about its functional use as a cooking technology. Its earliest production in the form of pottery oc-
curred relatively distant from the coast, along the mid-reaches of the Savannah River at Rabbit Mount.
But, due to an already well-established cooking technology there—indirect cooking (e.g., roasting) using
soapstone (or steatite) slabs—pottery did not flourish in the interior. Rather, it took root and spread along
the coast from the mouth of the Savannah, to the north and south along the St. Johns River, and to the
eastern and western (Orange/Norwood) seaboards of central peninsular Florida. The inclusion of fibers as
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
a tempering agent was a common feature of the earliest pottery in these regions; the style, surface design,
vessel form, and period of production, however, differed among regional cultures in Florida (Orange),
Figure 11. Thoms Creek pottery: A. Incised (Sanborn and Abbott 1999:Figure 6); B–D. Cord Marked
(Sanborn and Abbott 1999:Figure 6);E. Punctate from Coosaw Island Shell Ring Complex (Heide and
Russo 2003:Figure 14); F. Fingernail Punctate from Johannes Kolb site in SC (Judge and Steen
2004:Figure 26); G. Incised from Johannes Kolb site in SC (Judge and Steen 2004:Figure 26).
Figure 12. Stallings/St. Simons pottery: A–E. Stallings (Punctate and Drag and Jab) from Coosaw Island
Shell Ring Complex (Heide and Russo 2003:Figure 13); F. Stallings from SC (Chicora 2003); G–I. St.
Simons (Waring 1968a:Figure 58); J–N. St. Simons (Waring 1968:Figure 56); O–S. St. Simons (Waring
1968:Figure 57).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Figure 13. Orange pottery: A. Orange incised from Guana shell ring; B–C. Decorated fiber-tempered
from Tick Island (Moore 1893:608).
Georgia (St. Simons or Stallings), and South Carolina (Stallings). Some of the earliest pottery makers, in
fact, forewent the use of fiber, using sand as a temper instead. Sand-tempered Thoms Creek pottery domi-
nates the pottery at most of the South Carolina shell rings.
The period of this early pottery (4500–3000 B.P.) coincided with intensified shellfish collections along
rivers such as the Savannah and St. Johns, and along the Florida, Georgia, and South Carolina coasts in
particular. As such, archeologists have speculated that the development of pottery was intricately linked
to imperatives linked to shellfish exploitation at coastal and riverine environments. Goodyear (1988) sug-
gested that pottery increased the efficiency of processing small coastal food sources, such as small fish,
snails, and clams. Stoltman (1974:233) suggested that pottery increased the efficiency of storing and
cooking shellfish. In these views, the development of early pottery is directly connected to the construc-
tion of shell rings, which required large amounts of shellfish in short periods of time to sustain feasting
ceremonies held at ring sites. Sassaman (1993:216–217) has posited that the labor-saving aspects of using
pottery in processing shellfish were critical to the appearance of shell rings. The collection and processing
of shellfish, seen by Sassaman as primarily a woman’s domain, required vast amounts of time. To counter
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
this intensification of labor, women living along the coast developed pottery to facilitate the collection,
storage, and processing of shellfish. The innovation of pottery was so successful that women became the
more important contributor to the family’s subsistence diet, heretofore more reliant on game hunting. Not
only was food-collection time decreased with the adoption of pottery, the amount of food that could be
collected at any one time increased exponentially. This increase and the ready access to the virtually end-
less supply of rapidly regenerative oyster beds allowed for the hosting of ceremonies, with attendant
large-scale feasts. Shell rings were constructed for, and were the result of these feast ceremonies (Russo
2004), and pottery was critical to supplying the feast items in the form of on-demand shellfish resources
(Sassaman 1993:227).
On the other hand, Oxeye, Horr’s Island, Bonita Bay, the early stages of Hill Cottage, Buck Bayou, Ce-
darland, and Claiborne represent large shell rings or ring components that were constructed in the absence
of pottery. There also exist a number of shell rings that, while not totally devoid of pottery, have yielded
very little (e.g., Oemler, Reed). The construction of shell rings at these sites demonstrate that while pot-
tery may have facilitated the collection and processing of shellfish, it was not requisite for shell-ring con-
struction or feasting ceremonies dependent on shellfish. At Reed, Russo and Heide (2004) identified the
pottery as the earliest ever produced in the region. At the site, it was only sparsely found in excavations.
As such, the pottery was not likely used for the large-scale processing of oysters and other foods. Rather,
the novelty and rarity of the early pottery made it a prestige item at feasting ceremonies, where it was
likely used for display, either as an ostentatious gift or in the presentation of food.
Such special purposes for shell rings as reflected in their artifact assemblages have been discussed by a
number of archeologists (e.g., Michie 1979:96; Sassaman 1993:62), but few studies have been under-
taken. At Rollins, Saunders (2004) found that pottery had been adopted in the region for hundreds of
years prior to the construction of the ring. In the region, pottery was no longer a novelty and was quite
commonly used as utilitarian wares at all habitation sites surrounding the shell ring. At the ring itself,
however, the pottery found was far more frequently elaborately decorated, suggesting that the pottery
used in shell-ring contexts had special purposes. Saunders proposed these included displaying and serving
the food at ceremonial feasts.
As these examples show, the relationship between pottery and shell rings promises insights into the impe-
tuses for the adoption of pottery, its variable functions at ring sites, and changes in forms and styles over
the course of ring occupations. Pottery also provides a window into the technological transformations that
occurred across the southeast United States when old cooking methods were abandoned in favor of pot-
tery cooking. At a number of shell rings, baked clay objects have been recovered. These objects are often
viewed as indications of the first stage of ceramic technologies, immediately preceding and leading into
ceramic pot technologies. Manufactured from clay mixed with other soils and tempers, the baked clay
objects could withstand the heat of ovens and heating or be placed in water-holding containers, thus al-
lowing food to be cooked by baking, boiling, simmering, or other indirect methods.
At Oemler, Bony Hammock, Sapelo 1, Stratton Place, Sewee, Oxeye, Meig’s Pasture, Buck Bayou, and
most abundantly at Claiborne, baked clay objects have been found in ring contexts. At a number of rings,
ceramic pottery is not present at all (e.g., Oxeye, Meig’s Pasture, Buck Bayou, Claiborne), suggesting that
the baked clay objects were a preferred cooking technology during occupation. Radiocarbon ages suggest
that these sites actually predated the local adoption of pottery, although pottery was present in the South-
east at the time most these rings, except Oxeye, were built. Some have suggested that resistance to pottery
may have arisen at sites where the inhabitants had established strong relations in the steatite trade network
and, thus, viewed pottery as a threat both to the utility of steatite in cooking and to the prestige gained
from trading in it (Campbell et al. 2004:148; Sassaman 1993).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
A number of ring sites have baked clay objects, cooking stones or steatite vessel fragments (Figure 14)
found primarily in the lower levels of the ring, with pottery in the upper levels (Sapelo 1: Waring and
Larson 1968:274; Sewee: Edwards 1965:25), (Oemler: DePratter 1975:28; Bony Hammock: DePratter
1975:66; Hill Cottage: Sarney 1994). These contexts suggest that shell rings were being occupied pre-
cisely at the time when changes from indirect cooking technologies (i.e., baked clay cooking balls, cook-
ing stones, and steatite vessels) to direct cooking (i.e., pottery) were taking place. At these sites, measures
of changes in faunal exploitation between the transitions may lend insight into the persistent assumption
that the adoption of pottery increased exploitation efficiency. Changing technologies and cultural expres-
sion in vessel form, temper, and surface designs provide the potential for shell-ring pottery studies to ad-
dress contexts of national significance relative to peopling of the southeast United States (Table 3).
Evidence of Middle Archaic estuarine exploitation has been identified intermittently along the Southeast
coast. The smallness of some sites suggest at least seasonal use of the coast by interior, migratory terres-
trial hunter-gatherers. At coastal ring sites dating to the early Late Archaic, occupation occurred as the
first permanent settlements on the coast (Russo 1996). In either case, ancestral linkage to interior hunter-
gatherer groups is apparent at coastal sites. However, the tool kits of these interior hunter-gatherers,
which were replete with worked chipped and groundstone tools, could not be relied on along the coast
where the basic stone resources were lacking—lithic tools are comparatively rare at shell-ring sites (Table
4). The exploitation of new resources such as shellfish and fish required entirely different technologies
such as nets, net weights, fish hooks, and woodworking tools for constructing canoes. If not invented by
the Late Archaic shell ring builders, these technologies were expanded upon and fully developed by these
groups.
Shell replaced or supplemented lithics as a basic endurable resource for the manufacture of tools. The
coastal shellfish collectors fashioned shell into many and intricate objects that were used to capture, proc-
ess, and aid in the consumption of fishery staples. At Late Archaic shell rings, the range of tool types and
ornaments fashioned from shell is extensive. It includes pounders, hammers, gouges, chisels, adzes, celts,
planes, hoes, dippers, cups, spoons, scoops, pendants, beads, anvils, choppers, preforms, manos, net mesh
gauges, net weights, and various columella awls, fids, pins, and picks (Figure 15). All or most of these
types of shell tools are found at the earliest of the shell rings, indicating that the tools were essential to the
construction and use of the rings. That is, early on, shell tools were utilized in exploiting the resources
(shellfish) that formed the building materials for the rings themselves. The tool kit was so highly devel-
oped, that virtually no innovations in shell-tool technology would follow at coastal sites for the remaining
millennia of precontact history. Those shell tool types that were being used at the time of European con-
tact, can all be found in Late Archaic shell rings.
Shell technology was of undoubted importance to ceremonial life and the trade and subsistence econo-
mies of shell-ring builders. Beads made from gastropod columellas, as well as their outer whorls and
whole shells, whole snails, and bivalves are found at ring sites, but were also items of trade. Shells in-
digenous to the shell-ring coasts have been recovered at interior southeastern United States sites, and in
burial and other ceremonial contexts at sites as far away as the Northeast and Midwest.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Figure 14. Steatite objects and vessels from the Southeast and greater U.S. steatite trade network. A. Ob-
jects from Sarah’s Ridge ( Kane and Keeton 1993:Figure 30); B. Objects from Paris Island South (Kane
and Keeton 1993:Figure 30); C. Objects from Rocky River (Kane and Keeton 1993:Figure 30); D. Bowl
(Ward and Davis 1999:Figure 3.11); E. Vessel Form Variety (NPS 2004).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Table 3. Pottery
Figure 15. Shell Tools common at Late Archaic sites from the southwest Florida Coast. A. type B1 cut-
ting-edged tool; B. type A cutting-edged tool; C–D. unhafted hammer; E–G. perforated bivalves; H.
columella plan; I–L. columella sinkers; M. shell net mesh gauges—two views; N. shouldered adze—two
views; O. adze/celt—two views; P.cup (Marquardt 1992).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Shell tools were necessary to collect and process other resources. The shells of bivalves were perforated
and used as net weights. They were notched and sharpened for use as hoes, adzes, and celts. The long
columellas of whelks and conchs were ground down and fashioned into compound fish hooks, pounders
to cull oysters, and chisels to carve wood (Walker 1992:298–299). They were scribed for use as net
weights and attached to handles for use as hammers. The outer whorls of whelks and conchs were squared
off and used as mesh gauges in the fabrication of fish nets. They were sharpened and shaped into adzes
and celts, or chipped into spoons, dippers, and scoops for use in cooking and serving foods.
Despite their importance in varied aspects of shell-ring cultural life, not all shell rings have had shell tools
identified with them (Table 5). This may be due to archeologists’ failure to recognize shell tools, or the
difficulty in distinguishing food refuse from modified shell (e.g., Marrinan 1975:65). Only recently has
the wide range of shell tools been classified into formal types (Marquardt 1992; Torrence 1996). This ty-
pology identifies not only a large number of tool types, but specific uses and variations in use for a variety
of the types. Not only are chisels and hammers identified, but various types of columella cutting-edge
tools and whelk and conch hammers are distinguishable.
Equally likely, contributing to the absence of reported shell tools is the fact that relatively few shell rings
have been excavated, particularly in South Carolina and Georgia. In South Carolina, archeologists have
reported, on average, ten shell tools for every shell ring identified; in Georgia, the average is less than one
per ring. But only seven (Stratton Place, Lighthouse Point, Chester Field, Fig Island 1–3, and Sewee) out
of twenty-two known rings in South Carolina, and three (Sapelo 1 and 3 and Cannon’s Point) out of six-
teen in Georgia have been excavated to any significant degree. At these sites the averages are 30 and 7,
respectively. In contrast, Florida shell rings average far more shell tools—285 per ring (Table 5). If only
those rings that have been extensively excavated are counted (i.e., Horr’s Island, Hill Cottage, Rollins,
and Guana), the average increases to 640 per site. These data indicate that certain shell-ring–building cul-
tures used shell as tools more intensively and in greater diversity than others. Specifically, the shell-ring-
building cultures on the southwest coast of Florida were preeminent in their use of shell tools. A number
of reasons were behind this intensified use. The southwest Florida shell rings were the most distant from
lithic resources, located at the periphery of the steatite trade network (Campbell et al. 2004; Sassaman
1993). Here ring builders either did not use pottery at all, or did not use it to any significant degree.
The diversity of shell tools used by southwest Florida ring builders was great. Virtually all shell tools
found at other ring sites were whelk hammers (Table 5)(e.g., Edwards 1965; Marrinan 1975; Trinkley
1980a; cf. Saunders [2002:118–119], who additionally identified whelk columella punches, gouges, and a
net spacer at Fig Island). In contrast, the tools at Horr’s Island included pounders, hammers, gouges, chis-
els, adzes, celts, planes, hoes, dippers, cups, spoons, scoops, pendants, beads, anvils, choppers, preforms,
manos, net mesh gauges, net weights, and various columella awls, fids, pins, punches, and picks—a
nearly complete shell tool assemblage. The shell tools from Hill Cottage, although not as diverse, are nu-
merous. In part, this is due to the identification of shell tool debitage as well as the tools themselves.
Much in the same vein as chipped lithic tools, the manufacture of shell tools can result in identifiable
debitage. Analysis of this debitage can be used to identify the processes of reduction the shell went
through to make the tool (Marquardt 1992; Torrence 1996). Identifying this debitage is an important con-
sideration, which, unfortunately, most archeologists have not made. It can lead directly to recognizing the
activities that went on at shell rings and interpreting the functions they may have served, e.g., habitation
or ceremony.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Shell Shell
Ring/Site Other Ring/Site Other
Tools Tools
Shell-ring sites are renowned for bone tools and ornaments, specifically decorated bone pins. This is per-
haps another elaboration of material culture kindled in response to a paucity of lithic and other resources
common elsewhere in the southeast United States. Trinkley (1980a:298) suggests that “next to pottery,
the most visible artifact category is worked bone.” But he was speaking not only of shell rings but of Late
Archaic coastal sites in general, as well as interior sites. Unlike the shell tools found at shell-ring sites, the
bone tools have clear precedents in the Middle Archaic and contemporary usage pan-regionally (Figure
16). As such, and in the general absence of lithic points, it is the often abundant and near ubiquitous pres-
ence of bone tools at shell rings that provide a material connection to the involvement of the coast in trade
and other relations with the greater Southeast. However, a close look at the actual numbers of bone tools
found at shell rings suggests, that like shell tools, some of the ring-building cultures lay outside the main-
stream traditional bone tool technology. Florida, in general seems to yield fewer bone pins than other ring
builders (Table 6). This suggests they may have been marginalized from the greater Southeast regional
trade and communication networks.
Figure 16. Bone pins. A. Incised bone pin from Ossabaw Island (Waring 1968a:Figure 64); B-D. Incised
bone pins from the Chester Field site (Waring 1968a:Figure 64; E-I. Engraved bone pins from Tick Is-
land (Jahn and Bullen 1978).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The commonalities of material culture indicate extensive intercommunication between ring-building cul-
tures, while the lack of evidence for intensive use of certain of the technologies suggest limited participa-
tion in the wide-ranging trade and communication networks among some of the ring builders. In fact, it is
these differences in the relative use of specific cultural technologies that suggest an increasing regional-
ism was taking hold during the late Archaic. For example, greater numbers of exotic trade items, such as
steatite and chipped lithics and groundstone artifacts, as well as bone tools, suggests that the Atlantic
coast ring builders held closer ties to extra-regional trade and information networks involving interior
southeast U.S. groups than did Gulf coast ring-builders. In the Gulf coast area, the absence of pottery at
certain regional shell rings has been viewed as evidence that those trade and information connections
were resisted by certain groups or otherwise lay outside the greater southeast U.S. networks (Campbell et
al. 2004; Dickel 1992; Russo 1991).
At the same time, the shell and bone tool technologies abundant at Gulf coast ring sites, but less inten-
sively used along the Atlantic coast, suggest that the Gulf coast groups participated in closer trade and
communication networks among themselves than with their Atlantic contemporaries. Bullen and Bullen
(1976:12) have suggested that the intensive development of shell tool technologies at the Gulf coast Hill
Cottage shell ring indicate the occupants may have been borrowed their ideas from Late Archaic cultures
on the east coast of Florida. But, based on additional radiocarbon testing, the southwest Florida shell tool
traditions actually hold chronological priority over the Atlantic coast ring builders (Russo 1991; Torrence
1996). Diffusion of these technologies may have instead flowed from the Gulf coast to the Atlantic coast,
where they were widely, but not intensively adopted, perhaps due to competition with in situ lithic and
ceramic technologies that prospered through traditional exchange networks.
Some have suggested that shell rings’ use as ceremonial centers was directly related to the furtherance of
exchange networking and information gathering traditions. Cable (1997) suggested that shell rings were
used annually to host ceremonies at which the exchange of mates and material occurred. Russo (2004) has
suggested that the feasting that is so evident at shell rings is commonly associated with rituals involving
material and social exchange. Yet, exactly what items, information, or people were being exchanged at
shell rings remains something of a mystery. Based on their more limited artifact inventory, Michie
(1979:96) has suggested that South Carolina shell rings were more restrictively used than were non-ring
base camps whose artifact inventory included items such as lithic bifaces, debitage, hammerstones, stea-
tite not typically found at shell rings. As such, shell rings have been interpreted as ceremonial centers
(Sassaman 1993:62) with the apparent exclusion of certain utilitarian objects at rings implying aura
of sacred ceremonialism. This argument, however, has been largely undercut by Trinkley (1980a:311–
12), who have shown that shell rings have similar artifact assemblages as any of the amorphous base
camps cited by Michie. If exchange and information networks with interior groups involved
coastal shell ring cultures, the communal plazas of shell rings likely served as points of ex-
change.
Interpretations of shell rings have consistently avoided analogies with the contemporary “Shell
Mound Archaic” cultures of the mid-Southern interior river valleys (cf. DePratter and Howard
1980:8). Archeologists have formulated settlement models for these valleys in which small wandering
hunter-gatherers, widely dispersed away from the rivers much of the year, come together or aggregated
along the river at other times (e.g., summer and spring) to take advantage of seasonally abundant fish and
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
shellfish resources (e.g., Worth 1994). Undoubtedly, shell rings functioned as sorts of aggregation sites
during periods of ceremonies and feasts. But the coastal populace were settled permanently at the
rings and nearby along the coast throughout the year. Hence, the use of shell rings as subsistence ag-
gregation sites is not likely despite the great abundances of shell remains that signify otherwise
in interior sites. Shellfish were present along coastal estuaries throughout the year and rings were
used throughout the annual cycle. As such, the evidence suggests that rings were most likely
places social interchanges and material exchanges during ritual, ceremony, feasting and other social oc-
casions.
The investigation of the material culture shell rings offers great potential for resolving questions of trade,
the diffusion of ideas, the identification of cultural affiliations, and social connections among ring build-
ing and non-ring building traditions. Establishing the direction of flow and strength of connections be-
tween the various ring-building cultures is critical to addressing thematic questions related to the estab-
lishment of tribal identities, the peopling of the coastal zone, and the trade networks that linked the many
shell-ring-building cultures to the greater Southeast.
The shell at ring sites not only provide insight into the molluscan fauna that was consumed, but has
served to provide a protective environment for other fragile subsistence remains, such as bone and, to a
lesser extent, botanical remains, which has allowed a broad assessment of the food and other resources
used at shell rings (e.g., Marrinan 1975; Russo 1991; Trinkley 1980a). To date, more faunal analyses have
been undertaken on shell-ring assemblages than on those from any other type of coastal Archaic site.
These analyses have provided insight into the seasons of settlement at ring sites and, by extension, the
permanency of Archaic coastal settlement. Are shell rings permanently settled villages? Seasonal camps?
Periodically visited congregation or ceremonial sites? These are all critical questions that have been asked
and, to some extent, answered by faunal studies. The answers speak to the importance of studying the cul-
tural change that occurs when nomadic, egalitarian hunter-gatherers settle year-round in one place. Do
attendant changes in social organization occur that lead to complex and hierarchical social formations?
Fauna move and grow in cycle with the seasons of the year. The record of those seasonal cycles may be
found in the skeletal material left behind at shell-rings deposits. Growth marks in shellfish and fish skele-
tons, the absence/presence of species specific to the area only at certain times of the year, and the sizes of
individuals all provide signatures as to the collection season, and, by extension, when and for how long
shell rings were occupied throughout the annual cycle. From analyses of faunal and botanical remains,
which also reveal seasonal markers, shell rings can be identified as temporary or seasonal sites or perma-
nently occupied. Using faunal and other evidence, archeologists have variously identified shell rings as
permanently settled villages (Russo 1991; Trinkley 1980a), seasonally occupied aggregation sites (Michie
1979; Saunders 2004:261 FA), or year-round habitation/ceremonial centers (Russo 2004).
Identifying the seasons and resources used at shell-ring sites requires analysis of large numbers of subsis-
tence remains from a wide variety of shell-ring contexts (e.g., plazas, rings, pits, and hearths), which to-
gether potentially reflect the total spectrum of seasonal activities that might occur at a site. At Cannon’s
Point, Marrinan (1975) was the first archeologist to analyze subsistence remains from a shell ring using
fine-mesh recovery techniques (cf. Edwards 1965). Recovering faunal remains with fine mesh allows for
the recognition of vertebrate species that would otherwise have been lost in standard archeological exca-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
vations. The recovery of the broad spectrum of species utilization (mostly fish and shellfish) at Cannon’s
Point led to the recognition of year-round collection strategies there. Using similar analyses, Trinkley
(1980a:175) found evidence of year-round settlement at Lighthouse Point. Thompson (2006) identified a
similar pattern at Sapelo 3. Over one million fragments of shell and bone were analyzed from Horr’s Is-
land, leading Russo (1991) to conclude that year-round occupation took place there. Russo (1993) also
found multi-seasonal remains at Rollins, leading to the same conclusion.
In the face of limited data, however, others have concluded that shell rings were built from the debris col-
lected during temporary stays or visits to the rings (Michie 1979). At some ring sites, such as Fig Island
and Reed, the faunal analyses were too limited to conclude year-round site occupation, even though a
broad range of species were identified (Russo 2002b; Russo and Heide 2000). Rather than interpreting Fig
Island and Reed as seasonal occupations, however, the authors note that there is plenty of archeological
information remaining at Fig Island that may be studied and more data recovery would provide insight
into such questions.
Taken together, the large sizes of shell rings and the seasonality studies that have shown the rings to be
occupied throughout the year are the primary evidence that ring sites represent the first permanent settle-
ments on the southeast coast. As such, seasonality and subsistence studies serve to provide tests for com-
peting hypotheses that rings represent either seasonal or permanent settlements (DePratter 1979:35). Sub-
sistence studies, however, also provide evidence of the kinds of changes in subsistence strategies that
were necessary when terrestrially oriented hunter-gatherers became reliant on estuarine resources. The
collection of shellfish and fish required significant changes in technologies as outlined in Contexts 3 and
4. But they also required changes in social behavior. The construction of canoes, the use of nets, the own-
ership of shellfish beds and fishing grounds all arose when settlement on the coast became permanent.
These significant changes in social organization are, in part, reflected in the shell rings themselves as
identified in the kinds of food remains present and the distribution and mounding of shellfish to demar-
cate public and ceremonial space and mark ethnic or kin-based territories as outlined in Context 1.
Of some fifty shell rings, perhaps four have yielded human remains in secure archeological contexts.
Most of these remains are isolated fragments of bone, not full burials. The reasons for this scarcity may be
many, including the fact that relatively little archeology has been undertaken at shell rings. In the general
absence of burial remains, it may seem difficult to argue that shell rings may hold evidence of mortuary
practices that are of national significance. But clearly whatever ring builders were doing with the remains
of their dead, it differed dramatically from traditions that preceded (e.g., cemetery and pond burials) and
followed (e.g., mound burials) those of the ring builders. At shell rings, mortuary as well as isolated hu-
man remains are of national significance because they speak to questions of stability and change associ-
ated with the cultural practices of societies undergoing increasing numbers, establishing territories, and
symbolizing ethnic identities through material culture and architecture. The disposal of human remains
and associated burial furniture is a significant research agenda that can address issues of ethnic identity,
social organization, ceremony and ritual, demography and health, and settlement patterns.
Few burials have been identified at shell rings and other contemporary coastal sites. The sites containing
human remains have usually yielded fragmentary bones in ambiguous contexts, a situation that has not
provided much insight into patterns of burial (e.g., Bellamo 1995; Michie n.d.). In South Carolina, Geor-
gia, and Florida, so few coastal Late Archaic burials have been identified that any discussion of mortuary
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
practices is wholly and largely excluded from settlement models (e.g., Elliott and Sassaman 1995; Mi-
lanich 1994; Sassaman 1993; Sassaman and Anderson 1994:154).
Speaking of shell rings and associated shell middens, Trinkley (1980a:324) wrote that “skeletal materials
are perhaps the most neglected archeological data in South Carolina today.” He noted that only one non-
ring, midden site, Daws Island (38BU9), had revealed “evidence of intentional burials, and at even this
site none of the burials have been removed under satisfactory archeological conditions.” So even at Daws
Island, the burial patterns are not well understood.
Despite the lack of discovery and investigation of human burials from Late Archaic coastal sites, shell
rings are seen by some as ceremonial burial sites (e.g., Elliott and Sassaman 1995:146). This misunder-
standing may have evolved from Trinkley’s comments on burials. In his review of human remains from
South Carolina shell rings and contemporary midden sites, Trinkley clearly stated that “during shell-ring
excavation occasional [emphasis added] heavily fragmented human bones have been found.” He then
lists only five shell-ring sites, but also one non-shell-ring site as his examples, unintentionally inflating
the rate of occurrence of human remains at ring sites. Trinkley then discusses two other non-rings sites
with burials, Daws Island and Stallings Island. Subsequently, archeologists citing Trinkley have mistak-
enly suggested that “nearly all shell rings and other shell middens on the coast have yielded human skele-
tal remains” (Elliott and Sassaman 1995:146; Sassaman 1993:63; Sassaman and Anderson 1994:154).
Sarney (1994:147) inflated the occurrence rate of human remains at ring sites by misinterpreting the non-
ring Daws Island site as a shell ring and adding it to her list of shell rings containing human remains.
Speaking of shell rings, Sarney (1994:169) concluded that “at least 50 percent of the sites contain frag-
mentary human remains.” As the discussion below demonstrates, human remains from shell ring sites are
not nearly so ubiquitous.
Few human remains have actually been identified from the shell rings themselves. As Table 7 shows, hu-
man remains have been attributed to at least thirty shell-ring contexts from fifteen sites. Some of these
contexts can be summarily dismissed from consideration. As stated, one site, Daws Island, is mistakenly
identified as a shell ring (Sarney 1994:147). In addition, the human remains attributed to Guana are from
a context 200 meters away from the shell ring (Newman 2002), not from the ring itself. The burials attrib-
uted to the Reed shell ring were identified from a newspaper account as eroding out of a beach, which
may or may not be connected to the shell ring; while the context of the lone skull attributed to the ring—
the only curated human remain attributed to the site—is unknown. The flexed burials at Hill Cottage are
from a mixed context in the upper portions of the ring that date to Woodland periods, not the Archaic
(Sarney 1994). The crania fragment from Bonita Bay was not found in the ring, but rather in an upper
mound context, which may or may not be temporally associated with the adjacent shell ring. And the Se-
wee mandible fragment probably came from a midden context just outside the shell ring (Russo and
Heide 2003:2, 25).
Other human remains with shell-ring attributions may be Late Archaic and likely came from rings, but
suffer from contextual problems. The two human remains from Sapelo, a temporal bone fragment and a
deciduous tooth, are from contexts not clearly identified (Moore 1897). And their assignation to the Late
Archaic is problematic because non-Archaic materials have been recovered from both Sapelo Ring 1 and
Sapelo Ring 2. The Skull Creek 1 mandible fragment with two teeth in it seems to be from a secure ring
plaza context, Level 2, with ceramics and non-ceramic artifacts associated with the Late Archaic. But
Calmes (1967:23) fails to provide a listing of all artifacts from the provenience and notes that the excava-
tion unit was “shallow and cut through plow-disturbed zones.” This does not exclude the mandible frag-
ment from being attributable to a Late Archaic shell-ring context, it just presents doubts. Also suspect are
two of the contexts of the deciduous teeth found at Rollins. Saunders (2003:24) noted that one tooth came
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Table 7. Human Remains and Burials from Shell Rings and Nearby Sites
Ring vs.
Site/Ring # Element(s) Provenience References
Non-Ring
South Carolina
"jaw" fragment with 2
Skull Creek 1 (38BU8) 1 Y–12, level plaza unclear Calmes 1967:23
teeth in place
Daws ` (38BU9) 6 burials eroding shoreline non-ring Trinkley 1980a:322
Barrows (38BU300) 1 crania fragment Bu300, surface unclear observed in field
Coosaw 1 (38BU1866) 1 molar Ring 1, EU1, L12–23 cmbd ring Heide & Russo 2003:18
shaft, proximal left femur Fig. 2, ST5, s. profile of
Fig Island 2 (38CH42) 1 " Saunders 2002:140
fragment Hemmings’ trench
Sewee (38CH45) 1 mandible fragment G2, 12–18 in., outside ring non–ring Edwards 1965:18, 47
Georgia
Cannon’s Pt. (9GN57) 2 1 femur, 1 tibia 9S/15E, L2 ring Marrinan 1975:82–83
" " " 1 calcaneus 6S/15E, L1 " " " "
" " " 1 molar 6S/15E, L8 " " " "
" " " 1 cranium 18N/0E, L31 " " " "
" " " 1 pelvis 18N/0E, east baulk " " " "
" " " 1 incisor 18N/3E, L24 " " " "
" " " 1 parietal 24S/24E, 75–85cmbs non–ring " " "
Moore 1897:73;
Sapelo 1 (9MC23) 1 temporal bone fragment plaza ring plaza
Marrinan 1975:128
Sapelo 2 (9MC23) 1 "deciduous tooth" Ring 2, or possibly Ring 1 unclear Simpkins & McMichael 1976:99
Florida
intrusive in
Horr’s Island (8CR206) 1 burial Mound B Russo 1991
mound, not ring
Horr’s Island (8CR208) 2 burials Mound A/ring " " " "
Oxeye (8DU7478) 1 molar TR1, EU3, L9; FS29 ring Catalog on file SEAC
Rollins (8DU7510) 1 deciduous incisor EU140; FS 338 unclear Saunders 2003:24
" " " 1 " " EU2011; FS 202 ring " " "
" " " 1 tooth TR1, EU4, Z2/1; FS 172 " Saunders 2003:26
" " " 1 " TR1, EU6, Z1; FS 168 " Saunders, p. comm. 2006
" " " 1 " TR1, EU8, Z1; FS 309 " Saunders, p. comm. 2006
in mound, not
Bonita Bay (8LL716) 1 crania fragment Mound 8LL716, 30 cmbs Dickel 1992:142
ring
Carr et al. 1995; Fl. Atlantic
Reed (8MT13) 1 skull; uncounted remains unknown unclear
Univ. catalogue cards
1 crania, 1 long bone,
Guana (8SJ2554) 3 200 m northeast of ring non–ring Newman 2002
1 mandible
2 flexed burials; Intrusive ring Bullen & Bullen 1976:89;
Hill Cottage (8SO2) 2 Unit A, upper 3–12 in.
scattered remains burials? Sarney 1994:40
" " " 1 incisor 2 x 2 m unit, 95–115 cmbs ring Sarney 1994:51, 80–82, 88
" " " 1 molar 2 x 2 m unit, feat. 3 " Sarney 1994:53, 80–82, 88
" " " 1 incisor 2 x 2 m unit, 135–155 cmbs " Sarney 1994:53–54, 80–82, 88
# = number of fragments or burials reported; cmbd = centimeters below datum; cmbs = centimeters below surface; feat. = feature;
TR = trench; EU = excavation unit; L = level; Z = zone; ST = shovel test; FS =field specimen
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
from an excavation unit with a Woodland feature. In addition, that unit lies well outside the ring (Saun-
ders 2003:6).The other unit lies in a gap in the ring that was contaminated by “slopewash” (Saunders
2003:6, 24), although the associated artifacts are almost totally Orange period. Three other teeth from
Rollins were from secure Late Archaic shell-ring contexts.
Of three occurrences from Horr’s Island, one was an intrusive burial into Mound B, not the ring. Techni-
cally, it does not qualify as a shell-ring burial because it was placed into the mound after the ring and
mound were constructed. But it does shed light on a pattern of burial by the Horr’s Island ring builders.
That is, they did not place burials into their associated mounds, but their descendants may have. The other
two burials came from a context where Mound A adjoins the shell ring. As such, the burials could be at-
tributed to either the mound or the ring. But like the burial from Mound B, both were intrusive, post-
dating the construction of the mound and ring. This does not necessarily exclude them from classification
as ring burials. Like the Mound B burial, they can shed light on post-ring burial patterns. Radiocarbon
dates on two of the burials indicate they occurred hundreds of years after the construction of the ring
(Russo 1991:423). These dates suggest that the Horr’s Island ring may not have been used as a cemetery
during its occupation, but may have subsequently functioned as a sacred monument or cemetery where
burials were placed. These speak to the questions brought up in Contexts 1 and 2 concerning the evolving
functions of shell rings and associated mounds throughout their periods of use.
This synopsis leaves five shell-ring sites—Hill Cottage, Oxeye, Coosaw 1, Fig Island 2, and Cannon’s
Point—with seemingly unequivocal evidence of human remains in Late Archaic contexts The human re-
mains consist of seven teeth and fragments of two femur, one tibia, two crania, a pelvis, and a calcaneus.
(The Cannon’s Point parietal came from a non-ring context). If shell rings with only teeth are eliminated
(isolated teeth being more simply attributable to natural loss, rather than burial), then Fig Island 2 and
Cannon’s Point are the only shell rings that have definitively yielded fragmentary human remains, and
Sapelo1 and Skull Creek 1 are likely contributors. No complete or near complete burials have been recov-
ered from any ring site in contexts securely dated to the periods of ring construction. Regarding Cannon’s
Point, the ring site with the greatest number of human remains, Marrinan (1975:95) stated that the “skele-
tal material…is admittedly not well represented; nor is human skeletal material from any other ring site.”
If shell rings were not formal cemetery sites, what did shell-ring societies do with their dead? One sugges-
tion is that in Stallings/St. Simons coastal sites cremation was a Late Archaic mortuary practice (Elliott
and Sassaman 1995:121–122; Thomas and Larson 1979:64). Of course, another obvious explanation for
the presence of a few burned or other fragments, or the lack of formal burials is that most of the cremation
or burial activities may have occurred off site (Trinkley 1980a:323). For example, Daws Island (38BU9),
which did have burials, is near both the Barrows and Patent shell rings, so it may represent a cemetery
related to the two shell rings. Or, archeologists may have simply not looked in the right places for Late
Archaic burials. Mound burials, known during the Middle Archaic (Aten 1999; Russo 1994) and common
during the Woodland (e.g., Thomas and Larsen 1979), may be one place to look. Late Archaic mounds
have been found at Horr’s Island and Fig Island and possibly at the Bonita Bay, Cedarland, and Claiborne
rings (Bruseth 1991:15; Gagliano and Webb 1970:49). Although Mound A at Horr’s Island has revealed
only intrusive burials, the center of the mound was never excavated. Another possible burial location, par-
ticularly in Florida, may be mortuary ponds—a common Middle Archaic environment for burial (Mi-
lanich 1994).
Plaza burials may be yet another possibility. At Woodland coastal-ring middens, burials have been lo-
cated within plazas (Bense 1994). Certainly, ethnographically, village plazas are known to be cemetery
locations (e.g., Malinowski 1929). And the mandible from Skull Creek 1 was recovered from the plaza. If
Late Archaic shell-ring plazas were burial areas, archeologists may have missed them because relatively
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
few plazas have been investigated (e.g., Hemmings1970f; Russo 1991; Trinkley 1985). Also, the general
absence of shell in plazas provides a relatively poor environment for bone preservation compared to that
of the adjacent shell ring.
But plaza burials do not easily account for the occasional presence of human remains in shell rings. Mar-
rinan (1975:82) listed a number of possibilities for the inclusion of only crushed, fragmentary remains at
the Cannon’s Point shell ring. These include the borrowing of shell from off-site cemeteries to build the
ring and the “haphazard” burial of secondarily deposited remains. However, evidence for borrowed shell
is generally absent at rings (see Russo and Heide 2003 for a discussion of the evidence for borrowing).
And supporting evidence for secondary deposit (e.g., bundle burials) is lacking: Marrinan noted that no
evidence of burial pits or burial goods were associated with any of the human remains at Cannon’s Point.
The possibility that bodies were exposed on the surface of the shell rings and later collected for reburial
has also been proposed (Marrinan 1975:82; Trinkley 1980a:323). This would readily account for the oc-
casional presence of human bone in shell rings. But it still leaves open the question of where most of the
human remains ended up.
Of course, mortuary practices of shell-ring builders may not have involved in-ground burials at all. Cre-
mation was mentioned. But offshore burial, tree or scaffold burial, or abandonment are all possible expla-
nations for the absence of articulated remains in shell rings. Often the presence of isolated human bone
and teeth in Late Archaic shell rings, and shell middens in general, has been attributed to cannibalism.
Waring (1968a:191) links the “presence of cracked and fire-blackened human bone scattered throughout
the midden” at Bilbo as supporting evidence that cannibalism is a “true trait of the Archaic period in Ala-
bama, Georgia, and Florida” (see also Wyman 1875:60–78). Michie (1979:77) suggested that the scat-
tered human remains he found at the non-ring shell midden at the Bass Pond site “evoked suspicion” of
cannibalism. But it is important to note, that none of these suggestions directly involved any evidence
from shell rings. Although Marrinan (1975:95) cited Moore (1897) and Waring (1968a) as presenting
evidence of “post mortem consumption” of humans at shell rings, they did not. Moore simply identified
the temporal bone he found at Sapelo 1 without comment on its context, condition, or possible origin.
Waring (1968c:245) made no mention of human remains from any shell ring in particular. Rather, his
comments on possible Archaic cannibalism are general, based, in part, on Wyman’s (1875) observations
of human remains from shell middens, none of which were found at Archaic shell rings.
Only Edwards (1965) has offered cannibalism as one possibility to account for human remains at shell
rings, in particular the burned femur at Sewee. Marrinan (1975:82), on the other hand, criticizes assump-
tions of cannibalism when human remains are found in shell middens. She notes specifically that there
was “no solid evidence of cannibalism” in the Cannon’s Point human remains. Cut marks were apparently
the “solid evidence” to which she refers, and these were lacking on the remains she found. Phelps and
Burgess (1964:200), however, have suggested that in addition to evidence of cut marks or burning, other
evidence of cooking (e.g., the presence of clay cooking balls) and offerings (e.g., a bear claw necklace)
may be used to support the interpretation of fragmentary human remains as resulting from cannibalism.
Ritual may be associated with cannibalism. But it is also associated with burial, and specific contexts
must be analyzed to distinguish between the alternate possible sources of ritual.
It is clear that if cannibalism is to be proposed as an explanation for the presence of fragmentary human
remains, the context of the remains must be understood. In shell rings, the remains are most often associ-
ated with fish and shellfish food remains. Thus, a connection to consumption of humans as food items is
often made, as if humans were just another food item. However, if cannibalism is the source of the occa-
sional human remains at ring sites, more likely it would have occurred in ritual contexts, which may result
in little or no actual preparation or consumption observable on the remains. Such ritual use could result in
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
a random distribution of human remains at both shell-ring and non-ring sites. Speaking of fragmentary
remains at shell rings, Sarney (1994:158) stated that “the scarcity of defined burial pits or grave offerings
indicates a general lack of mortuary care.” This, of course, presumes that human remains are always de-
posited as a mortuary act. Human bones may be kept as trophy items, for use in tool manufacture, or as
spiritual tokens (Jacobi and Hill 2002). Their deposit in shell rings may be intentional or accidental, but
does not necessarily reflect a tradition of mortuary practice.
Because ring builders were members of a number of broadly-related but independent cultures, it is possi-
ble that each shell-ring culture treated human remains differently. Among Thoms Creek and St. Simons
cultures, off-site cemetery or cremations have been suggested (Trinkley 1980a:323). In southwest Florida,
shell rings may have served as cemeteries or sacred architectural memorabilia (see Dillehay 1990:233)
into which burials were placed, but only after ring abandonment (Russo 1991, 2004; Sarney 1994). At the
two Mississippi shell rings, Bruseth (1991:13, 17) noted that despite the extensive excavations, no human
remains were ever recovered from either ring. He did suggest, however, that off-site burial traditions may
have included mound burial, indicating the presence of a nearby sand mound “associated” with the rings
in some unspecified way (Bruseth 1991:13, 15; Gagliano and Webb 1970:49). Somewhat contradictorily,
he also suggested that a cache of steatite vessels and copper and lapidary objects from a shell-free area of
the Claiborne ring might represent funerary objects, absent human remains because of poor preservation
in the sandy soils (Bruseth 1991:17; cf. Gagliano and Webb 1970:59). Thus, as with all shell-ring cul-
tures, it is unclear what the mortuary practices of the Mississippi shell-ring builders were.
Despite assumptions to the contrary, not all Late Archaic shell rings contain isolated human remains; not
all isolated human remains represent artifacts of mortuary practices; and of the few known intentional
burials, none took place at the time of ring construction. These glimpses into the possible differences in
the presence/absence of human remains in shell rings speak to thematic issues related to site function, cul-
tural traditions, and ethnic identification.
To this point, all the historic contexts have reflected on how shell-ring cultures shared with and differed
from other cultures across the Late Archaic southeast United States. The architecture built, the pottery
made, the tools used, the environment settled, the resources exploited, the disposal of the dead—together
these contexts point the way to identifying social boundaries and tribal identities.
The rise of regionalism characterizes Woodland and Mississippian cultures and serves, in part, to distin-
guish them from earlier Paleoindian and Archaic archeological cultures. With greater populations arising
during the late Archaic, territories became restricted, and kin, tribal, and political associations formed to
exploit and protect their territories. Symbols of group identity became ubiquitous and necessary to aid in
the maintenance of these socio-political associations. For archeologists, pottery is the most common ma-
terial used to identify and distinguish among Woodland and Mississippian groups. But not all Late Ar-
chaic, and certainly none of the earlier Archaic or Paleoindian groups, made the pottery that provided ar-
cheologists with these potential symbols of group identity. Paleoindians and Early Archaic groups
undoubtedly had group identities and markers that symbolized those identities. But these groups were
largely mobile, covered enormous territories, and left behind little in the material record that would en-
able archeologists to identify group distinctions.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
With the Late Archaic, territories were smaller and more stable, and identifiable markers of group iden-
tity, such as distinctive pottery types, were left behind. Attempts have been made to recognize cultural
affiliations among Late Archaic pottery producing cultures. Most broadly, cultures that made Thoms
Creek, Stallings, and Orange series pottery have been linked geographically to South Carolina, Georgia,
and Florida respectively. But in these cases equating pottery series with ethnic identities results more
from loose use of the series’ names than the belief, for example, that all peoples making or using Orange
pottery in Florida were of the same tribal or other group. The expansive distribution of Orange pottery
over such a large geographic area as the Florida peninsula suggests trade, diffusion, or wide-scale use of
the same pottery by various ethnic or tribal entities rather than a single tribe. In this case, more than just
pottery is needed to identify tribal boundaries.
Combining pottery series and designs with other materials found at ring sites provides an initial entrée for
determining the earliest ethnic identities of groups that settled the coast in the Late Archaic. Common to
all groups is a dependence on coastal shellfish and other fishery resources, although specific species vary
among regions depending on local availability. Figure 17 shows broad culture areas in which pottery,
burials, tool technologies, and architecture can be use to distinguish among the various coastal Archaic
groups. Shell-ring shape, size, and presence/absence are significant traits that separate at least twelve cul-
ture areas.
Everglades
Sites unknown
No rings/mounds
1. Thoms Creek
The Thoms Creek region is found in South Carolina, extending from the middle coast at Sewee shell ring
to the Savannah River. New discoveries by DePratter (2005), however, indicate that shell rings may be
found even farther north. Pottery found at shell rings is predominantly sand-tempered Thoms Creek ware,
with minor amounts of fiber-tempered Stallings ware occasionally present. Radiocarbon dates range from
4180 to 2885 B.P., with the most recent dates coming from the Lighthouse Point ring—a site whose dates
seem out of sync with most other Thoms Creek ring dates, which typically range between 4100 and 3600
B.P. Rings range in size from 45 to 100 meters in diameter, are circular to C-shape, and may occur in iso-
lation, multiple groupings, or conjoined with other rings, as found at Fig Island 1, Coosaw 1 and 2 and
Skull Creek. The pottery and radiocarbon dates indicate contemporaneity of the rings at each of these
multiple- and conjoined-ring sites or, at least, an overlap in occupation. Hence, larger populations and
more complicated levels of social organization are found at these sites than at the single-ring sites. Thoms
Creek rings range from relatively low-lying and even shell distributions, such as at Fig Island 2 and Pat-
ent, to hypertrophic and asymmetrical distributions, such as at Fig Island 1. This constructional asymme-
try has been linked to asymmetry in social organization. The larger shell rings and mounds and ring com-
plexes reflect differences in group size and in status and power among group members (Russo 2004). At
Fig Island, Ring 1 exceeds 6 meters in height, while Ring 2 is only 2 meters high.
The use of shell tools at Thoms Creek shell rings is relatively rare, with only sixteen tools reported for
rings other than Fig Island and Sewee. It is possible, however, that these tools have gone unrecognized or
under reported at Thoms Creek shell rings. At Fig Island, for example, one of the specific project goals
was to recover a large sample and range of shell tools. In total, 125 shell tools and fragments were identi-
fied, more than at all other Thoms Creek shell-ring assemblages combined. Gouges, hammers, awls, net
spacers and spokeshaves were recovered (Table 5) (Saunders 2002:118–119). Bone pins, identified at ten
of the known Thoms Creek rings, are more evenly distributed (Table 6). Projectile points have been re-
covered from seven shell rings, and baked clay objects from two. (Tables 3 and 4).
In terms of the place of shell rings in the larger pattern of settlement, other non-ring sites associated with
shell rings are not well known. One model suggests that amorphous middens represent base camps, while
shell rings served as communal centers (Michie 1979). Another suggests that shell rings were the base
camps or villages of Thoms Creek coastal settlement (Trinkley 1980a:312). A third suggests that shell
rings may represent both villages and ceremonial centers, and it is up to the archeologist to figure out the
function of each shell ring empirically rather than typologically (Russo 2004).
Regardless of which model best fits the data, no “amorphous” middens are known to be directly linked to
shell rings, or even nearby. Some amorphous middens lie within a few miles of rings (e.g., Bilbo, Daws
Island, Venning Creek, Bass Pond). The pattern seems to be that shell rings and ring complexes are iso-
lated from other site types by considerable distances. The general paucity of large non-ring middens in the
coastal area suggests that rings functioned as nucleated centers, both permanently occupied and periodi-
cally enlarged by rural coastal populations, upland interior guests, and/or other visitors attending feasts
and other rituals held at the rings. Thoms Creek sites are found in the interior, upriver locales, suggesting
a cultural connection of ring occupants with folks from the interior coastal plain, although the precise na-
ture of those connections are not yet known. The most detailed attempt to link coastal and interior groups
diachronically and culturally has been forwarded by Sassaman (1993), who identified specific Thoms
Creek groups distinguishable from each other by pottery characteristics. Such analyses hold the promise
to define more precisely the ethnic groups that made up Thoms Creek ring-building cultures. Shell rings
in the Thoms Creek regions exhibit contact with the Stallings/St. Simons ring builders to the south by the
presence of Stallings pottery, which is usually more abundant in rings closer to the Stalling/St. Simons
boundary (Heide and Russo 2003).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
In summary, Thoms Creek shell rings are characterized by: circular and C-shapes up to 6 meters in
height; Thoms Creek sand-tempered pottery; moderate amounts of pointed bone tools and ornaments; oc-
casional baked clay objects; few shell tools; and few lithic projectile points. The rings seem to have func-
tioned as nucleated villages and ceremonial centers, with a few supporting larger populations and greater
ranges of status as evidenced by multiple rings and mound architecture. Other habitation sites and re-
source procurement stations in the area lack evidence of ceremonial or monumental architecture.
2. Stallings/St. Simons
The Stallings/St. Simons region extends from the Savannah River along the coast of Georgia to St.
Simons Island, some 30 miles north of the Florida border. The region is identified by coastal shell rings
whose size range, 35 to 85 meters in diameter, reflects those found at Thoms Creek sites. Like Thoms
Creek rings, Stallings/St. Simons rings are generally circular or C-shaped, with one ring, the West ring on
St. Simons Island, apparently U-shaped. (No contour map of the site exists.) It is perhaps here that influ-
ence from the southern ring builders in Florida, characterized by their U-shaped rings, can be seen. As in
the Thoms Creek region, multiple ring sites are present, as exemplified by the Sapelo complex. There,
two large rings, approximately 85 meters in diameter each, were separated by a small tidal creek. A third
smaller ring stood nearby.
Aside from ring sites, large coastal sites are few, while smaller sites are more numerous (Howard and
DePratter 1980:247). This suggests that nucleation occurred at the larger ring sites, which were probably
permanently inhabited (Marrinan 1975; Waring 1968c:245). As with Thoms Creek shell rings, periodic
aggregation related to ceremony and ritual likely occurred as groups from the coast and interior uplands
met at the rings. Unlike Thoms Creek, at least one large amorphous midden is known to be directly con-
nected with shell rings at the Sapelo complex (Simpkins 1975; Thompson 2006). This suggests larger
populations at Sapelo than the sizes of the rings alone would indicate.
In determining the range of this ring-culture area, no rings have been found along the south Georgia coast
between Cannon’s Point and Rollins, some 60 miles. This is a great distance compared to the frequency
of Thoms Creek and Stallings/St. Simon ring distribution further north, which ranges on average from 7
to 10 miles between rings. This suggests a cultural boundary, an idea reinforced by pottery. In the
Stallings/St. Simons region, Stallings/St. Simons fiber-tempered wares are predominantly found, except
in the southernmost rings and nearby sites where Florida Orange pottery may alternately dominate (e.g.,
Marrinan 1975). To the north, Thoms Creek wares are found as minority wares in the Stallings/St. Simons
rings (Thompson 2006), suggesting some sort of cultural interaction between the Thoms Creek and
Stallings/St. Simons regions. Shell tools are comparatively rare, with a total of only thirty reported from
seventeen rings. Bone pins are reported from six ring sites, but one of these is Walthour, which is proba-
bly not a Late Archaic ring. Possible baked clay objects have been reported from four rings, with one,
Sapelo 1, yielding the vast majority. Only seven lithic projectile points have been reported. Radiocarbon
dates range from 4590 to 3470 B.P., with most dates falling between 4200 and 3600 B.P. at the few rings
that have been dated. This age pattern reflects the time period associated with Thoms Creek rings.
In summary, Stallings/St. Simons shell rings are characterized by: circular and C-shapes up to 3 meters in
height; Stallings/St. Simons fiber-tempered pottery; moderate amounts of pointed bone tools and orna-
ments; occasional baked clay objects; few shell tools; and few lithic projectile points. The rings seem to
have functioned both as nucleated villages and ceremonial centers, with one supporting larger populations
and greater ranges of status as evidenced by multiple rings and associated habitation areas within the ring
complex. Other habitation sites and resource procurement stations in the area lack evidence of ceremonial
or monumental architecture.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
3. St. Marys
The St. Marys region extends some 80 miles from Cumberland Island, Georgia, to St. Augustine, Florida.
The northern half of the territory is marked by small, non-mounded shell middens with no shell rings,
while the southern half contains: two large U-shaped shell rings (Guana and Rollins); a large circular ring
(Oxeye); large, mounded shell middens; widely scattered small shell middens; and large extensive sheet
middens. The absence of shell rings in the northern part of the region is not understood. But Florida shell
rings are generally more widely dispersed. Compared to the distribution of Thoms Creek and Stallings/St.
Simons rings every 7 to 10 miles, the contemporary St. Marys rings are 30 miles apart. This distribution
suggests that ceremonies held at St. Marys rings drew from a more widely dispersed population base
along the coast.
Common to both the north and south portions of the region is the use of fiber-tempered Orange pottery at
all types of sites. At Rollins and Guana, lithic projectile points and baked clay objects are rare. But at the
pre-pottery Oxeye site, baked clay objects are common, suggesting a different cooking technology.
Pointed bone tools and ornaments were recovered from all three rings, while shell tools were present, but
not common.
Unlike most Thoms Creek and Stallings rings, each of the St. Mary’s rings lies adjacent to large amor-
phous shell middens (that is, villages) and are surrounded by smaller household shell middens. This sug-
gests association with both rural homesteads, or hamlets, and nucleated villages in the immediate area
(Russo 1993; Russo et al. 1993, 2002). The question of whether the rings themselves served as permanent
nucleated villages is still being debated (Russo 2004; Russo and Heide 2002; Russo and Saunders 1999;
Saunders 2003). But whether it was the rings that were permanently occupied, serving as both habitation
and ceremonial centers, or the nearby village with the rings serving as their ceremonial centers, the St.
Marys rings are larger in size and, presumably, home to larger occupations than most of the Thoms Creek
and Stallings ring sites. The three rings in the region range from 150 to 250 meters in outside diameter or
length, three times the average of the Thoms Creek and Stallings rings. The openness of the U-shape of
the two largest rings suggests that expanding or variable populations were attendant with ceremonial ac-
tivities. At Rollins, up to nine ringlets are attached to the main ring. These suggest communal areas for
smaller groups, where activities of smaller scope take place than those in the greater ring. It is intriguing
to postulate that these smaller rings may have served as guest, kin, or sodality-based areas for ceremonies.
They are contemporary with the main ring rather than the later add-ons.
Different shape, size, distribution pattern, and associated settlement features separate the Thoms Creek
and Stallings/St. Simons rings from the St. Marys rings. One reason for these differences lies in the re-
gion’s physiographic setting. Absent in the St. Marys region are the piedmont drainages of the interior
uplands along rivers entering the Atlantic perpendicular to the coast. The two main rivers in the region,
the St. Marys and the St. Johns, drain interior lowlands. Both lack recognizable Late Archaic sites away
from the coast. In addition, the Okefenokee Swamp presented a barrier between the interior uplands and
the St. Marys coastal region. Combined, these features suggest that permanent populations and ceremo-
nial aggregations of people at ring sites in the region were likely derived almost exclusively from people
living in the coastal zone, whereas periodic visits from interior groups are suggested for Thoms Creek and
Stallings/St. Simons coastal cultures.
While the construction of Oxeye around 4600 B.P. preceded the adoption of pottery, Rollins and Guana,
contemporary at around 3800 to 3500 B.P., yielded fiber-tempered Orange pottery. These rings are large,
have associated amorphous shell middens nearby that may represent villages, and are surrounded by
smaller sites suggestive of rural homesteads. Aside from surface designs, the pottery does not differ sig-
nificantly from Stallings/St. Simons wares. So too, the shell, bone, and lithic assemblages are similar and
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
occur in roughly the same frequencies. The rings do not seem to have had direct access to interior sites as
did the Thoms Creek and Stallings/St. Simons cultures.
4. Orange
The coastal Orange region is characterized by the absence of shell rings. It stretches some 200 miles along
the coast paralleling the north-flowing St. Johns River, which lies only 10 to 30 miles inland. The areas
along both the coast and the St. Johns River are densely packed with Late Archaic sites marked by the
presence of Orange pottery. If shell rings are aggregation sites for interior peoples who brought with them
exotic goods (e.g., lithics), then the absence of rings in the Orange region may be due to the lack of inter-
action with the interior populations who had ready access to prestige items (i.e., chippable lithics, which
are not found on the St. Johns River). Some have suggested that the interior and coastal sites actually rep-
resent two seasonal aspects of the same population (Milanich and Fairbanks 1980). While others have
suggested that both the St. Johns River and coastal sites were settled permanently, and though the popula-
tions may have been closely related (e.g., they made the same pottery), they represent distinct social
groups (Russo et al. 1992).
Important to the understanding of why and how cultures built and used shell rings, the Orange region
provides a controlled laboratory in which many of the temporal, environmental, and social aspects that
characterize most ring-building traditions are found, but in which rings were not constructed. If rings
were used for large-scale feasting, the large coquina shell and oyster shell mounds found in the region
may have been alternative feasting sites, supplanting or precluding the need for shell rings. In fact, the
region is best known for its enormous shell mounds, along both the St. Johns River and the Atlantic coast,
that far exceed the size and numbers of any contemporary constructions in the Southeast. These shell
mounds are known for the presence of human burials, both scattered remains and full burials (articulated
or bundled) (Rouse 1951:129), which suggests that mortuary traditions differed from those of ring-
building cultures. On the coast, burial traditions may have also differed. At least one ceremonial mound
site has been identified in which twelve conical mounds made of alternating layers of shell and sand con-
tain human remains. This center may have served as a sacred ceremonial burial site (Piatek 1994).
Shell tools are known from both interior and coastal sites, but comparative quantifications have not been
undertaken. Lithic points can be found, but are relatively rare, as are baked clay objects. Bone pins are
widespread (Rouse 1951). Fiber-tempered pottery consists of Orange and Tick Island series. Orange, in
particular, is abundant in the region.
5. Reed
The Reed culture of Southeast coastal Florida contrasts sharply with the Orange culture to the north. The
latter is characterized by the dense distribution of large and small shell-midden sites, but no shell rings,
whereas in the Reed region, the Reed shell ring is the only known site. It is among the largest, if not the
largest, Late Archaic shell ring in the Southeast. This suggests it entertained a relatively large permanent
population or large periodic aggregations of people. But where these people may have gathered from is a
mystery.
No fiber-tempered pottery has been recovered from the site, despite its proximity to the densely populated
Orange region to the north, where fiber-tempered pottery is abundant. Only spiculate and sand-tempered
St. Johns and Glades wares have been recovered (Russo and Heide 2000, 2002). Prior to the discovery of
this site, people who made these kinds of pottery were not to be found in the region until a thousand years
after the shell ring had been built. For these and other reasons, it may be presumptuous to call this lone
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
site a shell-ring region. But the site is distinguishable from other rings by its use of unique pottery types.
On the other hand, its large U-shaped shell ring connects it to traditions of ring builders found throughout
Florida. In the absence of other known sites, speculation is that people gathered at the ring from nearby
coastal sites, subsequently drowned. However, the nearby interior Everglades has recently revealed a
large Archaic population previously unknown (Russo 2005; Schwadron 2005) and this may be the source
of the draw for ceremony and trade. The site exhibits evidence of ceremonial feasting and habitation.
6. Everglades
The Everglades region contains no shell rings on its east and south sides. As stated above, a large, widely
dispersed, previously unknown Late Archaic population has only recently been identified in the interior
freshwater Everglades. Coastal occupations are virtually unknown.
7. Bonita
The Bonita culture area is described by two well-dated shell rings, Bonita and Horr’s Island, and possibly
five to eight undated rings lying south of Horr’s Island in the mangrove swamps of the Ten Thousand Is-
lands (Beriault et al. 2003). In addition, the Hill Cottage shell ring some 80 miles north of the Bonita ring
may be more appropriately included in the Bonita area, at least in its early stages of construction, rather
than assigned to a separate region. Both the Bonita and Horr’s Island rings, as well as the Hill Cottage
ring, are characterized by U-shapes. Heights/depths of ring deposits extend to 4 meters, although the Bo-
nita ring has been affected by road construction. The Bonita region rings, unlike those from other regions,
lack pottery, as does the lower levels of Hill Cottage. The rings both pre-date the adoption of pottery in
the Southeast (Horr’s Island) and are contemporaneous with its early introduction in their later stages of
use. The coeval traditions of shell ring and pottery production, however, did not exist in this region. The
region is also characterized by well-developed shell tool technology and mound ceremonialism. It is un-
known what ceremonies these mounds were associated with. But burials were placed in some of them
(Mound A and Mound B at Horr’s Island) after their construction. The presence of the mounds and rings
together suggest that the sites functioned both as places of ceremony and habitation.
The temporal range of the two shell rings, and the lower range of Hill Cottage is 4660–3870 B.P. Like the
Reed region, the shell rings in Bonita seem to lack ancillary sites from which populations may have been
drawn for aggregation at ring ceremonies. This may be a result deficiencies in the archeology of the re-
gion as opposed to a real settlement pattern. The fact is that without pottery being present, shell midden
sites without recognizable or temporally diagnostic artifacts are often glossed over in summary reviews of
site distributions. The coastal groups undoubtedly had some relation with interior groups for shell tool
artifacts have been identified at interior sites and cemeteries. But the nature of those relations are largely
unexplored. Trade seems even more limited at these rings than at others nearer lithic resources. Only one
lithic projectile point has been recovered at the rings, and the pointed bone tools are of types that gener-
ally lack the decorations of more northern ring-building cultures.
8. Cottage
The Cottage region is based solely on one ring, and the presence of other non-ring sites in the region that
have yielded fiber-tempered pottery, variously referred to as Orange or Norwood depending on the rela-
tive amount of sand also found in its paste. In terms of shell rings, the only trait that distinguishes the Hill
Cottage region from the Bonita is the presence of pottery and the more recent dates associated with the
pottery (3975–3625 B.P.). There are more sites known in the region that date to this period, and so the
pool of homesteads from which ceremonial aggregations may have drawn is not as puzzling as that found
in Bonita.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
9. Big Bend
The Big Bend of Florida, stretching from north of Tampa Bay to the Apalachicola delta, is marked by low
lying shorelines subject to extensive sea transgression with minimal rises in water levels. Because of this,
few Late Archaic coastal sites have survived rising sea levels over the last 5,000 years. Sites are widely
spread and consist of small shell middens and burials exposed as erosional features. Whether this regional
culture made and used shell rings is unknown.
10. Panhandle
The Panhandle of Florida has yielded at least two possible Late Archaic shell rings, Meig’s Pasture and
Buck Bayou, each dramatically different in shape and size from each other and from other Florida rings.
Neither site contains pottery of the period, but do contain, baked clay objects and, apparently, more than
the usual numbers of lithic artifacts (Thomas and Campbell 1991). These traits have prompted archeolo-
gists to associate the rings with Poverty Point rather than other shell-ring-building traditions. But without
enumeration of the artifacts it is difficult to determine if they differ in any substantial way from other
shell-ring-building traditions. In the absence of pottery and the promise of baked clay objects, the sites
distinguish themselves from all other ring builders. Too, the promise of a lithic tool kit unknown at other
rings makes these proposed rings very different. However, there are not many commonalities in shape and
size that might characterize the rings of this region. Meig’s Pasture is not a contiguous ring of shell, but a
series of pits. And the “plaza” it surrounds, has been suggested to have been a spring head. Buck Bayou,
on the other hand, is mounded shell in an arcuate shape 1.5 meters (0.45’) high. But it does not seem to
have much of a plaza at all. Shell is abundant in the “plaza” area. These factors should not dismiss the
sites or other potential sites in the region from consideration as shell rings. After all, Hill Cottage was
once thought to have surrounded a spring head, and a number of shell rings extensively damaged or
mined also seem to have limited shell free plaza areas. These factors just suggest more work is needed to
characterize the regional shell rings. Unlike other shell rings, these sites have been identified as trading
centers rather than habitation or ceremonial sites (Thomas and Campbell 1991). If true, that trait would
make them unique among shell-ring-building traditions.
As for intrasite development, their distributions of shell and artifacts indicate that the raised circles served
as areas of habitation with living floors, hearths, and special activity areas. At Claiborne, mounded and
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
hypertrophic deposits of shell reflecting quick deposits associated with large-scale feasts seem to be few
compared to layered deposits of shell containing abundant crushed shell, artifacts, and soil. The latter
suggests that daily maintenance activities characterize the site, rather than ceremonial large-scale feasting.
Burials may be present in the ring or in the nearby sand mound (Bruseth 1991), but none were ever found
to confirm this speculation. An abundance of steatite, chipped stone tools, and lapidary items sets both
sites apart from all other known shell rings. The presence of over 12,000 baked clay objects distinguishes
Claiborne. But the presence of a only a few baked clay objects at Cedarland (Gagliano and Webb
1970:51) is right in line with other shell rings. Together, however, the unique features at these two rings
serve to separate them as a distinct building tradition from those of the Atlantic coast, and, even from
those in the Florida Panhandle.
This brief summary of varying ring-building traditions presents ways in which shell rings and associated
traits are critical factors in exploring the rise of traditions, the development of group territories, and the
establishment of ethnic identities in the earliest coastal occupations of the southeastern seaboard.
The transformation of nomadic hunter-gatherers to coastal fisherfolk is a pivotal step in the evolution of
precontact United States cultures. The move reflects changing perceptions of the habitable landscape by
Late Archaic coastal settlers. Archeologists have suggested that prior to the Late Archaic, sea level was
too low or the rise too rapid to provide a stable environment for the growth of the estuarine shellfish and
fisheries resources that would thereafter come to support coastal settlements (Brooks et al. 1989; Miller
1998:72; Widmer 1988). Others have suggested that coastal settlements may have preceded the Late Ar-
chaic, but were later drowned and lost to archeological observation by risen seas (DePratter and Howard
1980:33; Russo 1996). Being among the earliest and largest of early coastal sites, shell rings represent a
keystone to our understanding of Archaic hunter-gatherer adaptations to coastal environments. Some
rings sit high and dry on bluffs and dunes above present-day coastal marshes. Others are being lost to
tidal erosion. Still others lie buried beneath the marsh. Rings were built on landscapes at a point when
significant environmental changes were occurring. The historic context, Settlement and Environment, is
critical to understanding the natural landscape upon which shell-ring builders founded early large-scale
coastal settlement, could prosper as sea levels stabilized, and foundered upon sea level rise. The context
falls under the National Park Service’s Framework Theme, Peopling Places and Transforming the Envi-
ronment.
Before the recognition that sea level has risen substantially over the last 5,000 years, shell-ring locations
in coastal marshes were seen as intentional. Rings were often interpreted as fish traps, designed to capture
fish swept into the rings at high tide (Edwards 1965). Today, the occasional occurrence of shell rings in
marshes is recognized as the result of risen seas. All shell rings were originally placed on dry land adja-
cent to marshes, but not in the marshes themselves. The central, shell-free areas functioned not as fish
traps, but as community nexuses, places of public intercourse. Rings and associated architecture were of-
ten placed on the highest points of land to proclaim territory, facilitate seeing and being seen, and, per-
haps, to mitigate against environmental stresses of weather and biology (e.g., storms and mosquitoes). In
low-lying environments, rings were often the tallest features on the immediate landscape. As such, the
construction of rings altered precontact environments and created new landscapes, raising dunes and
bluffs even higher, and providing high points on islands next to marshes. Today, these topographic gradi-
ents, combined with the abundance of shell, present a well drained, calcareous environment ideal for
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
plants and animals not otherwise native to the coastal zone. They function as environmental refugia of
plants and animals rare or absent elsewhere in the region (Dorroh 1971).
Rising seas are also seen as causing the abandonment of the coastal traditions centered around shell-ring
construction. The last shell rings were constructed around 3000 B.P., and at this time large changes in the
coastal environment and cultures that occupied them can be seen. As sea level rose, rings, mounds and
habitation sites were flooded causing great changes in established settlement patterns, and concomitant
changes in social organization. Shell midden sites became smaller across the southeast United States, pot-
tery styles changed, and large-scale public architecture ceased being built for centuries to follow. The oc-
cupation of the Southeast by societies that organized around permanent, nucleated settlements, con-
structed public works, and feasted en masse at ceremonial sites, changed to occupations by isolated
groups, small in scale and dispersed across the coastlines.
The rising seas at the end of the Archaic were not sufficiently rapid to destroy environmental productivity
of the larger coastal zone. But they were sufficient to adversely affect specific localities. Oyster beds be-
came too saline to support growth. Productive fishing grounds shifted laterally or landward to areas where
freshwater input and nutrients held in the advance of rising seas. And homesteads became more fre-
quently flooded with daily rains and tides. The movements of resource zones and the need to move habi-
tations change the social landscape. Rings were now marking territories that were not productive enough
to support the resident population. Movement to other territories was restricted due to increased popula-
tions and ethnic boundaries brought about by the exploitative success of ring-building cultures. Strife may
have arisen as populations sought to feed themselves by intrusion into others’ natural and social environ-
ments.
Five thousand years ago, shell rings were among the largest architectural features in the United States.
Shell rings and their associated sites represented pioneer forms of social organization capable of perma-
nent occupation, large-scale public works, and endurable residence among increasing populations. While
archeologists may argue over the function of shell rings, most agree they were constructed by peoples
whose immediate ancestors were the hunter-gatherers of interior forest lands. How these interior peoples
were able to change their subsistence and settlement patterns and master the coastal environments in a
relatively short period, and how they were able to leave a legacy architectural features that have defied
millennia of coastal storms and erosion, sets shell rings apart as national landmarks in the precontact his-
tory of the United States. The exploration of how the environment changed through time, how it com-
pelled social reaction, and how it affected end of the ring-building traditions of culture are also part of the
significance held in shell rings.
Arguably, the most important context related to the national historic significance of shell rings is the
promise to yield information critical to the understanding of the causes for and changes attendant with the
move from egalitarian to more complex forms of social organization. Few archeological sites have been
deposited or are as well preserved at those rare temporal points when such transitions have occurred as
are shell rings. The simple facts are these. Before the Late Archaic, there were no permanently settled
villages in the United States containing populations larger than those typically associated with migratory
hunter-gatherers (Russo 2004). Base camps, aggregations sites, and trading nexuses often resulted in large
Middle and Early Archaic settlements. But these temporary settlements dissipated after seasonal resources
beckoned the groups elsewhere. With the Late Archaic, and as particularly found at shell-ring sites, per-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
manent, year-round settlement of sites was a norm, ethnic boundaries defining and defending those sites
arose, populations grew, and complexity in site structure at some of the largest shell-ring sites achieved
levels theretofore unseen in the archeological record.
The equation of large-scale public works such as shell rings with social complexity has been a persistent
issue in the archeology of the Archaic southeast United States. Archeologists have pointed out that there
is no necessary connection between the construction of mounds, for example, and permanent settlement
or hierarchically organized societies (J. Saunders 2004). Others, however, have argued that a subsistence
infrastructure, relative sedentism, labor, and leadership with the capacity to build large public works be-
yond that found among migratory hunter-gatherers, is reflected in Late Archaic mound and shell-ring-
building societies (Russo 2004; Sassaman and Heckenberger 2004). Moreover, the structure of circular,
mounded settlements with specific mounds and ring walls placed in opposition to others represent asym-
metrical architectural constructs reflecting asymmetrical divisions in social relations (Russo 2004; cf.
Trinkley 1985).
Combined with shell rings, the construction of other large-scale architectural features has suggested to
some that shell-ring sites may represent the first evidence of changing social organization from simple
egalitarian hunter-gatherer bands to kin-based tribal organizations in which social status distinctions were
achieved and maintained through the sponsoring of public works such as the building of shell rings. Oth-
ers have suggested that shell rings are little more than the incidental refuse of simply organized hunter-
gatherers who chose to live in circular communities. As such, shell-ring sites offer a unique opportunity to
explore competing theories as to how societies changed from simple forms of social organization to more
complex forms, versus how Middle Archaic nomadic bands maintained a simply organized social struc-
ture while adapting to changing and unique environments.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Shell rings are large-scale architectural features of the Late Archaic located on the Atlantic and Gulf
coasts from South Carolina to Mississippi. Of the fifty-one (counting attached rings as one) known or
suspected rings, most are found in South Carolina and Georgia, with nine known in Florida, the two in
Mississippi have been destroyed. In all regions where they are found, shell rings are the earliest evidence
of permanent settlements on the coast. The primacy of their occupation represents a pivotal time in the
evolution of society in the precontact period United States. Shell-ring builders appear to be the first
groups to change from a migratory terrestrial-based hunting-gathering strategy of subsistence to settled
subsistence strategies. This switch to less mobile resource exploitation, combined with the abundant re-
sources, allowed for significant population growth for the establishment of multiple ethnic identities rei-
fied in ceramic, shell, and bone tool technologies and, most distinctively, for the large-scale public works
represented by the rings and associated architecture. Despite regional differences, shell rings hold in
common evidence of habitation, feasting, ceremony, and monumentalization as seen in the large-scale
mounding of shell surrounding centralized plazas.
Section E introduced ten major contexts and their applicability to the NPS Thematic Framework for un-
derstanding the significance of shell rings on a national level. These contexts, which are summarized in
Table 8, provide the basis for evaluating the significance of individual shell rings for NHL determination.
CRITERIA FOR THE EVALUATION OF SHELL RINGS AND ASSOCIATED LARGE-SCALE ARCHITECTURE
Information contained in the historic context may be used to evaluate significance of Late Archaic shell
rings for designation of National Historic Landmarks (NHL) and as properties for listing on the National
Register of Historic Places (NRHP). To be considered for listing in the NRHP, shell-ring sites and dis-
tricts must possess integrity of location, design, setting, materials, workmanship, feeling and association
at national, state, or local levels of significance relating to Criterion D, the potential to yield important
information.
NHL criteria reflect a more rigorous framework appropriate for evaluating properties possessing the po-
tential to contain information of the highest level of national significance. Guidelines for meeting the cri-
teria relevant for shell rings are found in 36 CFR 65.4. Of the six criteria, the most relevant criterion is
Criterion 6:
Have yielded or may be likely to yield information of major scientific importance by re-
vealing new cultures, or by shedding light upon periods of occupation over large areas of
the United States. Such sites are those that have yielded, or which may be reasonably ex-
pected to yield data affecting theories, concepts, and ideas to a major degree.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Table 8. NHL Shell-Ring Historic Contexts and NPS Thematic Framework Themes
Some of the contexts of national significance outlined in Section E that shell rings have the potential to
address include the invention of pottery, the development of coastal technologies, the origin of settled
communities, the rise of socio-political complexity, and the adaptations of cultures to and by the changing
environment in the face of global warming and the attendant rise in sea level.
As with NRHP Criterion D, NHL Criterion 6 is the typical criterion used to evaluate archeological prop-
erties for significance as NHLs. But shell rings differ from most precontact sites in that they can represent
extant and well preserved architecture representing large-scale monumental construction. As such, and
when of sound physical integrity, it is the architectural design of shell rings, not simply the scientific po-
tential, that provides the important information obtainable under Criterion D (NRHP) and Criterion 6
(NHL).
Important architectural traits are typically listed as criteria for consideration for the NRHP un-
der Criterion C, and Criterion 4 for National Historic Landmark designation. Criterion 4 states:
As such, the architectural traits of shell rings will be considered for their significance under Criterion C
and Criterion 4 as well as Criterion 6. Shell ring sites include not only scientific information but architec-
tural significance related to physical characteristics of design, construction and form that sets shell rings
apart from other precontact shell middens. The architectural significance is clearly the intent of Law-
rence’s (1990b) NRHP Multiple Property documentation when he states that resources eligible for the
NRHP must include integrity of original geometry.
REGISTRATION REQUIREMENTS
To be considered for nomination as a National Historic Landmark under this historic context, a shell ring
must demonstrate they possess all of the following criteria:
Lawrence (1990b) attempted to establish criteria for site integrity in NRHP Multiple Property documenta-
tion for South Carolina shell rings. He suggested that for a shell ring to be considered eligible for the
NRHP it must
have its base intact and near horizontal if tested or excavated and mapped, and will have
a significant thickness of preserved and undisturbed shellfish remains (normally 18
inches or greater), thus supporting the interpretation of the site as an original ridge or
topographic high.
It is not clear what he meant by “near horizontal,” but his suggestion that at least 18" (45 cm) of shell ris-
ing above the ground must remain for eligibility to the NRHP is probably too restrictive. Thompson
(2006) has clearly demonstrated that the shape of a shell ring can be determined as intact, and scientifi-
cally important information can be found below ground surface, even when all above ground, topographi-
cally raised ring walls have been removed. For the purposes of this historic context relative to NHL
status, scientific integrity means any remaining portion of the shell-ring site must have materials and fea-
tures still in undisturbed context, regardless if they are topographically elevated or below level ground.
These contexts must hold the potential to provide data for answering questions and addressing issues of
historic and anthropological importance on a national level.
Architectural integrity for NHL eligibility consideration requires a higher level of preservation than scien-
tific integrity. It requires that the shell ring remains intact to a degree sufficient to allow the determination
of ring shape, original height, and internal construction. All shell rings that have NHL level architectural
integrity will automatically hold scientific integrity for NHL or NRHP consideration. Note that rings
which may contain scientific integrity will not necessarily hold architectural integrity sufficient for NHL
status. Both architectural and scientific integrity must be obtained at a ring site for it to be considered for
NHL status. These issues of integrity apply not only to shell rings, but to any associated architecture such
as mounds, ringlets, ramps or causeways.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
LEVELS OF SIGNIFICANCE
Local Significance
At the local level, shell rings contain features, artifacts, or ecofacts that can address significant questions
of local interest, but lack such evidence relating to regional or national issues of importance. At such
rings, scientific and architectural integrity may be severely compromised, but sufficient undisturbed data
remain to answer thematic questions of subsistence, cultural affiliation, and age of the site. Such shell
rings may be considered eligible for the NRHP on a local level, but can only be considered of regional or
national significance as a contributing element in a multiple property listing of other shell rings holding
these higher levels of significance.
Regional Significance
At the regional level, shell rings contain features, artifacts, or ecofacts that may be used to address issues
and research questions of regional importance, but lack such data relating to important national issues.
These shell rings may contain sufficient data to address many of the contexts cited in this study, but they
lack sufficient integrity to address all. Sites that lack architecturally integrity sufficient to identify the
original ring form, for example, may still be considered significant on a regional level if the potential to
address other contexts is present. Such shell rings may be considered eligible for the NRHP on a regional
level, but can only be considered of national significance as a contributing element in a multiple property
listing of other shell rings holding this highest level of significance.
National Significance
Shell rings that have the potential to address all or most of the contexts in this study can be considered for
nomination as a National Historic Landmark. The national significance of these contexts is described in
Section E. To address all these, the shell ring must maintain sufficient architectural integrity to ascertain
its original architectural form. If this integrity is present, then the potential for recovering the scientific
data needed to address other contexts will be obtainable. The absence of architectural integrity will pre-
clude nomination as an NHL, although the rings may be considered eligible for the NRHP. They may also
be considered of national significance as a contributing element in a multiple property listing of other
shell rings holding this highest level of significance.
The data in Section F presents summary statistics and descriptions on all sites that have been identified as
Archaic shell rings potentially eligible for nomination as NHLs. Upon review, it turns out that some of the
sites previously thought by researchers to be Late Archaic shell rings are not, in fact, shell rings. These
sites and the shell rings that don’t meet the register criteria established in this section are discussed in Sec-
tion H as “Sites Considered but Rejected as Potentially Eligible for NHL Designation.”
For each shell ring, a site drawing and summary statistics are presented in a prefatory table. Following
each table is a brief overview of investigations conducted on that particular site over the last two hundred
years, including the methods used by the different investigators for obtaining and reporting collections
data. When available, complementary information about the excavations, site size, artifacts, and cultural
affiliations is provided. Precise artifacts counts have not been possible in some cases, because of the in-
vestigators’ contradictory or incomplete data. The symbol # indicates that no numbers were given for the
amount of artifacts. The symbol > indicates that while some researcher listed the number of artifacts, oth-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
ers did not, indicating only that some artifacts were recovered. Pottery type names are reported as the
original researchers identified them. Only known Late Archaic pottery types are listed in the ceramic to-
tals. As used here, the term bone pins refers to a variety of pointed bone objects and fragments, including
awls, decorated and undecorated pins, fids, and punches. In their reports, many investigators used the
term bone pins in a generic sense, and we defer to this generic use of the term herein.
In the summary tables, the site maps are the most detailed of those available. (With few exceptions, no
excavation units or features are shown.) Diameter refers to the largest outside diameter or length of a ring.
Height refers to either the highest topographic point above the plaza, or the thickest shell deposits in the
ring. Area Excavated refers to the total of all excavations and tests, including unscreened backhoe
trenches and excavations. References list the primary references only, particularly those that relate to in-
formation provided in the summary tables. Radiocarbon Dates are corrected (conventional) ages in years
before present (B.P).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Surface Topography
(after Trinkley 1980a:39)
contour interval = 0.1 m
Calmes’s (1967) is the only report containing primary excavation data. He dug a single 5-by-5-foot (1.5 x
1.5 m) unit next to a trench that bisected the ring. Trinkley (1980a:38) believed that Waring may have
originally dug that trench, which is 3 by 30 feet (0.9 x 9 m) on Trinkley’s map (Trinkley 1980a:39). A
contour map of the ring produced by the University of South Carolina. It showed two “old excavations.”
The smaller was 5 by 10 feet (1.5 x 30 m), and the larger, 15 by 20 feet (4.6 x 6 m) (Trinkley 1980a:39).
Calmes saw the ring as made from whole shell placed in contiguous piles. Each pile was separated from
the other by thinner, darkened midden on the piles’ slopes. Calmes hypothesized that these darkened
slopes were created when ring builders leveled off the top of the piles as the ring rose in height, pushing
the dark, crushed shell matrix of “living floors” over the sides of individual piles (cf. Cable 1995, 1997).
Some have criticized Calmes’s theory (Cable 1997; Marrinan 1975:124; Trinkley 1980a:38), but others
think his interpretation of piling plausible (Russo and Heide 2003:46–47). This fairly well-preserved and
protected site is definitely an Archaic shell ring and a potential NHL. It is one of only two shell rings
open to the public, albeit on a limited basis.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
by cult
d ivat
ce ion
du
re
0 30 m
Sketch
(after Lawrence 1989c: cover)
shell
borrowed
shell
borrowed
d
o we
rr
bo
e ll
sh
The site consists of two conjoined rings, variously referred to collectively as Ford’s Skull Creek Shell
Ring (Calmes 1967:7, 26; Judge and Smith 1991:38); Skull Creek Shell Ring (Calmes 1967:25); Skull
Creek Shell Rings (Trinkley 1980a:38); Skull Creek 1 & 2 (Trinkley 1980a:35); Skull Creek (Dorroh
1971:44; Trinkley 1980a:35); or Ford’s Skull Creek 1 & 2 (Trinkley 1980a:40). Individually the ring
components have been referred to as Skull Creek 1 & Skull Creek 2 (Trinkley 1980a:35); Large Ford
Ring and Small Ford Ring (Marrinan 1975:124); and Large Ford Shell Ring and Small Ford Shell Ring
(Calmes 1968:45). This listing may not be inclusive, and, for the record, Ring 2 is synonymous with the
smaller ring’s varied appellations. For brevity, we use Skull Creek to refer collectively and individually to
the rings.
Skull Creek was the first site where conjoined rings were identified. It demonstrated that not all shell
rings are the same in shape. Subsequently conjoined rings have been found at Rollins, Coosaw 1 and 2,
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
and Fig Island 1. The only other 8-shaped ring, however, was found at Coosaw. Calmes (1967) placed
one 10-by-10-foot (3 x 3 m) excavation unit in the plaza of the large ring. Additionally, he placed three 5-
by-5-foot (1.5 x 1.5 m) units, one in each of the rings and one in the plaza of the small ring.
Calmes (1967:7) described the site as “almost entirely destroyed.” A recent visit to the site shows that it
has been heavily impacted by mining, and perhaps looting. It is nonetheless in very good condition with
much of both rings remaining undisturbed. It received the highest rating for shell-ring preservation in
South Carolina’s Heritage Trust 100 Sites list (Judge and Smith 1991:25). It is potentially eligible for list-
ing in the NRHP and contains sufficient integrity for consideration as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0.1
Patent
Barrows
0 .1
0 20 m
Shell Thickness
contour interval = 0.2m
Shape C
Diam. x Ht. (m) 60 x 2
Area Excavated (m2) 2
Radiocarbon Dates 3550+/-60
Stone 0
Bone Uncounted pin
Archaic Ceramics Uncounted Thoms Creek
Other 0
References This document
James Michie, in a personal communication noted that “several uninvestigated shell rings are located in
the interior” of Daws Island (Sassaman and Anderson 1994). An investigation sponsored by the Heritage
Trust Program of the South Carolina Department of Natural Resources is ongoing by the National Park
Service (Russo) and the Louisiana State University’s Museum of Natural Science (Saunders). A single 1-
by-2-meter unit was placed in the south arm of the ring. Abundant Thoms Creek sherds were recovered.
A radiocarbon date on oyster shell was from the bottommost deposits in the unit. This site is related to
Patent Shell Ring (38BU301) in proximity, shape, and cultural affiliation. Possible contemporaneity re-
quires further investigation. Judge and Smith (1991:35) suggested that the ring was oval shape. And in-
deed the footprint is generally oval, but there is an opening on the southwest side and an interior area that
is lower than the ring indicating that the site is C-shaped with an interior plaza. However, much of this
plaza does have shell deposits. The origin of this shell (e.g., in situ midden, colluvium, or storm-
introduced shell) remains open to investigation. The interior of the ring is far less open and level than that
of the nearby Patent Ring. Due to its inaccessibility, the ring has not suffered from historic or modern dis-
turbances and is still very well preserved. It is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0.1
Patent
Barrows
0 .1
0 20 m
Shell Thickness
contour interval = 0.2m
Shape C
Diam. x Ht. (m) 60 x 1
Area Excavated (m2) 2
Radiocarbon Dates 3850+/-70; 3660+/-70
Stone 2
Bone 0
Archaic Ceramics Uncounted Thoms Creek
Other 0
References This document
This ring was investigated along with the Barrows Shell Ring (see previous section). Abundant Thoms
Creek sherds were recovered from a single 1-by-2-meter unit placed in the central wall of this C-shaped
ring. Two radiocarbon dates on oyster shells were taken from the bottommost deposits of shell in the unit,
70–80 centimeters below surface (3660 B.P.), and the upper deposits, 10–15 centimeters below surface
(3850 B.P.). Although the dates appear reversed, they overlap at two standard deviations, suggesting pos-
sible near-contemporaneity or a short interval between upper and lower deposits. Although measuring
about the same size horizontally as Barrows, the Patent ring contains much less shell by volume. It may
have been in the process of construction when abandoned, or simply used for a shorter period of time.
The dates from Patent are somewhat older than the one from Barrows, suggesting that this ring may have
been built before Barrows. Radiocarbon precision is such, however, that the two rings might also be
nearly contemporaneous. With this limited data, there are two possibilities. One, the rings represent a
shell-ring complex, much like such sites as Coosaw and Fig Island. Or, two, the rings may represent a
sequential move to a nearby location. Such moves may be necessary for many reasons and, according to
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
the ethnographic literature, are common for tribal societies that organize themselves in circles (e.g.,
Heckenberger et al.1999).
The ring is in excellent condition. It represents a potential NHL separately or in association with the Bar-
rows ring.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
2
0 20 m
Shell Thickness
contour interval = 0.1 m 1
Attached rings are difficult to describe in terms of metric attributes. The shared wall is often the highest
point in attached rings. Thus maximum ring height is the same for each ring. In terms of shape, Heide and
Russo (2003) suspected that Coosaw Rings 1 and 2 were originally conjoined in the shape of a figure 8.
Ring 1 has since eroded, leaving a C-shaped ring conjoined to Ring 2, a closed ring. A single 1-by-2-
meter unit was placed in Ring 1 to obtain radiocarbon dates and material culture.
Heide and Russo (2003:24) followed Saunders’s (2002) method in distinguishing Stallings from Thoms
Creek wares. However, they did not do fresh-break analysis on all sherds, but looked at surface evidence
of fiber-tempering to classify the sherds as Stallings (cf. Trinkley 1980b:18 for a discussion of the prob-
lem distinguishing Stallings from Thoms Creek, which, to date, has not been satisfactorily resolved).
Despite the erosion on one side of the ring and a looter’s hole, the site is in good condition. In combina-
tion with attached Ring 2 and as part of the multiple ring complex, Ring 1 is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
2
0 20 m
Shell Thickness
contour interval = 0.1 m 1
Coosaw Ring 2 is attached to Ring 1. It achieves its maximum height at the point of attachment. Its dis-
covery highlights the need for a systematic survey of ring areas to identify all associated structures that
may be hidden beneath marsh, forest, leveled lands, or houses and other structures. As with Ring 1, a sin-
gle 1-by-2-meter unit was excavated. (See Ring 1 for a fuller discussion.)
The ring is in good condition. In combination with attached Ring 1 and as part of the multiple ring com-
plex, Ring 2 is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
2
0 20 m
Shell Thickness
contour interval = 0.1 m 1
This ring seems to have been severely impacted from borrowing, plowing and scraping. What remains are
small portions of the aboveground sections of the shell ring, as well as the pits and features that existed
below the ring. At 0.6 meters, the deepest sections of the ring are mostly below ground level. A probe
revealed that a pit feature may extend down to nearly a meter. It’s possible that the ring had not been lev-
eled, but was a ring “in the making.” More work is needed to resolve the issue of its abbreviated stature.
Fifty-four Stallings sherds were recovered from the sole 1-by-2-meter unit placed in this ring. Like Sapelo
3 and Lighthouse Point, Coosaw Ring 3 has the potential to reveal much information about the initial
stages of ring construction. As part of the Coosaw Shell-ring Complex, Ring 3 is a potential NHL, al-
though alone it does not hold sufficient architectural integrity to qualify for inclusion as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Coosaw Ring 1 was identified in 2000 by that inveterate ring-hunter Chester DePratter. The following
year, Christopher Judge found the nearly level Ring 3 in a nearby field. A year later, while mapping Ring
1, Russo and Heide found Ring 2 under dense vegetation on adjacent property to which archeologists did
not previously have access. During the mapping of Ring 2, another ring (Ring 4) was spied farther east
under a house. It was not mapped. This chronology is given simply as background to Ring 4. No attempts
to map it were made because the property owner had not been consulted. Russo walked over it and esti-
mated it to be the size of the other Coosaw rings, 60 meters and approximately a half meter high. More
work is needed to determine its size, shape, cultural affiliation, and condition. It has not been evaluated
for listing in the NRHP. Depending on the results of further investigations, it may be eligible as a contrib-
uting element in the Coosaw Shell-ring Complex for listing as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Fig
Island 3
Fig
Island 1
Fig C
Island 2
A
B
0 50 m
Shell Thickness
contour interval = 0.3 m
Fig Island Ring 1 is arguably the largest, most complex shell-ring structure known. It consists of a volu-
minous ring over 6 meters in height above the surrounding marsh, four or more lower-lying attached
rings, including a concentric configuration containing a ring within a ring, and a mound with a shell
causeway linking it to the primary ring. Ring 1 was recently mapped, but only two excavation units were
placed (Saunders and Russo 2002). A single 2-by-2-meter unit on top of the primary Ring A wall was
taken to a depth of 95 centimeters. This unit only scratched the surface of the shell deposits, which are
nearly 6 meters deep from the top of the ring. A 1-by-2-meter unit was placed in the interior ring of the
attached, concentric Ring C configuration. In order to set a permanent datum, a shovel test (0.2 x 0.3 m)
was placed on the wall of the exterior concentric ring, and artifacts were recovered.
Of the 3,876 sherds reported from all three rings, only 1,788 were enumerated by type (Saunders
2002:131). At least 1,181 Thoms Creek and 110 Stallings sherds were identified from Fig Island 1. These
included sherds from both units and the shovel test (Saunders 2002:101, 131).
Shell tools were numerous and included mostly whelk cutting and pounding tools. Shell beads, clam
tools, and informal shell tools were also identified. Twenty fragments of bone pins were recovered.
This ring was not included in the 1970 NRHP for Fig Island (Hemmings 1970e). While Hemmings rec-
ognized the possibility of more rings under the dense vegetation on Fig Island 1, he did not classify the
site as a ring. Due to ambiguous wording, it is unclear if Fig Island 1 was even included as part of the
nomination of Fig Island 2 and 3. Detailed mapping of the site in 2002 identified at least five attached
rings, a mound and a shell causeway, plus a number of ramps that doubles as ring walls and in-
gress/egress paths to the rings. The mound and causeway are unique features among Atlantic coast shell
rings. The multiple architectural features at the shell complex make it unique among shell rings. As such,
alone or in combination with Fig Islands 2 and 3, the site is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Fig
Island 3
Fig
Island 1
Fig C
Island 2
A
B
0 50 m
Shell Thickness
contour interval = 0.3 m
Fig Island Ring 2’s major excavations were a 5-by-40-foot (1.5 x 12.2) trench dug by Hemmings through
the southern portion of the ring, and a 5-by-125-foot (1.5 x 38 m) trench across its east wall (Hemmings
1970f:10). He never reported any artifact enumeration. Marrinan (1975:121), however, reported that in
his unpublished field specimen catalog, Hemmings recorded 1,765 sand-tempered sherds (presumably
Thoms Creek) and seven fiber-tempered sherds (presumably Stallings). Marrinan (1975:122) cited the
catalog as listing “numerous” bone pin fragments, as well as bone awls, socketed antler projectile points,
and a single “bifacially retouched chert fragment.”
In the 2002 explorations, only two 50-by-50-centimeter column samples were placed in Fig Island 2
(Saunders 2002:106–107). One bone pin and two whelk tool fragments were recovered (Saunders
2002:123, 127); no ceramics were reported.
In the parlance of rings, the shape of Ring 2 is a closed circle. But Russo (2002a) suggested that it really
appears to be more of a hexagonal shape. This shape may also be reflected at the Sea Pines ring, suggest-
ing a culturally significant pattern. Fig Island Ring 2 is among the most symmetrical of the shell rings that
have been mapped, but it is not a symmetrical circle of shell. A path of shell was identified below the
marsh leading from a ramp on its northern wall to another ramp on Ring 3. These architectural features,
combined with radiocarbon dates and ceramics, link the two rings to contemporary usage.
The site was listed on the NRHP in 1970 (Hemmings 1970e). It remains in an excellent state of preserva-
tion. Singly or in association with the other rings and architectural features of the Fig Island complex, it is
a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Fig
Island 3
Fig
Island 1
Fig C
Island 2
A
B
0 50 m
Shell Thickness
contour interval = 0.3 m
Shape C
Diam. x Ht. (m) 49 x 1.9
Area Excavated (m2) 10
Radiocarbon Dates 3990+/-50; 4030+/-50; 4070+/-50
Stone 2 bead fragments, 1 debitage
Bone 7 pins
Archaic Ceramics 325 Thoms Creek, 7 Stallings
Other 17 shell
References Hemmings 1970e; Saunders and Russo 2002
Saunders and Russo (2002) placed a 1-by-8-meter trench and a 1-by-2-meter excavation unit. Probing
revealed a path of thinly scattered shell buried in the marsh between Rings 2 and 3. This suggested an
intentionally constructed pathway. Extensive probing revealed no buried shell on the north side of the
open C-shaped ring, indicating that its shape was intentionally constructed and not the result of erosion.
Artifacts were of the same types as found at Fig Island 1 and Fig Island 2. Two stone bead fragments
were recovered. The site was listed on the NRHP (Hemmings 1970e). It is well preserved. Singly or in
association with the other rings and architectural features of the Fig Island complex, it is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Shell Thickness
contour interval = 0.2 m
Edwards (1965) placed seven 5-by-5-meter excavation units at the site. Of these, one was placed in the
plaza, two others just outside the main ring. In his report, Edwards described at length the placement of
the units, the features and profiles, and his interpretations of ring function, which, he concluded, was a
fish trap. He stated that “almost 10,000 sherds” were recovered, suggesting that nearly all were Thoms
Creek series. Edwards typed the sherds as mostly Awendaw, a classification that has been largely sub-
sumed by Thoms Creek (Trinkley 1976, 1980b). He acknowledges that “some 40” might be fiber-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
tempered, but suggested that the presence of fiber was likely accidental and that all sherds likely belonged
to the sand-tempered Thoms Creek series (Edwards 1965:19). Trinkley (1976:19–20, 65) analyzed 335
sherds from this collection and identified them all as Thoms Creek. Russo and Heide (2003) placed a sin-
gle 1-by-1-meter unit in the ring and recovered 156 sherds of the Thoms Creek series.
Russo and Heide (2003) recovered no artifacts other than ceramics. Edwards (1965) identified ten ce-
ramic abraders, eleven baked clay objects with central holes, and “some three score of shell artifacts, the
vast majority fashioned of conch.” Based on photos of the shell artifacts provided in his report, the vast
majority may have actually been whelk (Busycon spp.) rather than conch (e.g., Melongena corona.). But
clam and oyster and ark shells are also shown in the photos, some of which are identified as “artifacts,”
others as “mollusks.” Their use or fashioning as tools is not stated. As at most shell rings, the paucity of
stone artifacts at Sewee was noted (Edwards 1965:25).
Edwards (1965) mapped the shape of the ring as a C. In fact, the shape had to have been a C for the ring
to have worked as a fish trap. He hypothesized that daily tides brought fish in through the opening in the
C, which was then sealed by a wooden weir to keep the fish in as the tide receded. Archeologists have
agreed with Edwards that the ring was intentionally built with an opening (Gardner 1992:49; Stephenson
1973); have noted, without comment, that the ring is “incomplete” (Marrinan 1975:119); or have sug-
gested that the opening was due to erosion (Cable 1995:108; Trinkley 1976). Russo and Heide (2003)
probed beneath the marsh in the opening and found it to contain up to a meter of shell. They interpreted
this to mean that the shape of the ring was actually originally a closed circle, albeit one with less shell
present in the apparent opening than elsewhere in the ring walls (cf. Stephenson 1973). Here we list the
current shape as a closed circle. But determination of the cause of the closure (e.g., intentional construc-
tion, colluvium) is certainly susceptible to further investigation.
The first radiocarbon date obtained from the site was reported as having come from charcoal (Anderson
and Logan 1981:54; see also Gardner 1992:49; Trinkley 1980b:14). The actual lab sheet, however, identi-
fied the assayed carbon as having come from oyster shell. As such, the date of 3295 has been “corrected”
in the summary table to 3675 +/-110 B.P., bringing it more in line with new dates obtained from the site
(Russo and Heide 2003).
Big chunks of the ring have been removed by mining, and erosion has had an impact on its east side. De-
spite this, the ring remains in good condition. Its architectural integrity is sufficient to ascertain its origi-
nal form. It was listed on the NRHP in 1970 (Hemmings 1970g).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Surface Topography
(after Marrinan 1975:26)
contour interval = 0.2 m
Shape C
Diam. x Ht. (m) 79 x 1.8
Area Excavated (m2) 63
Radiocarbon Dates 4085 +/-90; 4600 +/-90
Stone 2 PP; 1 GS; *flakes
Bone 25 pin; 15 other
Archaic Ceramics 639 fiber-tempered
Other 3 antler; 21 whelk; 2 other shell
References DePratter 1976:131–132; Marrinan 1975
Typically, in U- and C-shaped shell rings the highest deposits of shell are found at the closed portion of
the ring opposite the opening, while the lowest deposits are at the ends of the arms. Cannon’s Point,
sometimes referred to as Marsh Point, is unusual for a C-shaped ring in that its highest point is on the end
of one of the arms. This may be due, in part, to the fact that the end is connected to the mainland, while
the rest of the ring lies in the marsh.
Marrinan (1975:56) excavated nine 3-by-3-meter units around the ring, five in the ring and two in the
plaza. This represents the first shell ring where fine-mesh screen (1/8- and 1/16-inch) were used to re-
cover artifacts and faunal remains. Combined with the large numbers of units placed in the ring, the result
was, for its time, an unusually large collection of artifacts for a shell ring. Marrinan recovered 639 fiber-
tempered sherds and only nine grit-tempered sherds indicating predominant Archaic occupation. Addi-
tionally, thousands of pottery fragments too small for typological identification were also recovered. Bone
pins (i.e., pins, awls, fragments) were abundant (Marrinan 1975:220–222). Only two projectile points and
one ground stone implement were recovered— typical of the low numbers found at shell rings. But hun-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
dreds of lithic flakes were recovered, indicating that when fine-mesh recovery is used, evidence of lithics
may be found at shell rings, even if lithic points remain relatively sparse. Also recovered were twenty-one
whelk pounding tools and one altered clam and one oyster (Marrinan 1975:239–247). The ring has pro-
duced very early radiocarbon ages. It is unusual in that a relatively large number of human remains have
been recovered from it.
This potential NHL is a confirmed Archaic shell ring. Its condition in 1975 revealed a high integrity of
form and context. Its condition is unknown today.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Sketch
(after Marrinan 1975:31)
Shape U
Diam. x Ht. (m) 58 x 0.65
Area Excavated (m2) 8
Radiocarbon Dates 4015 +/-110; 4270 +/-90
Stone 56 flakes
Bone 14 pin; 5 other
Archaic Ceramics 92 fiber-tempered
Other 1 antler
References DePratter 1976:133; Marrinan 1975
Marrinan (1975) only put two 2-by-2-meter units in this ring, which were sufficient to determine its Ar-
chaic origins. (Only 20 percent of all the sherds, and all of those in the upper 15 centimeters of the units,
were from more recent precontact periods.) The shape of the ring has only been sketched, and it is unclear
how its outline was determined, but DePratter (1976: 133) mentioned that it rises over 2 feet (0.6 m)
above a sterile central plaza. Marrinan’s test units went through at least 65 centimeters of shell. The ring
is a potential NHL, but more investigation is needed to identify its boundaries, cultural affiliation, func-
tion, and current state of preservation.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Surface Topography
arbitrary base elevation
(after Simpkins 1975:19)
contour interval = 0.2 m
The Sapelo I ring is arguably the most studied shell ring. First identified by McKinley (1873) in 1872, it
was mapped shortly thereafter and dug by Moore (1897), who may have lost interest in it because it pro-
duced only “earthenware in fragments.” Prior to their excavation, Waring and Larson (1968:268) stated
that a 25-by-50-foot (7.6 x 15.2 m) hole was dug into the southern edge of the ring by the landowner’s
foreman. In addition, a 10-by-10-foot (3 x 3 m) pit was placed in the plaza, probably by Waring (Waring
and Larson 1968:270). In 1968, Waring and Larson reported on their 1950 excavation of a 100-foot-long
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
(30 m) trench that ranged between 10 and 15 feet wide (3 to 4.6 m). It consisted of four excavation units
placed across the west wall of the ring extending into the plaza. In 1975, Simpkins reported on two 2-by-
2-meter excavations he placed just outside the southern end of the ring (Simpkins 1975:63). In 2003, the
University of Kentucky excavated a 1-by-2-meter unit “in a borrow area on the south side of Ring I”
(Thompson 2006:107).
Although Thompson stated that they encountered “a low frequency of Archaic sherds,” he also noted that
only Archaic pottery was recovered from the excavation in which 20 centimeters of midden remained un-
disturbed by historic borrowing. Although Waring and Larson (1968) only listed the artifacts they recov-
ered from Unit 4 in the ring, Thompson (2006:98–103) tracked down pottery from Waring and Larson’s
Units 1–3, presenting quantified totals by level for each unit for the first time, and reanalyzing some of
the pottery that Waring and Larson reported on from Unit 4. Ceramic counts are incomplete because some
artifacts had been lost in storage over the years. Of the ceramics identified, 80 percent in Unit 1 (n=131)
were Archaic, 100 percent in Unit 2 (n=6), and 98 percent in Unit 3 (n=172).
In 1897, Moore provided a sketch map of Sapelo 1, showing it as a closed, fairly symmetrical circle. He
suggested that the two other rings at the site mentioned by McKinley (1873) were no longer visible. He
did not mention any evidence of shell borrowing at Ring 1 or around it. This is the first of many attempts
to find the other two rings (see following sections). In 1975, Simpkins produced a detailed contour map of
Ring 1 and the areas around it that were the supposed locations of Rings 2 and 3. Thompson (2006:82–
84) reported that he had completed a new surface contour map, along with several geophysical maps of
the ring complex. Both of their maps and excavations revealed that the south side of Ring 1 has been dis-
turbed by borrowing since Moore’s site visit at the turn of the nineteenth century.
Sapelo 1 is one of the best preserved and most studied shell rings. Its position as part of a three-ring com-
plex adds to its importance. It is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 25 m
Surface Topography
2.6
arbitrary base elevation
2.6 (after Thompson 2006:20)
contour interval = 0.2 m
2.6
2.4
2.4
2.4
1.2
McKinley (1873) was the first to mention a second ring north of Sapelo 1. But Moore (1897) failed to see
it a few years later. This began the debate as to whether Ring 2 can still be found at Sapelo, or ever ex-
isted. Larson (1998:30) suggested that McKinley may have been mistaken in his identification. Simp-
kins’s map revealed no obvious surface topography to indicate the presence of a ring in the supposed lo-
cation of Ring 2. However, Simpkins identified intact subsurface remains in the area (Simpkins and
McMichael 1976:97). In 2003, Thompson conducted a geophysical survey of the area and found the sub-
surface, unmounded remains of the ring. His map suggested a ring measuring between 75 and100 meters
in diameter (Thompson 2006:82). (The nature of such maps is difficult for any but experts to interpret,
and Thompson does not state a specific size for the ring.) The ring’s 90-centimeter height is taken from
McKinley (1873:422), who stated it was “now rising 3 feet.” A surface topography map made by Thomp-
son (2006) showed a slight elevated arc in the area of the ring he identified as Ring 2.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Simpkins (1975) placed three 2-by-2-meter units in Ring 2, while the University of Kentucky placed one
in 2003 (Thompson 2006:107). Thompson (2006:107–117) provided the counts and weights from both
excavation projects, although not all ceramics from Simpkins earlier dig could be located for analysis.
The upper levels of all these units contained mostly late precontact and historic artifacts along with fiber-
tempered pottery, suggesting occupation and disturbance after the shell ring had been abandoned and
mined. While Simpkins’s units contained abundant shell, Thompson’s unit apparently did not. Simpkins
more likely placed his units in the ring wall, while Thompson placed his in the plaza, suggesting a reason
for the differences in shell abundances (Simpkins 1976:19; Thompson 2006:113). A few lithic artifacts
are listed by Thompson, but they apparently came from mixed contexts.
The radiocarbon date listed in the summary table is problematic. Relative to three dates they obtained
from Sapelo Island, Noakes and Brandau (1974:132–133) stated that “Several shell ring middens exist on
Sapelo. UGa-73, -74 are from a relatively undisturbed ring of ca 50m diam and 2 to 3m high. UGa-75 is
from a remnant of one of the neighboring rings that were partially quarried.” Assuming that Sapelo 1 is
the “relatively undisturbed ring,” then Sapelo 2 seems the likely candidate for the “neighboring ring” that
was “partially quarried.” Sapelo 3 seems a less likely candidate because it expressed even less surface
topography and shell. In the table, we list the date as coming from Ring 2. Thompson (2006:183), how-
ever, interprets all three of Noakes and Brandau’s dates as coming only from a “Sapelo UID ring.”
Ring 2 has been severely damaged. Nonetheless, it contains much valuable information, especially as it
relates to the early stages of shell-ring construction. As such, it is potentially eligible for listing on the
NRHP. As a contributing element, it can also be considered for NHL status with Sapelo 1 and 3.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1.6
1.8 1.8
2
2.2
0 25 m
Surface Topography
arbitrary base elevation
(after Thompson 2006:20)
contour interval = 0.2 m
Shape oval?
Diam. x Ht. (m) 55 x 0.9
Area Excavated (m2) 22
Radiocarbon Dates 3730+/-50; 3730+/-60; 3560+/-50
Stone 1 PP, 81 debitage
Bone 0
Archaic Ceramics 663 St. Simons
Other 0
References McKinley 1873; Thompson 2006
After McKinley’s (1873) early identification of the Sapelo 3 ring, no archeologist recorded finding it
again until 2003 when Thompson, using a geophysical survey, identified shell and midden deposits laid
out in a circular formation at the location long-assumed to be that of Sapelo Ring 3. The shell he found
was not mounded, existed mostly in pits, and lay in widely scattered deposits—a formation hardly resem-
bling the contiguous mounded rings of shell that characterize shell rings. Thus, whether Ring 3, as with
Ring 2, should be considered a shell ring is open to interpretation (see Coosaw III and Meig’s Pasture).
Fortunately, Thompson ground-truthed his geophysical assays sufficiently to conclude that the footprint
of Ring 3 consisted of a series of discontiguous shell-filled pits separated by shell-free midden deposits.
He hypothesized that the shell-free areas were domicile locations (Thompson 2006:223).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Based on McKinley’s (1873) ambiguous statements that “circle No. 2…is 210 feet wide…on a mound
now rising 3 feet (0.9 m)” and that “circle No. 3 is 150 feet (46 m) wide, just like No. 2,” we concluded
that McKinley must have meant that Ring 3 once stood 3 feet (0.9 m) high, “just like No. 2,” since the
two rings obviously had different widths. In contrast, Thompson (2006:83) wrote that “‘just like No. 2’
could mean a variety of things,” and stressed that he does not take the statement “to mean that the height
of Ring II was 3 feet.” It is unclear what Thompson does think it means, if anything at all. But the issue is
important in ring studies. If Sapelo Ring 3 was not ever mounded, as Thompson suggests, it represents the
only known ring that was not. (The currently unmounded Coosaw Ring 3 had been leveled by plowing.)
We describe the pattern of the deposits as oval, but whether the ring was ever a closed structure or con-
tiguous deposit of shell is apparently a matter of debate. Despite McKinley’s statement that Ring 3 was a
mounded shell ring, Thompson concluded that it was never mounded, but rather a ring in the making that
would have become mounded had building continued.
Eleven test units were excavated in Ring 3 by the University of Kentucky. In 2003, eight 1-by-2 meter
units (Test Units 4, 5, 6, 8, 9, 11, 12, and 14), one 2-by-2-meter unit (Test Unit 7), and two 1-by-1 meter
units (Test Units 10 and 13) were placed in the ring and plaza (Thompson 2006:97, 113). We conclude
that the remains Thompson found represent the initial construction stages of a shell ring. Whether that
ring was planned or actually did later rise above ground level is open to further study. But Thompson has
shown that significant deposits capable of yielding important information (Criterion D) remain at the site.
As such, it is potentially eligible for listing on the NRHP. As a contributing element, it can also be con-
sidered for NHL status with Sapelo 1 and 2.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Mound A
(8Cr208)
Ring
0 100 m
Shell Thickness (estimated)
Mound B (8Cr209)
contour interval = 0.5 m (8Cr206)
Mound C
(8Cr207) Mound D
(8Cr211)
Horr’s Island is a shell-ring complex that includes a large shell ring; four shell/sand ceremonial mounds,
one of which is 5 meters tall and connected directly to the ring; a shell ridge not directly connected to the
ring; and a shell ramp leading into the ring (Russo 1991, 1994). All these prominent shell features have
been identified as Late Archaic through a series of radiocarbon dates, most of which fall between 4200
and 4800 B.P. (Only the ring, and not the associated feature dates are listed in the summary table.) The
height mentioned above refers to the depth of the thickest midden deposits at Mound A. Other ring depos-
its away from the mound measure up to 3.5 meters deep. The ring complex itself rises 7 to 14 meters
above the adjacent bay at sea level.
Because Horr’s Island was investigated under cultural resource contracts prior to its planned destruction,
much more area was exposed than is typical with other shell-ring investigations. Both McMichael (1982)
and Russo (1991) used heavy equipment to open large areas in the rings and mounds. Russo, in fact, used
the machinery to reopen trenches dug some ten years earlier by McMichael. One reason McMichael may
not have recovered artifacts (see further) is that he did not examine spoil produced by the backhoe. Russo
placed excavation units next to trenches and screened all the materials recovered through nested meshes
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
(1/16-, 1/8-, and 1/4-inch), with the 1/4-inch mesh producing the greatest numbers of artifacts, and the
smaller meshes greater numbers of vertebrate faunal remains.
McMichael placed a large number of shovel tests in advance of his backhoe operations. No artifact de-
scriptions or enumerations were reported. But it is likely that only shell, not artifacts were identified at the
sites (McMichael 1982:41), and he either did not recognize shell artifacts or did not attempt to recover
them. Subsequently, he placed four backhoe trenches across the ring and mounds, each between 1.5 and 2
meters in width and 4 and 8 meters in length (McMichael 1982:44, 50). Russo (1991) backhoed fourteen
1-meter-wide trenches varying in length from 10 to 96 meters across the rings and mounds. In addition,
Russo placed forty-one units along the trenches; these ranged between 0.5 by 1 meter and 3 by 8 meters.
Together, Russo and McMichael’s excavations totaled 741 square meters.
The Horr’s Island complex and the South Carolina and Georgia rings have in common the limited pres-
ence of stemmed Archaic projectile points and chert debitage (1991:369, 612). Groundstone tools were
far more abundant at the Horr’s Island site (Russo 1991:594–617). These included eighteen metate frag-
ments, four abraders/sharpeners, and forty miscellaneous stones of unknown purpose. At least four stones
were ground into balls, which may have served as net weights. All the artifacts were made from the local
limestone and sandstone, unusable for chipped stone tools, but apparently very useful in other ways. Un-
collected and numerous many large limestone slabs used to line open camp fires.
The Horr’s Island site produced a far wider diversity of shell tools (Russo 1991:594–617). These included
whelk preforms, cups, dippers, scoops, hammers of numerous types, grinders, chisels, celts, adzes, other
cutting-edge tools, columella awls, pins, and pounders, as well as hard clam anvils and notched tools used
for chopping and digging. Similar tools were made from conch (e.g., Strombus sp. and Pleuroplaca sp.).
Perforated arc and scallop shells were abundant and likely used as net weights and decoration; a few shell
beads and perforated oliva shells were recovered.
Further distinguishing the Horr’s Island assemblage was the absence of bone pins and ceramics. The latter
is understandable in that the site predates all of the Georgia and South Carolina shell rings by a few hun-
dred years and, as such, predates the adoption of pottery in the southeastern United States. The absence of
bone pins, however, is presumed to reflect differences in culture traditions.
The shell ring has been largely destroyed by development, and it is unclear how much of the ring remains
intact under the homes. On a recent visit to the site, there were indications that homebuilders are mining
the few undeveloped lots for shell and using it as landfill. Mounds A and D were preserved through
community covenants with the state of Florida. They remain intact, and presumably are protected.
Mounds associated with shell rings are significant architectural features, in part, because they are so rare
and because the hold high integrity. These aspects make the site a potential NHL. Any unaffected rem-
nants of the shell ring could be contributing elements to the mound nomination.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 40 m
Shell Thickness
contour interval = 0.2 m
Oxeye seems to represent Florida’s only closed-circle shell ring. Caution is warranted because less than
half the circle is apparent. The site lies in a marsh setting and is recognizable on the surface only by the
presence of a few small islands of shell in an arc protruding above the marsh. Sea level has risen and bur-
ied most of the ring beneath the marsh muck. Systematic soil probes, however, revealed a complete circle
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
of shell, with the eastern half and portions of the southern rim lying buried 0.1 to 2 meters beneath the
marsh (Russo 2004).
In total, eight 50-by-50-centimeter shovel tests were dug into the buried portions of the ring. Four of these
revealed shell, vertebrate remains, and well-preserved botanical remains, including green palm leaves,
hickory shell, and cut cedar and cypress branches dating to the Late Archaic. On the upland western side
of the ring, a 1-by-9-meter trench (Trench 1) was placed with units dug every other meter (Units 1, 3, 5,
7, and 9).
Oxeye is a pre-pottery Archaic shell ring. Its radiocarbon ages indicate it was constructed a few hundred
years before the introduction of fiber-tempered pottery in the region, thus it lacks Archaic-period ceram-
ics except in surface and disturbed contexts. No detailed report was written on the site’s 1998 excavations
and mapping (Russo and Saunders 1999). However, the field specimen catalogue noted that fiber-
tempered pottery of unstated amounts may exist in FS 25, which, at 80 centimeters below surface, is a
fairly deep provenience within the shell ring at Trench 1, Unit 7. While, at first glance, this deep recovery
seems to suggest that part of the ring midden dates to the Late Archaic ceramic Orange period, the field
form for that level suggests that the artifacts were in an area of disturbed context. At another part of the
site, nine very small, residual sherds were recovered from Shovel Test 2 at 1.5–1.7 meters below surface,
an even deeper context. These were tentatively identified as fiber-tempered Orange sherds. However, the
sherds were from a test in which the shell midden was buried beneath 1.5 meters of marsh. In essence,
these sherds were taken from the surface of the ring, a surface that has been subsequently buried under
thick marsh deposits. Aside from these, only four fiber-tempered sherds and one Deptford sherd have
been definitively identified, and these on the surface of the site. An earlier, preliminary survey placed four
shovel tests. Of the two Orange sherds recovered, one was from the surface; the other’s context was not
noted (Russo et al. 1992:69). In short, the ceramic evidence alone suggests that the site is a preceramic
Archaic shell ring, at which minor occupations of Orange-period peoples occurred on the surface.
Other artifacts were similar to those found in later South Carolina and Georgia shell rings. These included
106 baked clay objects and fragments collected in 1998 and 16 collected in 1993. While Russo struggled
to identify the clay fragments in 1993, with the larger sample in 1998, he recognized them as baked clay
objects (Russo and Saunders 1999). A number of the 1998 specimens had large, spherical shapes. Other
artifacts included a drilled deer bone, a piece of ochre, a cut antler tine, a worked alligator tooth, a bone
pin, a whelk adze, a whelk hammer, and a whelk “tool.” Only two small lithic flakes were recovered.
The Oxeye site is a potential NHL. It not only lies within the boundaries of the National Park Service’s
Timucuan Ecological and Historic Preserve, but the portions submerged beneath tidal waters are owned
by the state of Florida, which protects it from development. Because of its off-shore placement deep in
the marsh, over time large portions of the site have been buried and preserved under protective sediments.
Unique on the Atlantic coast because of its pre-pottery age and artifact assemblage, Oxeye could poten-
tially yield valuable information on the effects of sea-level rise on the environment through soil, faunal,
botanical, artifact, and architectural analyses. It could also provide answers to important questions about
changes from pre-pottery to pottery-making technologies. Finally, despite being partially submerged,
enough of the property is visible to convey its significance under Criterion 4.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
G
0 50 m
Surface Topography
contour interval = 0.3 m
F
H
I
E
D
C
B
A
The Rollins shell-ring complex consists of one primary C-shaped ring measuring 235 meters across, with
a 130-meter plaza. It is unusual for both its large size and the attachment of up to nine rings (otherwise
referred to as ringlets [Russo and Saunders 1999]) to the outside of the primary ring. Despite the diminu-
tive nomenclature, the ringlets are substantial constructions, some exceeding 80 meters in diameter.
During the initial survey, 219 test pits, measuring 0.5 by 0.5 meters, and two 1.5-by-1.5-meter excavation
units were dug to delineate the site (Russo et al. 1992:94). Later, Saunders dug ten 1-by-2-meter units
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
(mostly in ring and ringlet plazas) and a 1-by-16-meter trench across a western ring wall. The earlier sur-
vey recovered 1,191 Orange sherds and four non-descript lithics. Saunders recovered 8,331 Orange
sherds. Both projects revealed that ceramics from the ring were nearly exclusively Orange wares (Russo
1991:Figures 19–22; Saunders 2003:28). Subsequent Woodland cultures used the site, but their ceramics
were found mostly outside the main ring on the eastern side and in small numbers in the ringlet plazas.
Even here, however, Orange ceramics dominated the assemblages. The two Early Woodland radiocarbon
ages likely reflect Woodland intrusions into otherwise Archaic contexts. As with most rings, chipped
stone artifacts are few, save for debitage, while bone pin fragments are relatively common. Whelk tools of
a wide variety of types (punches, gouges, cutting-edge tools, and hammers) were relatively abundant in
the ring, though not in the plazas (Saunders 2003:23).
Rollins is managed as a part of the Talbot Islands State Park. It also lies within the boundaries of the Na-
tional Park Service’s Timucuan Ecological and Historic Preserve. It may have once been connected by a
lengthy causeway or ridge of shell to a large mound at a considerable distance from the ring, both of
which have subsequently been destroyed (Russo et al. 1993). Today, a central ring with attached ringlets
is a settlement plan found elsewhere at ring sites only at Fig Island 1. At Fig Island 1, however, the main
ring’s plaza is the smallest known (30 m), while the ring itself is the tallest (6 m). At Rollins, the plaza is,
horizontally, among the largest (135 m), and the ring walls not nearly as tall (3 m). The origins of these
differences could provide insight into the complex social structures that guided the construction of the
rings. As such, Rollins is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 50 m
Surface Topography
(after Kennedy 1980)
contour intervals irregular
Shape U or C
Diam. x Ht. (m) 250 x 1.7
Area Excavated (m2) 4
Radiocarbon Dates 2850+/-130; 3280+/-60; 3280+/-80; 3340+/-60; 3425+/-75; 3455+/-80
Stone 4 flake, 313 limestone, 3 sandstone
Bone 2 pin fragments
Archaic Ceramics 44 spiculate, 19 sand-tempered
Other 0
References Fryman et al. 1980; Russo 2004; Russo and Heide 2000, 2002, 2004
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The current configuration of the Reed Shell Ring is best described as a modified C. The northern arm, in
particular, is too short relative to the width of the ring to be classified as a U-shape. As the ring abuts the
Atlantic, there is good reason to believe that this arm and the longer southern arm have been shortened by
storm and beach erosion (Fryman et al. 1980; Russo and Heide 2000). Its current length has been meas-
ured as if it is a C-shape, the distance from the end of the northern arm to the end of the southern arm, i.e.
a half circle. But if it was a U-shape originally, its distance may be slightly shorter, based on a measure
from the closed end to the end of the southern arm. Kennedy (1966) produced a surface contour map of
the site. But its contour intervals are not consistent, or even readable in some cases. The midden’s height
listed in the summary table was based on excavation profiles (Russo and Heide 2000), not the height indi-
cated in Kennedy’s map. Although not particularly high, the ring is, horizontally, among the largest (cf.
Bonita Bay and Rollins). Its wall has been breached in a number of places by storms and possibly modern
drainage projects.
In the initial survey of the site, a single shovel test placed in the ring was aborted when a compacted layer
of shell was encountered (Fryman et al. 1980:40). A single St. Johns sherd was recovered from that test.
Two other shovel tests and probes in the plaza revealed it to be absent of midden or artifacts. Following a
literature search of the site, Carr et al. (1995:54–55) noted that St. Johns sherds had been recovered from
the ring. Catalog records at Florida Atlantic University identified these as five St. Johns sherds that were
surface collected and a human skull of unknown provenience. In a 1999 letter to Russo, William J. Ken-
nedy wrote that he collected the sherds while mapping the site in 1966.
In 1999, three 1-by-1-meter units were successfully placed in the ring, reaching the bottom-most ring de-
posits. A fourth unit in the plaza area revealed only storm tossed beach sand and debris (Russo and Heide
2000). The numerous limestone fragments may represent in situ deposits rather than manuports. Other
than ceramics, only four chipped flakes and 2 bone pin fragments were definitively identified. Shell tools
are usually abundant at rings along the Florida Gulf coast (e.g., Horr’s Island); the relative paucity of shell
tools at the Reed ring is more typical of Atlantic coast shell rings.
The Reed Shell Ring is unusual in yet other ways. Its radiocarbon ages makes it the last Late Archaic
shell ring ever built. Only Lighthouse Point has comparably late radiocarbon assays (Russo and Heide
2000:9). The southernmost Atlantic-coast ring, Reed stands in geographical and cultural isolation, lying
approximately 170 miles from its nearest ring, Bonita Bay on the Gulf of Mexico, and 250 miles from the
Guana Shell Ring on the Atlantic. In contrast, all other shell rings are found within a few miles of one
another, and never more than 25 miles distant (e.g., Meig’s Pasture and Buck Island).
Reed’s cultural isolation, however, is more than geographical, as reflected in its pottery. Its ceramics are
unique in that they represent two series, Glades sand-tempered and St. Johns spicule-tempered wares, nei-
ther of which shows up elsewhere in Florida for another 1,000 years. The Reed Shell Ring seems to be
among, if not the first site in south Florida where the adoption of pottery occurs. In this sense, the site is
like many other shell rings where regional pottery types, such as Thom’s Creek, Stallings, and Orange
wares, mark their initial or early appearances. The difference at Reed is that, to date, no other Late Ar-
chaic sites in the region have yielded these pottery wares. In fact, no other shell-ring site of any type is
known to reflect Reed’s age and ceramic traditions.
Because Reed Shell Ring has the earliest pottery in the region, is a founding site on the coast for the pe-
riod, is one of the largest rings, is the last Archaic shell ring ever built, and remains largely intact despite
probable erosion on its eastern side, it represents an invaluable resource for understanding the rise, func-
tion, and social use of shell rings. It qualifies as a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1.0
0.1
0 40 m
Shell Thickness
contour interval = 0.1m
Shape U
Diam. x Ht. (m) 170 x 1.3
Area Excavated
63.5
(m2)
2740+/-70; 2880+/-70; 3490+/-60; 3590+/-70; 3600+/-50; 3620+/-70; 3720+/-60;
Radiocarbon Dates
3740+/-70; 3820+/-70; 3860+/-60
Stone 9 steatite, 2 PP fragments
Bone 8 pin fragments, 2 bead
Archaic Ceramics >1,205 Orange
Other 21 whelk
References Russo et al. 2002; Saunders and Rolland 2006
This shell midden was first recognized by Tesar and Baker (1985) as a possible Late Archaic ring-shaped
shell midden based on an Orange sherd found at a tree fall and, apparently, a semi-circular shape of shell
on the surface. Newman and Weisman (1992) suggested the site was similar to South Carolina and Geor-
gia shell rings. Investigation of the subsurface aspects of the site occurred a decade later (Russo et al.
2002). The investigators conducted an intensive, close-interval testing and mapping program across the
entire site producing precision maps of the surface topography and shell distributions. In the field, the
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
shape of the ring could not be easily discerned in a single view due to its large size and heavy vegetation.
But the shell distribution map clearly showed a U-shaped ring with a central plaza.
Russo et al. (2002) systematically placed 50-by-50-centimeter shovel tests across the plaza and exterior of
the ring (n = 130), and ring edges (n = 39) at 10-meter intervals. Results from these tests have not yet
been written up. In addition, thirteen shovel tests were placed at 30-meter intervals around the apex of the
ring itself, and a 1-by-2-meter unit at the closed end of the ring (Russo et al. 2002:14). Saunders and Rol-
land (2006) placed four 2-by-2-meter units in the plaza area adjacent to the southwest ring wall. They
based this placement on shovel test data (Russo et al. 2002), which revealed an unusual amount of shell
for plaza otherwise free of dense shell deposits.
Russo et al. (2002:15, 29) recovered 1,003 fiber-tempered sherds, of which only 156 were sufficiently
large or well-preserved to be identified as decorated or plain Orange types. Saunders and Rolland
(2006:49–51) recovered 2,705 sherds, of which only 249 could be typed. Of these, 202 were of Orange
types. This should not be interpreted as meaning that only 7 percent of the total sherds were Orange.
Rather, their analysis was geared to distinguishing plain from decorated Orange sherds to get comparative
ratios; they chose not to attempt to identify sherds smaller than 3 centimeters in size. Thus, most of the
sherds were not reported as to type, but most were fiber-tempered. They also recovered four whelk tools,
three pieces of steatite, two bone pin fragments, and a bone bead from Orange contexts (Saunders and
Rolland 2006:54). (Sixteen expedient bone tools are not included in the summary table.) Non-ceramic
artifacts have not been reported from the Russo et al. (2002) survey. But preliminary data sheets indicate
five steatite, two projectile point fragments, fifteen lithic flakes, six bone pin fragments, a bone bead, and
seventeen whelk tools.
Guana Shell Ring lies within the boundaries of the Guana River State Wildlife Management Area and, as
such, is preserved and protected by the state of Florida. A dirt road, used by maintenance personnel and
rangers, runs through the site. Day hikers use the road as a trail. The low-lying site is well preserved, but
is not observable from the road except at the point where it crosses the closed end of the ring. With its
opening to the southeast, it is similar in shape (sans the ringlets) and orientation to the Rollins Ring, its
closest ring neighbor. With their proximity and similar artifacts and radiocarbon ages, the two sites seem
to reflect a common tradition of shell-ring construction and culture. Together or separately, they are po-
tential NHLs.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 40 m
Surface Topography
(after Bullen and Bullen 1976:5)
contour interval = 0.2 m
Shape U
Diam. x Ht. (m) 140 x 4.2
Area Excavated (m2) >36
Radiocarbon Dates n/a
Stone 7 balls, 1 limestone metate, 4 fossilized manatee rib mano
Bone 23 pin/awl, 2 shark bead
Archaic Ceramics 37 Orange/fiber-tempered
Other 847 whelk/conch, 3 shell bead, 18 perforated shell, 11 clam tools
References Bullen and Bullen 1976; Sarney 1994
Bullen and Bullen excavated one trench and five units, and twelve shovel tests. No artifacts were recov-
ered from the plaza. Prior to the construction of a sewage lift station in the ring, a 2-by-2-meter unit and a
1.5-by-0.5-meter unit were excavated (Sarney 1994:48). These units yielded 12 Orange/fiber-tempered
sherds, 4 bone pin fragments, a worked antler, 17 columella hammers, 3 gastropod pounders and 5 ham-
mers, 2 whelk dippers, a whelk adze, 4 perforated ark shells, 534 pieces of debitage from the manufacture
of whelk and conch tools, 1 limestone metate, 1 fossilized manatee rib mano, as well as other artifacts.
The summary table includes these artifacts, as well as those enumerated by the Bullens (1976:7–18).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Although Bullen and Stoltman (1972:50 [see Sarney 1994:19]) apparently suggested that the shape of the
Hill Cottage midden implied a relation to the Georgia and South Carolina Archaic shell rings, it was Sar-
ney (1994) that first described in detail the commonalities this site had with other known shell rings. Its
U-shape is typical of most other shell rings in Florida (Guana, Bonita, Horr’s Island, Reed, and Rollins).
It appears to have been constructed during both the preceramic period and the initial period when fiber-
tempered ceramics were introduced in the region. As such, it holds the promise of providing valuable in-
formation on the development of pottery, changes in social formations, and connections to other regional
cultural traditions. Hill Cottage is currently being held for the public’s benefit by a private concern dedi-
cated to preserving this and other historic resources on the managed property. It is part of a larger site
(8So2) listed on the NRHP. Its shape is preserved and its integrity is high, despite paths, buildings, and
excavation units still present. It is a potential NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
G. Geographical Data
The broadest geographical extent of known shell rings stretches along the coast from mid-South Carolina
to Mississippi. Because the cultural affiliations of the northernmost rings in South Carolina and the west-
ernmost rings in Mississippi are of groups known to extend farther north and west respectively, it should
be expected that other shell ring might be found beyond the known boundaries, from the northern border
of South Carolina to Louisiana. Shell rings have been found on barrier and other coastal islands, along
ocean shorelines, on and beneath marshes and mangroves, and along freshwater rivers entering into estu-
aries. Dependent on coastal shellfish for their construction, shell rings are perforce limited to the coastal
zone. Archeological sites in interior riverine areas have been identified as constructed from freshwater
shellfish remains in circular to semi-circular patterns (e.g., Phillips 1970:266, 270; Rouse 1951:128).
These sites fall outside and are not elements of the geographic boundaries of the Late Archaic shell-ring
building traditions.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Currently there are no Late Archaic shell rings that have been granted National Historic Landmark status.
One, the Long Field Crescent, was listed as a contributing element of the St. Catherine’s Island National
Historic Landmark in 1969, although it was not recognized at that time as a Late Archaic shell ring. In
addition to this site, twelve more Late Archaic shell rings are listed on the National Register of Historic
Places (Table 9). Eleven are in South Carolina. Nine of these were listed in 1970 under the theme “The
Original Inhabitants.” Spanish Mount was listed in 1974, but as a shell midden, not a ring. Later it was
suggested to be the remnant of a shell ring (Cable 1993). The NRHP nomination forms for the listed rings
served as primary sources of information presented here (specifically the Hanckel, Horse Island, Buzzard
Island, and Auld shell rings).
The Hill Cottage shell ring is identified by the state site number 8So2, which is an historic site consisting
of a number of standing structures, precontact shell middens and a burial mound. The larger site, 8So2 is
listed on the National Register (Matthews et al. 1975) with the Hill Cottage shell ring included within the
boundaries, but not specifically identified as a Late Archaic shell ring.
Lighthouse Point was nominated in 1990 (Lawrence 1990a) as part of the National Register of Historic
Places Multiple Property Documentation Form entitled “Late Archaic-Early Woodland Period Shell
Rings of South Carolina, ca. 1000–2200 B.C.” (Lawrence 1990b). The Keeper of the National Register
signed the document that same year, and at this time Lighthouse Point was listed. This multiple listing
document also included Guerard Point, Barrows, Patent, and Bull Island as rings that required more map-
ping and study to confirm their eligibility for listing on the NRHP. It was suggested that Guerard Point
probably lacked integrity of form to be listed. Broad River and Medicine were also listed as potentially
eligible. These last two sites have been mapped and tested by this historic context and determined not to
be Late Archaic shell rings.
Shell rings are circular to semi-circular deposits of marine shell found along the coasts of South Carolina,
Georgia and Florida. (Two rings were once located near the mouth of the Pearl River near the Mississippi
Gulf coast. However, these have been destroyed.) The shape of shell rings has fascinated archeologists
and other chroniclers since the 1800s, but relatively few rings have been studied in any detail. Over the
last two centuries four stages of investigative methodologies have characterized ring studies. Examination
of the literature produced from these periods of investigation has provided the documentation employed
to identify and evaluate shell rings’ potential for listing as NHLs under this historic context.
Stage 1, 1802–1925
Chroniclers and antiquarians did little more scientific inquiry than to estimate the rings’ shapes and sizes.
Their possible functions were speculated. The conjectures included the use of the rings as disposal areas,
villages, gaming arenas, political centers, storm refuges, and forts (e.g., Bragg 1925; Drayton 1802;
McKinley 1873; Moore 1897). Few excavations were ever conducted (cf. Moore 1897). Six ring sites in
Georgia and South Carolina were recognized at this time: Sapelo, Guerard, Stratton Place, Lighthouse
Point, Fig Island, and Buzzards Island. McKinley’s 1873 observations provided some of the only infor-
mation on the sizes of Sapelo rings 2 and 3 for years to come.
Stage 2, 1932–1975
The first scientific investigations and preservation efforts on shell rings began during this period. The
studies of this stage are typified by trenches cut across ring walls. While a large number of shell rings
were identified during this era, the period is also equally notable for its under-reporting and limited analy-
ses of the rings.
Large-scale excavations were conducted at the Chester Field shell ring in 1932 and 1933 although the
results of these excavations were belatedly and never fully reported (Flannery 1943; Ritter 1933). In
1950, a trench was excavated through a wall of the large ring on Sapelo Island. Again, those results were
not reported for many years after the fact (Waring and Larson 1968). By the 1960s small excavations had
been conducted at Skull Creek, Sea Pines, and Sewee (Calmes 1967; Edwards 1965). These reports or
summaries of the findings were never published, and notes, maps, and photographs have since been lost.
In 1970, the South Carolina Institute of Archaeology and Anthropology (SCIAA) conducted a survey of
shell rings in South Carolina. In all, they identified 18 shell rings, nine of which were nominated and
listed on the NRHP (Table 9)(Hemmings 1970f:9). Unfortunately, no full report on these findings was
ever published, and it remains unclear as to exactly which shell rings were surveyed (cf. Hemmings
1970f), or why only 9 were nominated. A trench was placed across the east wall of the Fig Island 2 repre-
sent the largest excavation conducted during the survey.
These small excavations and trenches across ring walls were conducted to recover artifacts and datable
materials in order to determine the cultural affiliation of shell rings. The development of cultural chro-
nologies and identification of cultural regions was a typical subject of study at the time, and ceramics
from shell rings in Georgia and South Carolina were often used to establish temporal and cultural linkages
(e.g., Caldwell 1952; Waring 1968b; Griffin 1943). Most rings in South Carolina were determined to have
been made by Thoms Creek pottery-producing peoples (Trinkley 1980b), while those in Georgia were
identified as being made by Stallings/St. Simons wares producing peoples (DePratter 1975).
Notes on the internal structures of the rings were made, and the insight that rings were constructed from
piles of shell intermixed with “kitchen midden” was established (e.g., Calmes 1967; Edwards 1965; War-
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
ing and Larson 1968). Speculation not only as to the function of rings continued (ceremonial versus habi-
tation were the dominant discussion themes), but also as to how rings were constructed (accretion versus
intentional piling). The first (and only to that point) contour maps of shell rings were made of the Sewee
ring (Edwards 1965), Fig Island 1 and 2 (Hemmings 1970f) and the Reed shell ring in Florida, although
the latter was not recognized as a shell ring at the time (Kennedy 1980[1966]). However, Hemmings did
provide numerous sketch maps of the rings he had surveyed, which today remain the best ideas we have
on of some of these rings’ shapes and sizes.
While scientific investigations on shell rings were somewhat limited in scope and reporting, a large num-
ber of shell rings were identified during this era. Seven ring sites in South Carolina were first reported
during this period: Sea Pines, Skull Creek, Chester Field, Hanckel, Horse Island, Auld, and Sewee.
Twelve ring and possible ring sites in Georgia were reported: five on Skidaway Island (Large 9, Small 9,
21, Odingsell, and Skidaway), two on Wilmington Island (Oemler, Walthour); three on St. Simons Island
(Bony Hammock, Cannon’s Point, and West), and Busch Krick and Barbour Island. By the mid 1970s, all
shell rings in Georgia and South Carolina known today were recognized except for the Coosaw rings and
Fig Island 1.
Stage 3, 1975–1999
In 1975, Trinkley reported on excavations he had conducted at the Lighthouse Point shell ring. This report
represented the most extensive excavation of a shell ring up to that time, and, more importantly, the re-
sults were presented to academia and the greater public in an accessible journal (Trinkley 1975). Addi-
tional work at this site and the Stratton Place shell ring was performed 1979, the detailed results of which
were presented in a Ph.D. dissertation (Trinkley 1980a). Combined with Marrinan’s (1975) dissertation
on her extensive testing of the Cannon’s Point and the less extensively tested West shell ring in Georgia,
these two seminal works brought shell ring investigations into a new era where detailed analyses of bo-
tanical and faunal remains, soils, sea level and other aspects of the environment characterize the best of
the studies (DePratter 1975; Marrinan 1975; Trinkley 1980a; Trinkley and Ward 1978).
At the time, only one of the nine Archaic shell ring sites in Florida were thought to represent a shell ring
related to the tradition of the Georgia and South Carolina shell rings. At Hill Cottage, Sarney (1994) iden-
tified similarities in the “U” shape of the ring, the presence of fiber-tempered pottery, of shell and bone
tools, and radiocarbon dates. At the Reed shell ring, discussion of similarities to Georgia/South Carolina
shell rings were offered as one of a number of possibilities to account for the ring’s semi-circular shape
(Fryman et al. 1980). At Horr’s Island, the shape of the ring was clearly defined but not yet linked to the
Georgia/South Carolina traditions. Work at the site yielded extensive data on subsistence reported in a
Ph.D. dissertation (Russo 1991), on soils reported in a master’s thesis (Scudder 1993), on botanical re-
mains (Newsom 1991), and on environment and sea level changes (Russo 1991, 1994).
Subsistence and seasonality studies endeavored to determine the attraction the coastal environment held
for ring builders. Seasonality studies determined that rings were likely occupied for multiple seasons or
throughout the year (Marrinan 1975; Russo 1991; Trinkley 1980a). With few exceptions, only sketch
maps were produced of shell rings. Of the few surface contour maps that were made, none clearly out-
lined the parameters of the shell rings or specified the manner in which the maps were made so as to pro-
vide the reader an objective criteria for determining the maps’ relative accuracy (Marrinan 1975:26;
Russo 1994; Thomas and Campbell 1991:105; Trinkley 1980a). Determining where shell-ring sites began
and ended and where non-ring surface topographies blended into the sites on these maps was always
problematic.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Questions of the effect of sea level rise on shell rings are prominent during this period. Investigators note
that many shell rings are located in coastal marsh, are partially submerged beneath high tide waters, or are
eroded by wave action. Borrowing evidence from geological disciplines they suggest that these and all
shell rings were originally located on dry lands, but have been submerged since their abandonment by
risen seas. Some archeologists suggest that the rising of sea levels during the mid-Holocene, in fact,
flooded the coastal shores allowing the first formations of estuarine environments necessary for Archaic
settlements (DePratter 1976:115; Michie 1979:96; Widmer 1976:46, 1988). Others argue that not only
environmental, but social impetuses must be considered for explaining the wide-spread settlement and
increased social complexity evident at Late Archaic ring sites (Russo 1991; Trinkley 1980a).
Only touched upon during this period is the question of social organization and settlement patterns of the
shell-ring builders. A variety of theories emerge from several studies. Some archeologists proposed that
larger shell rings served as base camps in a regional pattern of settlement that included smaller shell-
bearing sites that functioned as temporary camps (Widmer 1976:46). Others contended that shell rings
served as special purpose meeting places (evidenced by a limited artifact inventory), and other large, non-
ring, shell-bearing sites functioned as the base camps (Michie 1979). And some interpreted shell rings as
permanent villages, either of simply organized egalitarian hunter-gatherers (Trinkley 1985), or more
complexly organized groups capable of building large-scale public works and monuments (Russo 1991).
These issues would extend into and be expanded upon in the next stage of shell-ring studies.
Even though extensive investigations occurred at a number of shell-ring sites, only three new rings (all in
Florida) were identified during this period: Guana (Tesar and Baker 1985), Buck Bayou (Thomas et al.
1991), and Meig’s Pasture (Curren et al. 1987). Only a brief note on the Guana ring suggested a possible
connection the Late Archaic shell-ring-building traditions. Buck Bayou was connected to the Poverty
Point trade network bringing the Mississippi Valley cultural tradition of arcuate settlement into the dis-
cussion of coastal shell formations through the similarities of shape in the Claiborne and Cedarland rings.
But no connections to the Atlantic coast shell-ring traditions were made with these circular deposits of
shell. Except for Guana, all investigations of these sites during this period yielded surface contour maps.
Stage 4, 1999–Present
What differentiates this period of shell-ring investigation from earlier ones is the attempts to determine
the boundaries and structure of shell rings through systematic shovel test surveys, systematic probing,
surface contour mapping, and geophysical remote sensing. During this time there is an increased descrip-
tion of the internal structures of shell rings as part of an effort to gain an understanding of the ring build-
ers’ methods of construction and reasons for building rings. Through comparative analyses with other
rings, ethnographic analogies, and the efficacy of different theoretical understandings, rings become iden-
tified as quite diverse in terms of shape, size, function, cultural affinity, and the social organization of
their makers. Rings are seen not as changeless settlement components, but as vital settlement features
whose particular uses differed among rings and changed through time.
During this brief seven year period, four of Florida’s shell-ring sites (Rollins, Oxeye, Guana, and Bonita
Bay) and ten additional rings in South Carolina (Coosaw 1–4; Fig Island 1–3; Sewee; Barrows; Patent)
were mapped, re-mapped, or formally recognized as shell-ring sites for the first time. While Barrows,
Patent, and Fig Island were previously identified, their confirmation as Archaic shell rings only recently
occurred (Saunders et al. 2006; Saunders and Russo 2002). Similarly, all Florida shell rings, except Ox-
eye, had been identified as archeological sites earlier, but their identification as shell rings did not occur
until they were mapped; radiocarbon dated, and compared to Georgia and South Carolina ring formations
(Russo 2004). In addition to these, the “lost” rings of Sapelo, rings 2 and 3 (Larson 1998:30) were redis-
covered using geophysical survey techniques (Thompson 2006).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Today, up to 45 ring sites with up to 51 individual rings (not counting conjoined rings or mounds) are
recognized in South Carolina, Georgia, Florida and Mississippi. And more are being found in areas previ-
ously unknown to contain them (DePratter 2005; Thomas 2006).
All historic reports and accounts of shell rings were examined for this historic context. Lawrence’s
(1989a-c, 1991a, b) compendia fairly well covers the documentation for Georgia and South Carolina shell
rings prior to 1991. In addition nomination forms for the NRHP-listed shell rings were consulted. Site
reports, chapters, and articles for the Florida shell rings were consulted. Some rings had little or no pub-
lished documentation (e.g., Oxeye, Long Field Crescent) and field notes, catalogue records, and personal
communications provided most of the information on these sites. For this historic context, ongoing work
for two shell rings (Barrows and Patent) and two sites (Broad River and Medicine) considered previously
to be shell rings (Lawrence 1990b) were discussed for the first time. State Historic Preservation and other
government offices provided site forms and other unpublished information. The major references used in
conducting research on this historic context are listed in Section I.
Sites were chosen for consideration and discussion if in the reviewed literature they had been referred to
as shell rings dating between 5000 and 3000 B.P. This is the Late Archaic period in the southeast U.S. In
South Carolina, some archeologists refer to the latter part of this period as the Early Woodland based on
non-temporal criteria (e.g., the presence of pottery) which have been suggested to be a Woodland cultural
manifestation (e.g., Trinkley 1980a; Lawrence 1990b). For this historic context, we eschewed the distinc-
tion, not because it is without merit, but because other states with shell rings do not generally follow that
usage. In addition, other states have Early and other Woodland ring middens that may or may not contain
shell (Russo et al. 2006), but that are not of the same tradition or character as the Archaic shell rings.
With this in mind, we concluded that the use of the term Woodland to describe the historic contexts of the
shell rings in this context would only provide confusion with Refuge, Deptford, and Swift Creek ring
middens (Bense 1998; Milanich 1971; Russo et al. 2006).
Sites of the appropriate age, shape (arcuate to circular), internal structure (i.e., made of shell) and high
integrity were considered potentially eligible for listing as an NHL. Their descriptions are found in Sec-
tion F, as are details of the evaluative criteria.
In some cases, authors referred to sites as shell rings incorrectly. The sites were not shell rings, but rather
were shell middens without a ring shape. Other references identified sites as shell rings mistakenly listing
the wrong site numbers. Some sites provisionally called rings by authors were, upon inspection, shown
not to be Late Archaic shell rings. These and other sites were included in the descriptions of rings for con-
sideration as potential NHL shell rings under this historic context. Those that did not meet the criteria for
listing as NHLs are described in this section as “rejected,” with the reasons presented for exclusion from
further NHL consideration. Some of these sites, upon further study, may be eligible for nomination as
NHLs. A summary of the NRHP/NHL potential for all sites considered for NHL status is found in Table
10. A summary of ring sites identified in this historic context as potential NHLs is found in Table 11.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
In Lawrence’s (1990b) NRHP Multiple Property listing for South Carolina shell rings he states that to be
eligible for the NRHP a shell ring must past tests of original geometry, age, and integrity. In terms of in-
tegrity he suggests that a ring must have at least 45 cm (18") of mounded shell at its base. This would pre-
sent problems for some ring sites such as Coosaw 3 and Sapelo 3, which have been leveled but maintain
their original geometry below ground in non-mounded shell midden deposits. Both these sites have been
shown to have invaluable information below ground, although they are not recognizable as rings above
ground (Heide and Russo 2003; Thompson 2006). Therefore, in this historic context to be eligible for the
NRHP, a shell ring must minimally contain important information under Criterion D, regardless if any
above ground ring remains intact or not. To be considered eligible as a NHL, however, a shell-ring site
must maintain integrity of its original geometry above ground. Shell rings are architectural forms, and
their integrity is linked not only to in situ shell deposits above and below ground useful for providing in-
formation for scientific studies, but to the integrity in form in terms of shape, height, and horizontal dis-
tributions of shell sufficient to ascertain their original architectural design.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Table 11. NRHP and NHL Status of Sites Identified as Shell Rings by Investigators
SOUTH CAROLINA
Sea Pines 38BU7 listed 1970 yes well preserved and protected
Skull Creek, Large 38BU8 listed 1970 yes mined; more study needed
Skull Creek, Small 38BU8 listed 1970 yes mined; more study needed
Daws Island 38BU9 not eligible no eroded; not a Late Archaic ring
Guerard Point 38BU21 not evaluated no mined, subsurface intact
Chester Field 38BU29 listed 1970 no erosion and development
Barrows 38BU300 not evaluated yes well preserved and protected
Patent Point 38BU301 not evaluated yes well preserved and protected
Broad River 38BU302 not evaluated no not a Late Archaic ring
Medicine 38BU303 not evaluated no not a Late Archaic ring
Bull Island 38BU475 not evaluated ? more study needed
Coosaw 1 38BU1866 eligible yes eroded, but largely intact
Coosaw 2 38BU1866 eligible yes well preserved and protected
Coosaw 3 38BU1866 eligible yes as MP mined, subsurface intact
Coosaw 4 38BU1866 not evaluated yes more study needed
Hanckel Mound 38CH7 listed 1970 ? more study needed
Lighthouse Point 38CH12 listed 1990 no architectural integrity lacking
Horse Island 38CH14 listed 1970 ? more study needed
Buzzards Island 38CH23 listed 1970 ? more study needed
Stratton Place 38CH24 eligible ? mined, current condition unknown
Auld 38CH41 listed 1970 ? more study needed
Fig Island 1 38CH42 listed 1970 yes well preserved and protected
Fig Island 2 38CH42 listed 1970 yes well preserved and protected
Fig Island 3 38CH42 listed 1970 yes well preserved and protected
Sewee 38CH45 listed 1970 yes mined, but largely intact
Crow Island 38CH60 not evaluated ? more study needed
Spanish Mount 38CH62 listed 1974 no if a ring, too disturbed for NHL
Hobcaw 38CH? Not evaluated ? more study needed
GEORGIA
Oemler 9CH14A not evaluated ? more study needed
Walthour 9CH16 not eligible no not a Late Archaic ring
Cane Patch 9CH35 not evaluated ? more study needed
Skidaway 9, Large 9CH63 not evaluated ? more study needed
Skidaway 9, Small 9CH63 not evaluated ? more study needed
Skidaway 21 9CH75 not evaluated ? more study needed
Skidaway 9CH77 not evaluated ? more study needed
Odingsell 9CH111 not evaluated no more study needed
Ossabaw 77 9CH203 not evaluated ? more study needed
Bony Hammock 9GN53 not evaluated ? more study needed
Goodyear Mound 9GN54 not evaluated no not a Late Archaic ring
Cannon’s Point 9GN57 eligible? Yes well preserved in 1975
West Ring 9GN76 eligible? Yes well preserved in 1975
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Table 11. NRHP and NHL Status of Sites Identified as Shell Rings by Investigators (cont.)
GEORGIA (CONT.)
Long Field Crescent 9LI231 listed 1969 ? impacted, more study needed
Sapelo I 9MC65 eligible yes well preserved and protected
Sapelo 2 9 MC65 not evaluated yes as MP mined, subsurface intact
Sapelo 3 9 MC65 not evaluated yes as MP mined, subsurface intact
Busch Krick 9 MC87 not evaluated no too eroded for NHL
Barbour Island 9 MC320 not evaluated ? more study needed
FLORIDA
Horrs Island Md. B 8CR206 not eligible no destroyed
Horrs Island Md. C 8CR207 not eligible no destroyed
Horrs Island Md. A 8CR208 eligible yes well preserved
Horrs Island 8CR209 not eligible no mostly destroyed
Horrs Island Md. D 8CR211 eligible yes well preserved
Oxeye 8DU7478 eligible yes well preserved, but drowned
Rollins 8DU7510 eligible yes well preserved and protected
Bonita Bay Md. 8LL716 eligible ? well preserved, more study needed
Bonita Bay Ring 8LL717 eligible ? developed, but largely intact?
Reed 8MT13 eligible yes eroded, but largely intact
Meig's Pasture 8OK102 unknown no destroyed?
Guana 8SJ2554 eligible yes well preserved and protected
Hill Cottage 8SO2 listed 1975 yes developed, but largely intact
Buck Bayou 8WL90 unknown ? more study needed
MISSISSIPPI
Cedarland 22HC30 not eligible no destroyed
Claiborne 22HC35 not eligible no destroyed
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
In describing the baked clay objects he found at Charles Towne, South (1969) mentioned that baked clay
objects were also present at Daws Island (38Bu9). He identified the Daws site as “a small shell mound or
ring fragment” (South 1969:24). Sarney (1994:147–170) interpreted this to mean the site was a shell ring.
She discussed it at length relative to artifactual commonalities with other shell rings, including the pres-
ence of “fragmentary human remains.” She believed that the remains were indicative of shell rings be-
cause they occurred at half the ten ring sites she reviewed (Sarney 1994:169).
Although the site has Late Archaic artifacts and shell midden, it has not been otherwise described as a
shell ring (Michie n.d.). It is included in this listing for the sake of completeness. The site is not a shell
ring, and may no longer exist in any form.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Knowledge of Guerard Point initially came from a single paragraph written by Moore (1897:151), who
stated that the site was “of the same class as that on Bull Island, S.C., and the great one on Sapelo Island.”
He identified the ring as “roughly circular” and noted that it represented the only shell midden in the area.
At the time it had been greatly lowered by plowing. In 2002, William Green of the South Carolina Ar-
chives and History Department, Rebecca Saunders of Louisiana State University, and Mike Russo of the
National Park Service found that the site had been completely leveled and lay under sod as part of a sod
farming operation. Green had visited the site earlier and identified Archaic ceramics and oyster shell after
the land had been leveled. In 2005, Gantt et al. (2005) reported on investigations at the site and the sur-
rounding area. In a survey employing 30-meter-grid shovel tests, they placed six tests in the site area, four
of which were positive for shell and artifacts. In their sketch, Gantt et al. (2005:63) demarcated 38Bu21 as
an 80-by-45-meter elliptical area, although their text suggests it measures 80 by 60 meters. It is unclear if
this is meant to indicate the size and the shape of the shell ring’s remnants. Unfortunately, Gantt et al.
(2005) combined their description of artifacts from the site with that of a historic site surrounding it. Al-
though three Stallings sherds were recorded among other precontact and historic artifacts, it is unclear if
they came from the shell ring (Gantt et al. 2005:63). The depth of the deepest deposits was given as 65
centimeters. But the ring once stood much taller than the extant shell deposits indicate.
A 1-by-3-meter unit was placed in the ring location (Gantt and Styer 2006:52). From it came ninety-nine
Stallings sherds. In addition, fifty-six eroded, residual sand-tempered sherds were recovered. It seems odd
that all the residuals were sand-tempered when all the other sherds were fiber-tempered. Whether the
sand-tempered sherds are residual Thoms Creek sherds or other wares is unclear (Gantt and Styer
2006:57). The unit showed that beneath the shell ring, the strata were characterized by sand and shell,
with observable shell filled pits and posts.
Further testing is needed to confirm the shape and size of the ring. The remaining undisturbed portions of
the site likely represent features originally found below the ring rather than remnants of the above ground
ring itself. From these initial stages of shell-ring construction, archeologists have been quite successful in
gleaning valuable information on the shape and character of plaza and sub-ring activities (Thompson
2006; Trinkley 1985), as well as size and shape of the former overlying ring. As such, Guerard Point ring
is potentially eligible for listing in the NRHP under Criterion D. It lacks architectural integrity to qualify
as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 10 m
Measured Drawing
(after Ritter 1933)
Shape C
Diam. x Ht. (m) 54 x 1.8
Area Excavated (m2) 77
Radiocarbon Dates n/a
Stone 2 to 5 PP?
Bone 7 pins/awls
177 Thoms Creek, 63 fiber-tempered (Trinkley 1980a:283), 158 fiber-tempered (Griffin
Archaic Ceramics
1943)
Other # antler
References Flannery 1943; Ritter 1933
The C-shape may be wholly or partly the result of erosion. The site is situated on the bank of the Broad
River, which in 1991 was observed to be eroding the western side of the site (Judge and Smith 1991:37).
It must be stated, however, that there is no record of the site having ever been a closed circle.
The site is important in the history of shell-ring studies because it was the first to be intensively investi-
gated and reported, albeit posthumously (Anderson 1977:377). Warren K. Moorehead excavated the site
in 1932 (Flannery 1943), while Ritter (1933) mapped it. It is difficult to determine the exact sizes and
how many units were placed based solely on Ritter’s maps and Flannery’s descriptions. But a surprisingly
large amount of the site was excavated both within the ring and within the plaza. Flannery (1943:150)
noted that seven “test holes” were placed in the plaza. However, two more “test holes” were found on the
Ritter map. One was 10 by10 feet (3 x 3 m), but the others seemed to be less than 3 feet square (0.9 m2)
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
each. Within the ring a 25-by-10-foot (7.62 x 3 m) area was excavated along with two units that were
maybe 3 by 2 feet (0.9 x 0.6 m) square and 2 by 2 feet (0.6 x 0.6 m) square. Four other excavations fea-
tures were also shown on the map to have been placed in the ring. These include two trenches (Trench A
and B), which intercept to form an irregular L-shape. Trench A is about 3 by 33 feet (1 x 9.1 m), and
Trench B about 6 by 30 feet (1.8 x 9.1 m) at its longest points. In addition, two units (Pit 2 and Pit 4)
measure approximately 10 by 5 feet (3 x 1.5 m) and 15 by 10 feet (4.6 x 3 m).
The site was among the first in which pottery was analyzed with type names still in use. Comparing his
more recent analysis of Chester Field ceramics to that of the earliest study (Griffin 1943), Trinkley
(1980a:283) points out that Thoms Creek and Stallings tempers and designs can overlap, and conflation
of types may result from biases and experience attendant with the identifier. His analysis of Chester Field
ceramics yielded 74 percent Thoms Creek and 26 percent Stallings pottery, while Griffin’s (1943) yielded
100 percent Stallings wares. Chester Field, long thought to be a classic Stallings shell ring, may actually
be a Thom’s Creek ring according to Trinkley’s analysis.
Unpublished notes in the Charleston Museum from the Moorehead and Ritter investigations identified
two to five lithic points (i.e., “spear heads”), an unspecified number of chipped flakes, a hammerstone, at
least seven fragments or whole decorated and undecorated bone pins and awls, over one thousand untyped
potsherds, and an unspecified number of worked antler prongs.
Chesterfield is currently listed on the NRHP. The degree of erosion and impact from development (Judge
and Smith 1991:38) likely excludes it from consideration for listing as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 15 m
Shell Thickness
contour interval = 5 cm
Shape semi-circle?
Diam. x Ht. (m) 100 x 0.3
Area Excavated (m2) 1
Radiocarbon Dates 1240+/-50
Stone 0
Bone 0
Archaic Ceramics 0
Other 0
References Lawrence 1991a:64–66
Two sites (38Bu302 and 38Bu303) near Patent and Barrows shell rings have recently been shown to be
neither Archaic nor shell rings based on: radiocarbon dates, the absence of Archaic artifacts, and the lack
of a contiguous circular to semi-circular shape to their shell deposits. These are mentioned here for clari-
fication of Lawrence’s (1991:65) suggestion that 38Bu302 might be a shell ring. Detailed surface and
shell distribution, as identified through probing, indicated that the site represents a scatter of small depos-
its of shell (mostly oyster and clam) over an area roughly 100 by 30 meters. Excluding the southernmost
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
deposit, the shell distribution does present a series of deposits in a roughly circular, non-contiguous pat-
tern 30 by 40 meters across. A 1-by-1-meter unit was dug through dense shell midden in the northernmost
section of the site to sterile soil at 30 centimeters below surface. No artifacts were recovered, but one ra-
diocarbon age on oyster shell from the bottom strata was obtained, indicating occupation some 2,000
years after the nearby Patent shell ring was abandoned. The site is not an Archaic shell ring. It may be
eligible for listing in the NRHP, but not as a shell ring. It has not been evaluated for the NRHP.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 5m
Shell Thickness
contour interval = 5 cm
Shape arc?
Diam. x Ht. (m) 17 x 0.3
Area Excavated (m2) 0.25
Radiocarbon Dates 1280+/-50
Stone 0
Bone 0
Archaic Ceramics 0
Other 0
References Lawrence 1991a:67–69
Two sites (38Bu302 and 38Bu303) near Patent and Barrows shell rings have recently been shown to be
neither Archaic nor shell rings based on: radiocarbon dates, the absence of Archaic artifacts, and the lack
of a contiguous circular to semi-circular shape to their shell deposits. These are mentioned here for clari-
fication of Lawrence’s (1991:65) suggestion that 38Bu303 might be a shell ring. Detailed surface and
shell distribution, as identified through probing, indicated that the site represents a small arc of shell
(mostly oyster and clam) with an opposing shell pile lying to its south. Like many Archaic shell rings, it
appears as an arc with a sterile plaza, but is much smaller at 17 meters across. A 1-by-1-meter unit was
dug through dense shell midden in the northernmost section of the site to sterile soil at 30 centimeters be-
low surface. No artifacts were recovered, but one radiocarbon age on oyster shell from the bottom strata
was obtained, indicating occupation some 2,000 years after the nearby Patent shell ring was abandoned.
The site is not an Archaic shell ring. It may be eligible for listing in the NRHP, but not as a shell ring. It
has not been evaluated for the NRHP.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 40 m
Sketch
(after Lawrence 1991:80)
Two mentions by Moore (1898:147, 151) that Bull Island contains a shell enclosure just like that found
on Sapelo suggest that an Archaic shell ring may be located here. However, since other time periods also
produced shell enclosures, we cannot be sure this one dates to the Late Archaic. Lawrence (1991a:77–80)
identified the site as a shell ring, however, did not identify any cultural affiliation. A sketch map shows
the site as a series of four arcs of shell (described in text as oyster) forming a circle 91 meters in diameter
with 10- to 15-meter gaps between each arc. The text describes the geometry of the ring as elliptical.
Since an ellipse can be circular, both the map and text seem to be in agreement. The map is referenced by
Lawrence to Aichele 1981, although access to the reference could not be obtained. No excavations have
taken place at this site, which needs more study before it can definitively be classified as an Archaic shell
ring. As such, it remains potentially eligible for listing in the NRHP and as an NHL, but has not been
evaluated.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
eroded
eroded
0 15 m
Sketch
(after Lawrence 1989b:cover)
Shape C
Diam. x Ht. (m) 62 x 2.4
Area Excavated (m2) 0
Radiocarbon Dates 3550+/-60
Stone 0
Bone 0
Archaic Ceramics 48 Thoms Creek
Other 0
References Hemmings 1970b
The 1970 NRHP nomination form on the Hanckel site states that “the Leadenwah Creek has removed
slightly more than one half of the ring; the southeast quarters remain.” This statement assumes the ring
was formerly a closed circle. A sketch of the ring shows it as a C-shape and suggests that the arms of the
C have been shortened by erosion (Lawrence 1989b: cover), implying it was formerly a longer-armed C-
shape, not a closed circle. The nomination also states that the ring measures 158 feet (48 m) from crest to
crest. However, the 1989 sketch map shows it to be about 62 meters in outside diameter. Although no ex-
cavations are reported for the site, Trinkley (1976:17, 65) analyzed forty-eight Thoms Creek sherds that
were apparently surface collected at the ring and are housed in the South Carolina Institute of Archae-
ology and Anthropology (SCIAA) collections.
The Hanckel ring was listed on the NRHP in 1970 (Hemmings 1970b). Its current condition is unknown.
More study is needed before it can be considered for NHL determination.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 18 m
Sketch
(after Trinkley 1985:108)
This ring holds a unique place in ring studies. It was the first identified in print (Drayton 1802) and the
first in which the plaza area and sub-ring deposits were explored extensively (Trinkley 1980a). As of
2006, it remains extant but in modified form, having been pushed around and infilled by developers. The
diameter and circumference were taken from the early notes by Drayton (1802), for by the time Trinkley
reported on the site, much of the ring had been destroyed. In the 1960s, Donald MackIntosh placed a 5-
by-5-foot (1.5 x 1.5 m) test pit on the northwestern portion of the ring (Trinkley 1980a:161). Trinkley
(1980a) excavated a 102-square-meter area, as well as a 112-square-meter area beneath the ring where
most of the shell ring had been removed. Two smaller units totaling more than 20 square meters were
placed in the plaza area. Additionally, four 10-by-10-foot (3 x 3 m) units totaling 36 square meters were
placed on the outside edge of the ring. This is one of two shell rings dug by Trinkley (1980a, 1997) on
which he based his theory that shell-ring formation resulted from the gradual accumulation of midden
debris—a theory less formally presented in earlier ring studies (e.g., Flannery 1943; Waring and Larson
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1968:273). The ring, listed on the NRHP in 1970, is currently preserved in its considerably disturbed state
by the city of Charleston per an agreement with the neighboring home-owners association. While
Trinkley excavated a considerable amount of sub-ring deposits, undisturbed portions of the ring and sub-
ring structure may still remain to provide significant information as described under Criterion D. Damage
to the architectural integrity, however, seems to have been sufficiently extensive to preclude the site from
consideration for NHL status.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 15 m
Sketch
(after Lawrence 1989b:cover)
shell
borrowed
Shape C
Diam. x Ht. (m) 61 x 3
Area Excavated (m2) 0
Radiocarbon Dates n/a
Stone 0
Bone 0
Archaic Ceramics 85 Thoms Creek
Other 0
References Anonymous 1969; Hemmings 1970c; Trinkley 1976:17, 65
The NRHP nomination states that the ring is 156 feet (48 m) from crest to crest and rises 4 feet (1.2 m)
above the plaza. An anonymous article (probably written by Caldwell [Trinkley 1980a:33]) states it is 10
feet (3 m) high and 200 feet (61 m) in diameter (Anonymous 1969:1). The nomination assumes that the
ring was originally a closed circle, but the “marine erosion has removed 15 percent of the ring,” an inter-
pretation iterated in a map on the cover of Lawrence’s (1989b) report. Trinkley (1976:17) states that
Caldwell and Waring excavated at the ring, however no information about their work survives. The
Anonymous (1969:1) article suggests that an unstated number of sherds were likely Thoms Creek as op-
posed to Stallings. Trinkley (1980a:33) speculates that the sherds may have been excavated by Caldwell.
The ring was listed on the NRHP in 1970 (Hemmings 1970c). Its current condition is unknown. More
study is needed before it can be evaluated for NHL consideration.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 15 m
Sketch
(after Lawrence 1989a:cover)
The NRHP nomination form indicates the site is a “ring shape” and “appears to be largely intact.” The
nomination states that the ring is 178 feet (54 m) from crest to crest (Hemmings 1970a). Judge and Smith
(1991:36) suggested that the average diameter is 54 meters and the maximum diameter 62 meters. A
sketch, however, shows the ring to be about 76 meters in outside diameter (Lawrence1989a: cover).
Bragg (1925:4) reported the site as an “unbroken circle of shell.” However, the mapped and described
dimensions indicate it is oval. No excavations have been reported for the site, and no ceramics or other
artifacts have been enumerated, although Trinkley (1980a:63) stated that Awendaw Finger Pinched domi-
nated the decorated types. Judge and Smith (1991:37) suggested that the ceramics included Stallings,
Awendaw, and Thoms Creek. However, Trinkley provided sherd counts only for Thoms Creek and re-
lated series (1976:65). The nomination mentions the presence of “a number of worked shell artifacts” in
the Charleston Museum, and Bragg (1925) reported whelk tools along with pottery on the surface of the
site. The current condition of the ring is unknown. It was listed on the NRHP in 1970 (Hemmings 1970a).
More study is needed for consideration as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Surface Topography
(after Trinkley 1980a:251)
contour interval = 0.1 m
Shape C
Diam. x Ht. (m) 57 x 0.6
Area Excavated (m2) 135
Radiocarbon Dates n/a
Stone 1 PP, 7 hammerstone, 1 flake, 1 pendant, 8 heating stone
Bone 1 pin, 6 antler
Archaic Ceramics 1,506 Thoms Creek, 15 Awendaw
Other 10 whelk, 4 sherd abrader, 2 clay lump
References Trinkley 1980a; Lawrence 1991b
There is some discussion regarding the shape and size of the shell ring. The NRHP nomination (Lawrence
1991b) and Trinkley (1980a:246) both stated that the south to southeast portion of the ring had been dis-
turbed. (The NRHP nomination was prepared in 1991, but the ring was never listed due to the owner’s
objection.) The nomination seems to be repeating information provided by Trinkley (1980a). Trinkley’s
(1980a:251) contour map labeled the midden in this area as indistinct. He stated that “the southern third of
the shell ring” was utilized as “causeway fill” and was disturbed by driveway construction and use. An-
other map showed the ring missing in that area and labeled it “shell borrowed” (Lawrence 1989a). How-
ever, that map depicts the ring as a closed circle, save for the borrowed southeast area, while Trinkley’s
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1980 map shows it as C-shaped with the opening to the northeast. The nomination provides a modifica-
tion of Trinkley’s map to demonstrate the shape and boundaries of the site. We have chosen to classify
the ring as C-shaped since both Trinkley and the nomination depict it as such. There is no textual descrip-
tion in either source to further resolve the issue.
Trinkley (1980a) is the only one to have excavated and reported on the site, although collections from the
site have been made, some of which were contributed to the Charleston Museum. Thirteen 10-by-10-foot
(3 x 3 m) units were placed in the plaza and two outside the ring to the north (Trinkley 1985:107–109).
While excavations did not provide much information on shell-ring construction, they did show that the
shell ring was placed on a humic layer that extended into the plaza. Although the plaza was nearly sterile
of features (one pit), it did yield abundant artifacts. This suggested that the plaza was used differently than
the ring, not for habitation or refuse disposal (Trinkley 1980a; Lawrence 1991b:2). Trinkley’s observa-
tions are important in that most ring investigations have assumed the plaza is sterile, without having dug
it. Obviously this and other excavations of the plaza (e.g., Russo 1991; Thompson 2006; Waring and Lar-
son1968) demonstrate that they are not sterile, but that the density of refuse in ring plazas (and presuma-
bly the intensity of activities associated with midden accumulation) is greatly reduced.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
In 1970, Hemmings wrote the nomination for NRHP listing of this ring, but not much more has been writ-
ten. A sketch map found on the cover of Lawrence’s report (1989a) shows the site to be a closed ring
about 73 meters in diameter. Apparently, Dorroh (1971:36) borrowed this sketch for her redrawing, com-
ing up with about the same diameter. Dorroh (1971:36) indicated an “old excavation” on the southwest
portion of the ring measuring about 20 by 40 feet (6.1 x 12.2 m). This may be a Waring excavation unit
(see below). Gregorie (1925:16) suggested that the circumference at 600 feet (183 m), if circular, would
make the diameter about 58 meters. This agrees with Judge and Smith’s (1991:36) estimate of maximum
diameter at 184 feet (56 m). Somewhere before 1965, Waring must have visited (and dug at) the site, for
that year Waddell published a radiocarbon date on oyster shell collected by Waring (Sassaman and
Anderson 1994:93). Waring’s comment on the age also reveals that he identified Awendaw pottery from
the site. No enumeration of recovered pottery has been reported.
The ring was listed on the NRHP in 1970 (Hemmings 1970d). Its current condition is unknown. Further
study is needed to evaluate its potential for NHL consideration.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
v
Trinkley (1980a:246) stated that “two other shell rings, Buzzards Island (SoC 57) and Crow Island
v
(SoC 268), are found within 2,000 feet of Stratton Place.” In the NRHP nomination sheet for Stratton
Place, Lawrence (1991b:1) stated that “other Early and Late Woodland period shell middens lie close by,
and one undoubted shell ring (Buzzards Island, 38CH23) and one likely shell ring (Crow Island, 38CH60)
both occur within a half mile of Stratton Place.” An aerial photograph shows Crow Island to be C-shaped,
with about the same diameter as that of the nearby Buzzards Island shell ring. But no ground truthing has
been conducted to verify its shape, content, or cultural affiliation. Crow Island remains a possible Archaic
shell ring. More study is required to determine its potential for NHL consideration.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 6m
Surface Topography
contour interval = 0.5 m
Shape ?
Diam. x Ht. (m) 28? x 2.9
Area Excavated (m2) 24
Radiocarbon Dates 3820+/-185; 4170+/-350
Stone 1 PP
Bone # pins
Archaic Ceramics >5,008 Thoms Creek, 424 fiber-tempered
Other 0
References Cable 1993
Virtually all archeologists have referred to this site as a shell midden or mound except Cable (1997). It
dates to the time of Archaic rings, contains the same type of artifacts found in Archaic rings, and contains
the same species of shell that make up the matrices of Archaic rings. Yet it is not ring shaped, semicircu-
lar, or otherwise arcuate. Cable (1993) explained that this was due to the fact that most of the ring has
eroded away and that the ring plaza was infilled with shell before it was abandoned. In other words, at
one time it was a shell ring, but ultimately, during its last usage, the ring had been filled in. Thus, the site
would more accurately be described as a shell midden or mound before it was abandoned because any
ring function in which the plaza was used for public intercourse would have been precluded by the infill-
ing. Cable’s (1993, 1997) reasoning for this conclusion regards ceramic distributions too complicated to
discuss here (cf. Russo and Heide 2003). His idea that rings were constructed with quarried shell/midden
and ultimately infilled to make a mound provides a testable alternative hypothesis to the theories that
rings were built from the direct deposit of shell resulting from habitation (Trinkley 1980a) or ceremonial
feasting (Russo 2004) activities.
Sutherland (1974) placed a number of trenches of varying sizes into the midden in addition to a 1-by-1-
meter unit (TU11). Trench 1, consisting of four separate 2-by-2-meter sections, was 8 meters long. The
other trenches, although irregular in shape, measured approximately 2 by 1 meters (Trench 2), 1.5 by 1.5
meters (Trench 3), and 2 by 1.5 meters (Trench 4), according to a map (Cable 1993:159; Sutherland
1974:26).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Trinkley (1976:15) studied 3,693 Thoms Creek sherds from the Sutherland excavations and surface col-
lections. Cable (1993:173) studied 5,432, apparently from the same collection. He identified 424 fiber-
tempered sherds, assigning the rest to two sand-tempered series, Thoms Creek and Horse Island (Cable
1993:180–181). It is unclear if the sherds from these two studies are from the same collection. To play it
safe, we have listed only the larger sample numbers in the summary table above. Sutherland (1974:31)
noted that a Savannah River projectile and a few incised bone pins were recovered.
Listed on the NRHP in 1974, this Late Archaic site is significant regardless of any connection to its pos-
sible shell-ring status. Any consideration for nomination to the NHL under this historic context would
require critical review of the registration requirements. It appears that Spanish Mount would not meet the
requirements for inclusion in that it lacks integrity related to its original shell-ring shape.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Gregorie’s (1925:18) mention of Hobcaw near Charleston Harbor is the only indication that this site is
possibly a shell ring. After describing the Auld shell ring, she stated that “on Mr. J. M. Muirhead’s planta-
tion, Hobcaw, there is a similar circle.” A map associated with the article shows a shell heap at “Hobcaw
Point.” Using this as a reference, others have similarly mapped the location of the supposed shell ring
(Heide and Russo 2003:2; Russo and Heide 2003:5). It is unclear if this particular site has ever been re-
corded in the state files. More investigation is needed to determine if this site is a shell ring.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
ell ed
sh port
?
ns
d
t ra
de
ero
0 15 m
Sketch
(after Lawrence 1989c:cover)
Waring (1968a:182) identified the ring as one of two “oyster shell piles” (also see Walthour, next section)
in the marsh with “curious atoll-like, ring formation(s), the centers of which contain nothing but low
marsh and no shell.” DePratter (1976:104) described it as a ring-shaped midden 30 meters in diameter
with a narrow opening on its western side. But there is some confusion with another nearby site
(9CH14A) that lies to the south. This midden (9CH14A), he described as “tear-drop” shaped, not as a
shell ring (DePratter 1976:106–107). However, the Georgia Archaeological Survey form for 9CH14A
describes the midden as a shell ring, and 9CH14 only as a deposit, with the words “shell ring” crossed out
(DePratter 1978). This may just be a mix-up between the two sites. But the shell-ring form describes the
ring from 9CH14 as 45 meters in diameter, not 30 meters as suggested by DePratter (1976:105). Origi-
nally, Waring (1968a:182) stated that the ring is no more than 23 meters in diameter. For the purposes of
this study, we have assumed that the ring variably referred to as either 9CH14 or 9CH14A is the same as
that described by Waring. The three different size estimates cause problems. In terms of height, Waring
suggested the ring is “not more than four feet high” (1.2 m), while DePratter (1976:105) stated that the
ring stood up to five feet (1.5 m). The survey form stated 1.2 meters. We have chosen to use DePratter’s
measures for our summary statistics because they are associated with the most detailed description of the
site. Waring placed one 10-by-10-foot (3 x 3 m) unit in the ring; DePratter surface collected (Marrinan
1975:126–126). The total area of excavation was 9 square meters. The site is Archaic and is likely a shell
ring. More work is needed to confirm this assessment.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
6.0
0 12m
Surface Topography
(after DePratter 1991:38) D
contour intervals irregular 6.7
C
6 .9
6.9
E
6.4
6.4 B G F
6.6
6.2
A
5.8
Shape U?
Diam. x Ht. (m) 30 x 1.2
Area Excavated (m2) 567
Radiocarbon Dates n/a
Stone 2 PP, 3 GS
Bone 14 pin, 2 other
Archaic Ceramics 147 St. Simons
Other 5 whelk
References DePratter 1991:21–42; Waring 1968a:182
The Walthour Ring is the second “curious atoll-like, ring formation” mentioned by Waring (1968a:182).
They estimated the “atoll” to be no more than 23 meters across. But a contour map produced between
1937 and 1942 showed the atoll-like ring of shell mounds to be about 30 meters across (DePratter
1991:38). And if two additional shell heaps outside the ring are included, the site is 60 meters in length.
The same map shows the mounds of shell to rise as high as 3.5 feet (1.1 m) above the surrounding plain,
not far from the 4 feet (1.2 m) suggested by Waring.
There is some doubt as to whether this site should be classified as an Archaic shell ring, however. The site
is made up of a series of seven isolated shell heaps, not a continuous ring of shell. While isolated shell
deposits can be found below rings (e.g., Thompson 2006), aboveground heaps are typically connected to
form a ring wall. The shape of this site, if a ring, is hard to define; Caldwell and McCann described the
arrangement of the shell heaps as “haphazard” (in DePratter 1991:37).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Of the 1,915 sherds classified by DePratter (1991:41), only 8 percent were Archaic. The contexts from
which the bone pins and other artifacts were obtained, and their possible associations with Archaic mate-
rials is unclear. The site has been suggested to represent “a Wilmington component overlying a…sparse
Deptford occupation” (DePratter 1991:32). At least thirteen excavation areas were placed; impacting all
seven shell heaps that make up the site. Seven 10-by-10-foot (3 x 3 m) units were placed around the site,
with five larger units within the site. The larger units ranged in size from 10 by 20 feet (3 x 6 m) to 20 by
70 feet (6 x 21 m). Our measures suggest that the total excavated area of the site is approximately 567
square meters, comparable to DePratter’s (1991:39) 6,400 square feet (595 square meters).
The site has an Archaic component and contains shell, but is not likely an Archaic shell ring. It was ex-
tensively disturbed by archeological investigations in 1939 and 1940.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
DePratter (1976:72) suggested that “a portion of a circular rim appears to be present, so the site may have
originally been a ring.” The midden has been mined, with the remaining portions some 150 feet (46 m) in
diameter, with a “ridge running 50 to 60 feet (15–18 m) along the western edge of the midden area” (De-
Pratter 1976:107). He noted the site was 10 feet (3 m) thick. DePratter (1976:72) stated that the site con-
tained fiber-tempered sherds and that of the over 1,000 sherds recovered, many were decorated (DePratter
1976:109). A single 3-by-15-foot (0.9 x 4.5 m) trench was placed in the midden.
The site is Archaic and contains shell in a semi-circular distribution. It is thus considered a possible shell
ring. More study is need for consideration as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Very little work has been done on these rings save a walk-over reconnaissance. Although Beasley (1970)
suggested that the absence of pottery might indicate that the rings were preceramic Archaic in origin, no
excavations have been placed at the site. More work should be undertaken to confirm cultural affiliations
and site type.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The location of this possible ring was reported to Beasley (1970:13), but virtually nothing is known about
it. Apparently, no archeologist has ever visited it, and, as with 9CH63, no excavations have been con-
ducted here. More study is need for consideration as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Surface Topography
(after Howard and DePratter 1980:250)
contour interval = 0.2 m
Shape C
Diam. x Ht. (m) 77 x 2.3
Area Excavated (m2) 1.35
Radiocarbon Dates n/a
Stone 0
Bone 0
Archaic Ceramics 173 St. Simons
Other 0
References DePratter 1976:115–117; Howard and DePratter 1980:250–252
Portions of the site shape have resulted from wave-deposited shell, perhaps shell eroded from the northern
section of the site. If only the tallest ridges are taken into consideration, the “ring” seems less ring-like
than square. The two ridges form an almost perfect 90 degree angle. Howard and DePratter (1980:252)
suggested the site was originally a closed circle that eroded into its current shape DePratter (1976:115–
116) described this “main part” of the ring as “an angular C-shaped ridge.” He suggested it bounds a
plaza with no buried shell. If the entire site is considered, however, it forms what might best be described
as an S-shape. Only a few post-Archaic sherds have been recovered. A single 3-by-5-foot (0.9 x 1.5 m)
test pit was placed by DePratter. Although its shape, shell contents, and Archaic artifacts identify it as a
shell ring, determining its original shape and size will require more study before it can be considered for
NHL determination.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 10 m
Surface Topography
(after DePratter 1975:24)
contour interval = 0.1 m
Shape C
Diam. x Ht. (m) 77 x 1.5
Area Excavated (m2) 1.6
Radiocarbon Dates n/a
Stone 2 PP
Bone 1 pin
Archaic Ceramics 210 St. Simons
Other 4 baked clay fragments
References DePratter 1975:23–26; 1976:117
There is some doubt, due to its shape, as to whether Odingsell qualifies as an Archaic shell ring. On De-
Pratter’s (1975:24) contour map, it looks more mound-like than ring-shaped. DePratter (1975:24) called it
a crescent-shaped shell midden, but did not refer to it as a ring. On the Georgia Archaeological Survey
form, the site is described as a “small C-shaped shell midden. Only on an attendant “record of materials
form” is it called a “fiber-tempered shell ring.”
Most artifacts from the site were obtained through surface collections. The single excavation unit, meas-
uring 3 by 6 feet (0.9 x 1.83 m), was placed by DePratter near the western end of the shell midden. Only a
few non-Archaic artifacts were found on or near the surface of this unit. However, numerous plain and
punctated St. Simons ceramics, four baked clay fragments, and a single bone pin were recovered (DePrat-
ter 1975:25–26). More investigation is needed to determine if Odingsell is in fact a remnant of a shell
ring.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
DePratter (1976:72) mentioned this site as “a second ring-shaped midden,” south of the first, Cane Patch,
9CH35. DePratter (1974:33) stated that this crescent of nearly pure oyster shell with little soil abuts the
shoreline on each end, and that the center of the crescent is 100 feet (30 m) from the marsh edge, while
the ring ranges from 3 to 15 meters in wall width. With this in mind, 30 meters plus a 15-meter wall width
would result in an estimated 45-meter ring diameter.
DePratter (1976 and 1974) did not indicate that any excavations were carried out here. Artifacts were ap-
parently surface collected. Based solely on DePratter’s assessment, this is considered a possible Archaic
shell ring. More study is need for consideration as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
It is unclear what the shape of Bony Hammock was at the time of its discovery and how much, if any, of
the site remains. DePratter noted that the oyster-shell ring was C-shaped at one time, but had been greatly
reduced by mining (DePratter 1976:130). The baked clay artifacts recovered from the lowest levels of
DePratter’s 3-by-6-foot (0.9 x 1.8 m) excavation unit indicated an Archaic component, but he noted that
the midden was not abundant with ceramics. More study is need for consideration as an NHL.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The Georgia Archaeological Survey form calls this site a Savannah-period shell heap, while Barnes
(2000:15) refers to it as an Archaic shell ring. DePratter (1976:131) identified the Cannon’s Point shell
ring (9GN57) as 9GN54, which may be the source of confusion. The Goodyear Mound (9GN54) is not a
shell ring.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Thomas (2006) states that the site is a medium-sized, crescent-shaped shell midden—a shape, he notes, is
characteristic of the Late Archaic. The ring is at least 70 centimeters thick, or high, based on the shell
sample obtained from between 60 and 70 centimeters, which was used for a radiocarbon age from Test Pit
1. Three test pits were excavated, and 2.6 cubic meters of material recovered. This suggests that the test
pits may have been 1-by-1-meter units. The St. Simons ceramics were obtained from these test pits
(David Hurst Thomas, personal communication 2006). As at most other rings, multiple seasons of occu-
pation are evidenced in the faunal remains at Long Field Crescent, with year-round occupation not pre-
cluded in the absence of larger analyses.
A draft paper discussing comparative zooarcheological assemblages from two shell-ring sites states that
“two of the sites are classic sea island Archaic shell rings. The northernmost assemblages are from the
Archaic St. Catherines Island Shell Ring (Long Field Crescent, 9LI231; Reitz 1990) and St. Simons Is-
land Marsh Ring (9GN57; Marrinan 1975)” (Reitz and Quitmyer 2006). The authors conclude that the use
of vertebrates cannot conclusively be linked to either ritual feasting or quotidian consumption. Based on
the investigators’ studies, the site appears eligible for nomination to the NRHP as shell ring based on Cri-
terion D. It is currently listed as a contributing element for the St. Catherine’s Island NHL, although it is
not identified as an Archaic shell ring. Upon further study, if high site integrity is present, it may be iden-
tified as a potential NHL under this historic context.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 40 m
Surface Topography
(after Crusoe and DePratter (1976:8)
contour interval = 0.1 m
Shape horseshoe
Diam. x Ht. (m) 40 x 2.4
Area Excavated (m2) 40
Radiocarbon Dates 3625 +/-80; 3470 +/-85
Stone 1 flake
Bone 0
Archaic Ceramics 566? fiber-tempered
Other # antler, # whelk
References Crusoe and DePratter 1976:7–9; Marrinan 1975:126
Crusoe and DePratter (1976:7) called the ring “horseshoe-shaped,” with its shape likely the result of se-
vere erosion. It appears, however, angular like 9Ch77. They excavated a 6-by-6-foot (1.8 x 1.8 m) unit
and a 3-by-15-foot (0.9 x 4.6 m) unit. All ceramics were fiber-tempered except for a few sand-tempered
sherds recovered in the top 6 inches (Marrinan 1975:126). Marrinan suggests that a crushed zone repre-
sented a living floor. As is typical of such zones in shell rings, this one is found as the first (lowest) layer
of shell, while all layers above it consisted mostly of whole, piled oyster shell (Crusoe and DePratter
1976:7). Marrinan (1975:126) also reported an unspecified number of antler and whelk tools. The extent
of the site’s damage still needs to be determined. But it remains a potential source of important informa-
tion on shell rings on the regional and national levels. As such, under Criterion D, it is potentially eligible
for listing in the NRHP.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
On a state site form, DePratter (1984) noted that this site was “once a complete ring” with the appearance
of an “arch-shaped shell ridge.” He stated that the “would-be center of the ring has very little shell.” Ce-
ramics apparently came from a surface collection, as did a few post-Archaic sherds. No excavations have
been placed. This may have been the site Moore (1897:71) mentioned when he described Sapelo Ring 1,
as “one of those symmetrical works of the aborigines made by piling shell through a period of time to
form some definite shape such as [the] great ridge on Barbour’s Island not far from Sapelo….” Excava-
tion and mapping are required to confirm the site’s identification as an Archaic shell ring and possible
consideration for NHL status.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 100 m
Surface Topography
(after Beriault in Dickel 1992:143, 154)
contour interval = 0.1 m
This site is also known in the gray literature as Estero Bay Park, the Boone’s Farm Site, and Bonita Bay
Shell Works Site (Dickel 1992; Hughes 1998:1). The shape of the ring and nearby mound was obtained
from precisely mapped surface topographies by John Beriault (Dickel 1992:143). At 235 meters in length,
the ring is among the largest horizontally, but is not very high. Its relatively short stature may, in part, be
due to historic mining and leveling.
Dickel (1992:158–159) stated that he placed “a single test on the higher ridge elevation on the SW arm.”
He did not state how much of the site he dug during his survey, except to note that in the general site area
“subsurface shovel tests were placed randomly, about every 20 meters on the higher sand ridges” (Dickel
1992:53). The ring, of course, is not a sand ridge, but is made of shell. We can assume he dug more tests
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
than these from his statement that “tests all along the interior and exterior margins of the shell ridges re-
vealed an interesting fact…” (Dickel 1992:158). Hughes (1998:2) suggested that most of Dickel’s tests
were placed not in the shell ring but in the surrounding area to determine where the site ended, and were
largely negative, accounting, perhaps, for the paucity of artifacts recovered. In a personal communication
(2006), Dickel confirmed that most of his “shovel tests” were actually posthole tests placed just outside
the edges of the ring walls. He also said that the “single test on the higher ridge” was actually a looter’s
hole in which he cleaned off the profiles, and from which no artifacts were recovered (Dickel personal
communication 2006).
Dickel (1992:142) placed one test of unknown size in the adjacent sand/shell mound (8LL716), but termi-
nated it at 30 centimeters below surface due to an encounter with a fragment of human cranium. The cul-
tural association of the human remains is unclear since only one abbreviated test was placed. That test
also yielded one post-Archaic ceramic. The mound’s proximity to the ring suggests some contemporane-
ous association, but the ceramic does not. In other words, if the ceramic is associated with the burial, nei-
ther may be associated with the construction of the mound. However, the mound may be Archaic, and
both the human remains and ceramic may be intrusive. Similar intrusion is found at the nearby Horr’s
Island, Mound A where burials post-dated mound construction.
In 1993, Houck (1996) excavated six 1-by-2-meter units in the ring prior to road construction across the
southwestern arm of the shell ring. After the units were completed, a backhoe was used to open adjacent
areas. Four 1-meter-wide trenches, measuring 9 meters, 2 meters, 1 meter, and 38 meters long, were
placed next to the units. In addition, Houck excavated a 2-by-3-meter unit. Heavy machinery was used to
scrape the entire right-of-way across the ring in order to reveal features, but none were found. The precise
area uncovered is unclear, but it might have been as much as 5,000 square meters (Houck 1996:11).
In 1996, fifty-eight 0.5-by-0.5-meter shovel tests and one 1-by-2-meter excavation unit were placed in the
plaza prior to construction for a parking lot (Hughes 1998:5). No artifacts were recovered in the 1/4-inch
mesh screens used. Of all the testing and excavations at the site, only Houck described recovering any
artifacts. The few whelk and limestone artifacts he found reflect types also found at the nearby Horr’s
Island ring, while the bone pin artifacts are of types typically found at other shell rings.
A large parking lot, roadways, and a private museum, dedicated, in part, to bringing the site’s cultural
importance to the public’s attention, occupy much of the surface of the site. Nevertheless, most of the ring
and the nearby mound remain intact. Under Criterion D, the site is eligible for listing in the NRHP. More
study is needed to determine the cultural affiliation of the mound and the integrity of the ring before they
can be considered for NHL status under this historic context.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 20 m
Sketch over Surface Topography
(after Curren et al. 1987:7)
contour interval = 0.1 m
Shape C
Diam. x Ht. (m) 77 x 0.9
Area Excavated (m2) 150+
Radiocarbon Dates 3036+/-60; 3630+/-50; 3690+/-50; 4030+/-90; 4070+/-80; 4100+/-80
Stone 2 sandstone hone
Bone 1 awl
Archaic Ceramics 0
Other 2 whelk, # baked clay object fragments
References Curren et al. 1987; Thomas and Campbell 1991
The radiocarbon ages and apparent C-shape are the primary traits that suggest this site may be a Late Ar-
chaic shell ring. Militating against its identification as a shell ring is the absence of shell in much of the
ring and the general lack of shell mounding anywhere. Shell, when found, seems to have been placed in
pit features.
In 1976, Benton reported that the site was a 70-by-20-meter linear shell midden with two 5-by-5-foot (1.5
x 1.5) excavation units dug by amateur archeologists. The two extant units were cleaned and profiled by
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Benton (1976:36), who also placed a small test pit in the site. Only two whelk artifacts were recovered
(Benton 1976:38, 40). The amateur’s report stated that fired clay fragments were found in the tests, and
that the second test produced a bone awl and two small sandstone hone fragments (Benton 1976:47;
Sharon 1967:47).
In a letter dated October 16, 1984, Thomas challenged Benton’s description of the site as a linear midden.
Basing his analysis on 125 shovel tests of unstated sizes, Thomas suggested that the site was horseshoe
shaped, measuring some 120 by 90 meters with an opening to the northwest (Thomas and Campbell
1991:105). He noted, however, that the northwest end of the arm of the C did not contain shell, thus ren-
dering the ring more arc- than C-shaped (see Thomas and Campbell 1991:106). Thomas’s shovel tests
yielded “several flakes of quartzite and several fragments of baked clay” (Thomas 1984:2).
Apparently using the Thomas map as a guide, Curren placed six trenches across the ring and into the
plaza (Curren et al. 1987:18). The trenches were each 1 meter wide, but varied in length as follows:
Trench 1 = 34 m; Trench 2 = 32.5 m; Trench 3 = 30.5 m; Trench 4 = 20.5 m; Trench 5 = 12 m; and
Trench 6 = 18 m. These showed the shell deposits to be sporadically distributed in pits and sheet middens,
with little to no mounding. (The height in the summary table refers to the depth of the deepest shell pit
rather than the height of shell above a level plaza). No contiguous ring of shell exists, which may indicate
the site was a ring in the making (see Sapelo 3) or a ring that had been mined for its upper level shell,
leaving behind mostly sub-ring features (see Coosaw 3 and Lighthouse Point). If a true shell ring, it is
unusual. Curren suggested that the plaza area may have represented a spring head; if so, this would sug-
gest that the site is not an Archaic shell ring—a site type defined by mounded shell surrounding a central
plaza, not a spring. The shape of the ring presented by Curren differs somewhat from that presented by
Thomas and Campbell (1991:106). Curren noted thinner ring wall widths overall and shell in the north-
west arm where Thomas and Campbell showed none existed (Curren et al.1987:6). Curren also showed a
differently configured shell distribution in the northeast arm, where, curiously, the archeologists seem not
to have tested (Curren et al. 1987).
Thomas (1984) stated that the site was scheduled for destruction in advance of development. It has pre-
sumably been destroyed. Identifying the site as a shell ring remains problematic due to its lack of
mounded shell, the vagaries in investigation and reporting regarding its shape, and the possibility that no
plaza is present. Clarification of these aspects needs to be determined in order to ascertain eligibility un-
der the NHL historic context for Archaic Shell Rings of the Southeast United States. The point may be
moot, however, if the site has been destroyed. More study is needed to determine its present condition.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
0 40 m
Surface Topography
(after Thomas and Campbell 1991:105)
contour interval = 0.2 m
Shape C
Diam. x Ht. (m) 120 x 1.2
Area Excavated (m2) 4
Radiocarbon Dates n/a
Stone # steatite
Bone # pins
Archaic Ceramics 0?
Other # baked clay object, # shell beads
References Thomas and Campbell 1991
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
The site is a shallow C-shape, with its ring walls wider than the plaza—an unusual set of parameters for a
shell ring. The ring seems to have suffered from mining, perhaps altering its dimensions in unknown
ways.
Classifying Buck Bayou as an Archaic shell ring is somewhat problematic. Only one significant primary
publication describing the site exists (Thomas and Campbell 1991), and its goal was not to describe the
structure of the site or detail the artifacts recovered, as much as to present an overview of the artifacts and
an interpretation of their significance to a regional trade network. The report noted the artifacts recovered
from a 2-by-2-meter unit, linking them culturally not to the Atlantic coast Archaic shell-ring cultures, or
even contemporary Late Archaic shell-ring cultures on the Gulf coast, but to the lower Mississippi Valley
Poverty Point culture. Primary to this linkage is the presence of Poverty Point objects (i.e., baked/fired
clay), which the authors described as being “scattered throughout the shell midden at Buck Bayou”
(Thomas and Campbell 1991:108). No enumeration of the artifacts recovered from the site is given. Other
markers of Poverty Point associations, such as Jaketown perforators and Motley and other points, do not
seem to have been recovered at Buck Bayou, although they have been found at other nearby sites in lim-
ited numbers (Thomas and Campbell 1991:108–109). Steatite, a common artifact at Poverty Point sites, is
present at Buck Bayou in unstated numbers, as are bone pins and beads (Thomas and Campbell 1991:108,
111). The authors wrote that “there seems to be a moderately active bone and shell tool industry associ-
ated with Elliott’s Point”—the Late Archaic culture to which they assigned Buck Bayou. Thomas and
Campbell (1991:111) illustrated two shell beads recovered from Buck Bayou. It is not clear if any fiber-
tempered ceramics were recovered from the site, although they are present in the area (Thomas and
Campbell 1991:113).
While all the features described (arcuate settlements, baked clay objects, steatite) are common to interior
Poverty Point sites, they are also markers of the coastal Archaic shell rings described herein, and thus not
exclusively associated with Poverty Point cultures or trade networks. Other objects at Buck Bayou, such
as the relatively limited numbers of chipped stone, microflint, or lapidary artifacts, are common to coastal
Late Archaic shell rings and distinct from Poverty Point sites. Also moderately active bone and shell tool
industries are common to coastal Archaic shell rings and uncommon at Poverty Point sites. Therefore,
Buck Bayou may, indeed, have held trade and cultural links to Poverty Point, but its connections to the
characteristics of other regional shell-ring cultures seem equally strong.
More work needs to be undertaken at Buck Bayou to determine the causes for the intrusion of the thick
ring walls into the apparent plaza. Artifacts must be enumerated and their stratigraphic positions mapped
in order to determine if the site is an Archaic shell ring or the palimpsest of multiple occupations. Investi-
gation of its current condition needs to be undertaken before consideration as an NHL is granted.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Cedarland
Mound
Claiborne
0 100 m
Surface Topography
contour interval = 1 m
Shape C
Diam. x Ht. (m) 165 x 4
Area Excavated (m2) 22.5+
Radiocarbon Dates 3200+/-130
Stone # flakes, # PP
Bone # awls
Archaic Ceramics 0
Other # shell
References Bruseth 1991; Gagliano and Webb 1970; Webb 1982
Before its destruction, the site was definitively identified as a C-shaped ring of shell surrounding a plaza.
As such, it fits the basic registry requirements for shell-ring classification. It differs from Atlantic coastal
shell rings in a number of ways. For example, the plaza may not have been entirely “clean” or “sterile”
but rather filled with occupation debris and living floor features (Bruseth 1991:9; cf. Gagliano and Webb
1070:49). The relative abundance of chipped stone artifacts from the site seems to differ from typical Ar-
chaic shell-ring assemblages. Unfortunately, artifacts were largely recovered by local amateurs and were
not enumerated in professional reports, other than to state relative abundances. Bruseth (1991:12), for
one, notes that “abundant” artifacts included sandstone slabs, while “common” artifacts included a variety
of stemmed projectile points, bannerstones, drills and microliths—typically considered uncommon at
other ring sites. However, other “common” artifacts identified in the assemblage include typical Archaic
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
shell-ring objects, such as whelk gouges and scrapers, hard clam anvils, and “occasional” antler and bone
tools (Gagliano and Webb 1970:68).
Gagliano and Webb (1970) dug an 85-meter-long trench across the ring (no width data was available),
along with a 10-by-20-foot (3 x 6 m) excavation unit. Additionally, two 5-by-5-foot (1.5 x 1.5 m) units
were placed by Bruseth in 1972 (1991:7, 11). It is unclear from where the artifacts came: the controlled or
uncontrolled collections from the site, or the “restricted area” around the site.
Because the site has been destroyed, it does not meet registry requirements for shell rings under this NHL
historic context. However, knowledge of its structure, artifactual contents and contexts, and environ-
mental setting are important to our understanding of the social and environmental forces behind the rise
and use of shell rings during the Archaic period.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Cedarland
Mound
Claiborne
0 100 m
Surface Topography
contour interval = 1 m
Shape C
Diam. x Ht. (m) 200 x 2
Area Excavated (m2) unknown
Radiocarbon Dates 3100+/-110; 3385+/-140; 3470+/-160; 3990+/-80
Stone # flakes, # PP, #steatite
Bone # awls
Archaic Ceramics # fiber-tempered
Other # shell, >12,000 baked clay objects
References Bruseth 1991; Gagliano and Webb 1970
Similar to the Cedarland site, Claiborne was definitively seen as C-shaped semi-circular ring of shell sur-
rounding a plaza prior to its destruction. As such, it too fits the basic criteria for classification as a shell
ring. However, it differed from Atlantic coastal shell rings in a number of ways. One, the predominant
shell was not oyster, but wedge clam (Rangia sp.). Two, the ring seems to have been built on the slopes of
bluff ridges, rather than through the piling of shell on level ground (Bruseth 1991:16). Three, the site is
made up of a substantial amount of midden with little or no shell. The amount and kinds of artifacts also
differed: over twelve thousand Poverty Point objects were recovered, steatite vessels were common, lapi-
dary artifacts abundant, and lithic projectile points common. Exactly how many of each class of these arti-
facts were recovered has not been reported (Bruseth 1991:14–18; Gagliano and Webb 1070:66–69). The
Claiborne site has been subjected to pot hunters, amateur archeologists, and bulldozers. Little professional
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
archeological work was carried out there. As a result, little to no data is available regarding the number of
excavations.
We believe the Cedarland and Claiborne sites, which sit adjacent to each other, are sequential constructs
of the same people, with Cedarland being an earlier Late Archaic village and Claiborne, a more recent
Poverty Point Archaic village. While the limited Archaic radiocarbon ages recovered from Claiborne do
not support this interpretation, the artifact assemblages seem to. Technically, these sites can be interpreted
as shell rings and, as such, are evidence that the Poverty Point coastal cultures participated in a tradition
of ring building that was widespread throughout the coastal southeastern United States.
Because the site has been destroyed, it does not meet registry requirements for shell rings under this NHL
historic context. However, knowledge of its structure, artifactual contents and contexts, and environ-
mental setting are important to our understanding of the social and environmental forces behind the rise
and use of shell rings during the Archaic.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
References
Anderson, David G.
1977 A History of Prehistoric Archaeological Investigations in the Coastal Plain of South Carolina. South
Carolina Antiquities 9(2):367–398.
Anonymous
1969 Horse Island. University of Georgia Notes, November. Reprinted [1989a] in Studies of Southeastern
United States Aboriginal Shell Rings, Part 1, edited by David R. Lawrence, pp. 103–110. Department
of Geological Sciences, University of South Carolina, Columbia.
Aten, Lawrence E.
1999 Middle Archaic Ceremonialism at Tick Island, Florida: Ripley P. Bullen’s 1961 Excavations at the
Harris Creek Site. The Florida Anthropologist 52(3):131–201.
2002 Late Archaic–Early Woodland Period Shell Rings of the South Atlantic Coast. National Register of
Historic Places, Multiple Property Documentation Forms. United States Department of the Interior,
National Park Service, Atlanta.
Barnes, Mark R.
2000 Late Archaic–Early Woodland Period Shell Rings of the South Atlantic Coast. National Register of
Historic Places Multiple Property Documentation Form. National Park Service, Atlanta.
Beasley, Dana
1970 Archaeological Investigations on Skidaway Island, Chatham County, Georgia. Proposal to the
Branigar Organization. Ms. on file, South Carolina Institute of Archaeology and Anthropology,
Columbia.
Bense, Judith A.
1994 Archaeology of the Southeastern United States: Paleoindian to World War I. Academic Press, New
York.
1998 Santa Rosa-Swift Creek in Northwestern Florida. In A World Engraved: Archaeology of the Swift
Creek Culture, edited by Mark Williams and Daniel T. Elliott, pp. 247–273. University of Alabama
Press, Tuscaloosa.
Benton, Dale
1976 Archaeological Survey of Bluewater Development, Okaloosa County, Florida. Miscellaneous Project
Report Series, No. 40, submitted to Henningson, Durham, and Richardson. On file, Florida Department
of State, Division of Archives, History, and Records Management, Bureau of Historic Sites and
Properties. Tallahassee.
Beriault, John G., Robert S. Carr, Mark Lance, and Steven Bertone
2003 A Phase I Archaeological Assessment of the Ten Thousand Islands Collier County, Florida. AHC
Technical Report No. 434. Archaeological and Historical Conservancy, Miami. DHR Grant No.
F0221. Submitted to the Rookery Bay National Estuarine Research Reserve.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Beta Analytic
2006a Report of Radiocarbon Dating Analysis for Samples Beta-213395 and Beta-213398. Beta Analytic,
Miami. On file, Southeast Archeological Center, National Park Service, Tallahassee.
2006b Report of Radiocarbon Dating Analysis for Samples Beta-213396, Beta-213397, Beta-213399, and
Beta-213400. Beta Analytic, Miami. On file, Southeast Archeological Center, National Park Service,
Tallahassee.
Bragg, Laura M.
1925 An Indian Shell Culture in South Carolina. The Charleston Museum Quarterly 1(2):3–7.
Brooks, Mark J., Peter A. Stone, Donald J. Colquhoun, and Janice G. Brown
1989 Sea Level Change, Estuarine Development and Temporal Variability in Woodland Period Subsistence-
Settlement Patterning on the Lower Coastal Plain of South Carolina. In Studies in South Carolina
Archaeology: Essays in Honor of Robert L. Stephenson, edited by Albert C. Goodyear III and Glen T.
Hanson, pp. 91–100. Anthropological Studies 9. South Carolina Institute of Archaeology and
Anthropology, University of South Carolina, Columbia.
Bruseth, James E.
1991 Poverty Point Development as Seen at the Cedarland and Claiborne Sites, Southern Mississippi. In
The Poverty Point Culture: Local Manifestations, Subsistence Practices, and Trade Networks, edited
by Kathleen M. Byrd, pp. 7–25.Geoscience and Man 29. Louisiana State University, Baton Rouge.
Cable, John S.
1993 Cultural Resources Survey and Archeological Site Evaluation of the Edisto Beach State Park, Colleton
County, South Carolina. New South Associates Technical Report 165, Stone Mountain, Georgia.
Submitted to South Carolina Parks, Recreation, and Tourism, Columbia.
1995 Intensive Archaeology Survey of Salt Pond, South Tibwin, Palmetto Plantation, and Brick Church
Tracts, Wambaw Ranger District, Francis Marion National Forest, Charleston County, South
Carolina. Cultural Resource Management Report 94-07. New South Associates Technical Report 243,
Stone Mountain, Georgia. Submitted to Francis Marion and Sumter National Forests, U.S. Department
of Agriculture, Forest Service, McClellanville, South Carolina.
1997 The Ceremonial Mound Theory: New Evidence for the Possible Ceremonial Function of Shell Rings.
From South Carolina Archaeology Week poster, Shell Rings of the Late Archaic. South Carolina
Institute of Archaeology and Anthropology, University of South Carolina, Columbia.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Caldwell, Joseph R.
1952 The Archaeology of Eastern Georgia and South Carolina. In Archaeology of the Eastern United States,
edited by James. B. Griffin, pp. 312–321. University of Chicago Press, Chicago.
Calmes, Alan
1967 Test Excavations at Two Late Archaic Sites on Hilton Head Island, Beaufort County, South Carolina.
Paper presented at the Southeastern Archaeological Conference, Macon, Georgia, Nov. 9, 1967.
1968 Test Excavations at Three Late Archaic Shell-Ring Mounds on Hilton Head Island, South Carolina.
Southeastern Archaeological Conference Bulletin 8:45–48.
Cherkinsky, Alexander
2002a E-mail correspondence to Gregory Heide concerning Sample GX-29516. On file, Southeast
Archeological Center, National Park Service, Tallahassee.
2002b E-mail correspondence to Michael Russo concerning Samples GX-26118, GX-26119, GX-25976, and
GX-25977. On file, Southeast Archeological Center, National Park Service, Tallahassee.
2006a E-mail correspondence to Rebecca Saunders concerning Sample GX-25750. On file, Southeast
Archeological Center, National Park Service, Tallahassee.
2006b E-mail correspondence to Rebecca Saunders concerning Samples GX-30737, GX-30378, GX-30379,
and GX-30340. On file, Southeast Archeological Center, National Park Service, Tallahassee.
Crane, Horace R.
1956 University of Michigan Radiocarbon Dates I. Science 124:664–672
Curren, Caleb, Amy J. Belanger, Cheryl Claassen, Lee A. Newsom, and Michael Russo
1987 Archaeology at Bluewater Bay (8Ok102): A Late Archaic Period Site in Northwest Florida. University
of West Florida, Office of Cultural and Archaeological Research, Report of Investigations 9.
Pensacola.
DePratter, Chester B.
1974 An Archaeological Survey of Ossabaw Island, Chatham County, Georgia: Preliminary Report. Ms. on
file (Research Manuscript 344), University of Georgia, Laboratory of Archaeology, Athens.
1975 An Archaeological Survey of the P. H. Lewis Property on Skidaway Island, Chatham County, Georgia.
Manuscript on file, University of Georgia, Laboratory of Archaeology (Research Manuscript 343).
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1976 The Shellmound Archaic on the Georgia Coast. Master’s thesis. University of Georgia, Athens.
1978 Georgia Archaeological Survey form on Oemler (9Ch14). University of Georgia, Department of
Anthropology, Georgia Archaeological Site File, Athens.
1979 Shellmound Archaic on the Georgia Coast. South Carolina Antiquities 11(2):1–69.
1991 W.P.A. Archaeological Excavations in Chatham County, Georgia, 1937–1942. University of Georgia
Laboratory of Archaeology Series Report Number 29, Athens.
2005 Survey of Marsh Middens, Georgetown County, South Carolina. Legacy 9(1–2):18.
Dickel, David
1992 An Archaeological and Historical Survey of Bonita Springs, Parcel Three, Lee County, Florida. AHC
Technical Report 43. Archaeological and Historical Conservancy, Davie, Florida.
Dillehay, Tom D.
1990 Mapuche Ceremonial Landscape, Social Recruitment and Resource Rights. World Archaeology
22(2):223–241.
Dorroh, Rita J.
1971 The Vegetation of Indian Shell Mounds and Rings of the South Carolina Coast. Master’s thesis,
Department of Biology, University of South Carolina, Columbia.
Drayton, John
1802 A View of South Carolina, as Respects Her Natural and Civil Concerns. W. P. Young, Charleston.
Edwards, William E.
1965 A Preliminary Report on the Sewee Mound Shell Midden, Charleston County, South Carolina. Report
submitted to the U.S. Forest Service. Ms. on file, South Carolina Institute of Archaeology and
Anthropology, University of South Carolina, Columbia.
Flannery, Regina
1943 Some Notes on a Few Sites in Beaufort County, South Carolina. Anthropological Papers No. 21. From
Bureau of American Ethnology Bulletin 133, pp. 143–153. Smithsonian Institution, Washington.
Ford, James A.
1966 Early Formative Cultures in Georgia and Florida. American Antiquity 31(6):781–799.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Fraser, Douglas
1968 Village Planning in the Primitive World. George Braziller, New York.
Gardner, Jeffrey W.
1992 Sewee Fire Survey: Archaeological Survey of 135 Acres in the Sewee Historic Area, Wambaw Ranger
District, Francis Marion National Forest, Charleston County, South Carolina. Draft report produced
by Brockington and Associates, Atlanta, for the USDA National Forest Service, Wambaw Ranger
District, McClellanville, South Carolina.
Geochron Laboratories
1999 Report of Analytical Work on Sample GX-25750. Geochron Laboratories, Cambridge, Massachusetts.
Submitted to Rebecca Saunders, Louisiana State University, Baton Rouge.
Goodyear, Albert C.
1988 On the Study of Technological Change. Current Anthropology 29:320–323
Griffin, James B.
1943 An Analysis and Interpretation of the Ceramic Remains from Two Sites Near Beaufort, South Carolina.
Anthropological Papers, No. 22. From Bureau of American Ethnology Bulletin 133, pp. 155–168.
Smithsonian Institution, Washington.
Heide, Gregory
2002 Mapping of the Fig Island Shell Rings. In The Fig Island Ring Complex (38CH42): Coastal
Adaptation and the Question of Ring Function in the Late Archaic, by Rebecca Saunders and Michael
Russo, pp. 67–84. Grant No. 45-01-16441, South Carolina Department of Archives and History,
Columbia.
2003 38BU1866: Preliminary Investigations at the Coosaw Shell Rings, Beaufort County, South Carolina.
Paper presented at the 29th Annual Conference of South Carolina Archaeology, Columbia, South
Carolina. February 22, 2003.
Hemmings, E. Thomas
1970a Buzzard’s Island (38CH23). National Register of Historic Places Inventory–Nomination Form.
National Park Service, Washington.
1970b Hanckel Mound (38CH7). National Register of Historic Places Inventory–Nomination Form. National
Park Service, Washington.
1970c Horse Island (38Ch14). National Register of Historic Places Inventory–Nomination Form. National
Park Service, Washington.
1970d Auld Mound or Yough Hall Plantation Shell Ring (38CH41). National Register of Historic Places
Inventory–Nomination Form. National Park Service, Washington.
1970e Fig Island (38CH42). National Register of Historic Places Inventory–Nomination Form. National Park
Service, Washington.
1970f Preliminary Report of Excavations at Fig Island, South Carolina. Notebook 2(9):9–15. Institute of
Archaeology and Anthropology, University of South Carolina, Columbia.
1970g Sewee Shell Ring (38CH45). National Register of Historic Places Inventory–Nomination Form.
National Park Service, Washington.
Houck, Brett A.
1996 Archaeological Excavations at 8LL717, Bonita Springs, Lee County, Florida. Archaeological
Historical Conservancy, Technical Report 78 (revision of 1993 report). Submitted to Bonita Bay
Properties. Copy on file, Bureau of Archaeological Research, Florida Division of Historical Resources,
Tallahassee.
Hughes, Daniel
1996 Archaeological Excavations of the Proposed Estero Bay Park Parking Area, Lee County, Florida.
Archaeological Historical Conservancy, Technical Report 154. Submitted to Bonita Bay Properties.
Copy on file, Bureau of Archaeological Research, Florida Division of Historical Resources,
Tallahassee.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1998 Excavations at 8LL717, Bonita Springs, Florida. Paper presented at the Annual Florida Anthropology
Society Meetings, Gainesville.
Kelly, Robert L.
1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Smithsonian Institution Press,
Washington.
Kennedy, W. J.
1980 Manuscript Map: Reed Mound (originally drawn, 1966). In Cultural Resource Reconnaissance of
Hobe Sound National Wildlife Refuge, Martin County, Florida, by Mildred L. Fryman, David Swindell
III, and James J. Miller, pp. 43. Cultural Resource Management, Inc., Tallahassee. Submitted to
Interagency Archeological Services Division, National Park Service, Atlanta. Contract A-55034(79).
On file, Division of Archives, History, and Records Management, Tallahassee.
Larson, Lewis
1998 Introduction. In The Georgia and South Carolina Coastal Expeditions of Clarence Bloomfield Moore,
edited by Lewis Larson, pp. 1–85. University of Alabama Press, Tuscaloosa.
Lawrence, David R.
1990a Lighthouse Point Shell Ring (38CH12), National Register of Historic Places Registration Form. United
States Department of the Interior, National Park Service, Washington.
1990b Late Archaic-Early Woodland Period Shell Rings of South Carolina, ca. 1000–2200 B.C. National
Register of Historic Places Multiple Property Documentation Form. United States Department of the
Interior, National Park Service, Washington.
1991a Inventory of Aboriginal Shell Rings of South Carolina. Ms. on file, South Carolina Institute of
Archaeology and Anthropology, Columbia.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1991b Stratton Place Shell Ring (38CH24). National Register of Historic Places Registration Form.
University of South Carolina, Columbia.
1989b Studies of Southeastern United States Aboriginal Shell Rings, Part 2. Department of Geological
Sciences, University of South Carolina, Columbia.
1989c Studies of Southeastern United States Aboriginal Shell Rings, Part 3. Department of Geological
Sciences, University of South Carolina, Columbia.
1991a Studies of Southeastern United States Aboriginal Shell Rings, Part 4. Department of Geological
Sciences, University of South Carolina, Columbia
1991b Studies of Southeastern United States Aboriginal Shell Rings, Part 5. Department of Geological
Sciences, University of South Carolina, Columbia
Malinowski, Bronislaw
1929 The Sexual Life of Savages in North-Western Melanesia. Horace Liveright, New York.
Marquardt, William H.
1992 Shell Artifacts from the Caloosahatchee Area. In Culture and Environment in the Domain of the
Calusa, edited by William H. Marquardt, pp. 191–227. University of Florida Institute of Archaeology
and Paleoenvironmental Studies, Monograph 1, Gainesville.
Marrinan, Rochelle
1975 Ceramics, Molluscs, and Sedentism: The Late Archaic Period on the Georgia Coast. Ph.D.
dissertation, Department of Anthropology, University of Florida, Gainesville.
McKinley, William
1873 Mounds in Georgia. Smithsonian Annual Report for 1872, 27:422–428.
McMichael, Alan E.
1982 A Cultural Resource Assessment of Horr’s Island, Collier County, Florida. Florida State Museum
Miscellaneous Project Report Series, No. 15. Gainesville.
Michie, James L.
1979 The Bass Pond Site: Intensive Archaeological Testing at a Formative Period Base Camp on Kiawah
Island, South Carolina. Research Manuscript Series 154. South Carolina Institute of Archaeology and
Anthropology, University of South Carolina, Columbia.
n.d. The Daw’s Island Shell Midden: Cultural Diversity on the Lower South Carolina Coast During the
Late Archaic Period. Unpublished ms. on file, South Carolina Department of Natural Resources,
Columbia.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Milanich, Jerald T.
1971 Deptford Phase: An Archaeological Reconstruction. Ph.D. dissertation, University of Florida,
Gainesville.
Miller, James J.
1998 An Environmental History of Northeast Florida. University Press of Florida, Gainesville.
Moore, Clarence B.
1893 Certain Shell Heaps of the St. John’s River, Florida, Hitherto Unexplored. American Naturalist 27
(319):605–624.
1897 Certain Aboriginal Mounds of the Georgia Coast. Journal of the Academy of Natural Sciences of
Philadelphia 2:71–73.
1898 Certain Aboriginal Mounds of the Coast of South Carolina. Journal of the Academy of Natural
Sciences of Philadelphia 11:147–166.
Newman, Christine L.
2002 Report of Human Remains at Guana River Wildlife Management Area, St. Johns County, Florida.
Florida Bureau of Archaeological Research, C.A.R.L. Archaeological Survey, Tallahassee
Newsom, Lee
1991 Horr’s Island Archaeobotanical Research. In Final Report on Horr’s Island: The Archaeology of
Archaic and Glades Settlement and Subsistence Patterns, by Michael Russo, pp. 591–645. Florida
Museum of Natural History, Department of Anthropology, Gainesville. Submitted to Key Marco
Developments, Marco Island, Florida.
Reitz, Elizabeth J.
1990 Vertebrate Faunal Remains from the St. Catherines Island Transect Survey. Ms on file,
Zooarchaeology Laboratory, Georgia Museum of Natural History, University of Georgia, Athens.
Ritter, Woldemar H.
1933 Shell Mound on Land Owned by Chester Fields, Burton, Port Royal Island, Beaufort County, South
Carolina. Plan of Mound and Location of Test Pits. Map and associated field notes on file at the
Charleston Museum, South Carolina.
Rouse, Irving
1951 Survey of Indian River Archaeology. Yale University Publications in Anthropology, Yale University
Press, New Haven.
Russo, Michael
1991 Archaic Sedentism on the Florida Coast: A Case Study from Horr's Island. Ph.D. dissertation,
Department of Anthropology, University of Florida, Gainesville.
1992 Chronologies and Cultures of the St. Marys Region of Northeast Florida and Southern Georgia. The
Florida Anthropologist 45:107–126.
1993 Subsistence and Seasonality: Zooarchaeological Measures. In Timucuan Ecological and Historic
Preserve Phase III Final Report, by Michael Russo, Ann S. Cordell, and Donna L. Ruhl, pp. 136–172.
SEAC Accession Number 889. Florida Museum of Natural History, Gainesville.
1994 Why We Don’t Believe in Archaic Ceremonial Mounds and Why We Should: The Case from Florida.
Southeastern Archaeology 13:93–108.
1996 Southeastern Mid-Holocene Coastal Settlements. In Archaeology of the MidHolocene Southeast, edited
by Kenneth E. Sassaman and David G. Anderson, pp. 177–199. University Presses of Florida,
Gainesville.
2002a Architectural Features at Fig Island. In The Fig Island Ring Complex (38CH42): Coastal Adaptation
and the Question of Ring Function in the Late Archaic, by Rebecca Saunders and Michael Russo, pp.
69–72. Grant 45-01-16441, South Carolina Department of Archives and History, Columbia.
2002b Faunal Analysis at Fig Island. In The Fig Island Ring Complex (38CH42): Coastal Adaptation and the
Question of Ring Function in the Late Archaic, by Rebecca Saunders and Michael Russo, pp. 141–153.
Grant 45-01-16441, South Carolina Department of Archives and History, Columbia.
2004 Measuring Shell Rings for Social Inequality. In Signs of Power: The Rise of Cultural Complexity in the
Southeast, edited by Jon L. Gibson and Philip J. Carr, pp. 26–70. University of Alabama Press,
Tuscaloosa.
2005 Zooarchaeology of Tree Islands. In Archeological Assessment of the Eastern Everglades Expansion
Lands, Everglades National Park, Florida (draft), by Margo Schwadron, pp. 207–260. SEAC
Accession Number 1928, EVER Accession Number 771. Ms. on file, Southeast Archeological Center,
National Park Service, Tallahassee.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
2002 The Joseph Reed Shell Ring (8Mt13). The Florida Anthropologist 55(2):67–87.
2003 Mapping the Sewee Shell Ring. Report submitted to the Francis Marion and Sumter National Forests,
South Carolina. Copy on file, Southeast Archeological Center, National Park Service, Tallahassee.
2004 The Emergence of Pottery in South Florida. In Early Pottery: Technology, Function, Style, and
Interaction in the Lower Southeast, edited by Rebecca Saunders and Christopher T. Hays, pp. 105–
128. The University of Alabama Press, Tuscaloosa.
Sassaman, Kenneth E.
1993 Early Pottery in the Southeast: Tradition and Innovation in Cooking Technology. University of
Alabama Press, Tuscaloosa.
Saunders, Joe
2004 Are We Fixing to Make the Same Mistake Again? In Signs of Power: The Rise of Cultural Complexity
in the Southeast, edited by Jon L. Gibson and Philip J Carr, pp. 146–161. University of Alabama Press,
Tuscaloosa.
Saunders, Rebecca
2002 Field Excavations: Methods and Results. In The Fig Island Ring Complex (38CH42): Coastal
Adaptation and the Question of Ring Function in the Late Archaic, by Rebecca Saunders and Michael
Russo, pp. 98–140. Grant No. 45-01-16441, South Carolina Department of Archives and History,
Columbia.
2003 Feast or Quotidian Fare? Rollins Shell Ring and the Question of Ring Function. Submitted to the
Florida Department of Archives and History for 1A-32 9697.04. Tallahassee.
2004 The Stratigraphic Sequence at Rollins Shell Ring: Implications for Ring Function. Florida
Anthropologist 57(4):249–268.
Schwadron, Margo
2005 Archeological Assessment of the Eastern Everglades Expansion Lands, Everglades National Park,
Florida (draft). SEAC Accession Number 1928, EVER Accession Number 771. Ms on file, Southeast
Archeological Center, National Park Service, Tallahassee.
Scudder, Silvia J.
1993 Human Influence on Pedogenesis: Midden Soils on a Southwest Florida Pleistocene Dune Island.
Master’s thesis, University of Florida, Gainesville.
Simpkins, Daniel L.
1975 A Preliminary on Test Excavations at the Sapelo Island Shell Ring, 1975. Early Georgia 3(2):15–37.
South, Stanley
1969 Baked Clay Objects from the Site of the 1670 Settlement at Charles Towne, South Carolina. In
Proceedings of the Twenty-Sixth Southeastern Archaeological Conference, Bulletin No. 11, edited by
Bettye J. Broyles, pp. 13–26.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Stephenson, Robert R.
1973 Continuation form for 38Ch45, April 16. South Carolina Institute of Archaeology and Anthropology,
University of South Carolina, Columbia.
Stoltman, James B.
1974 Groton Plantation: An Archaeological Study of South Carolina Locality. Monographs of the Peabody
Museum No. 1. Harvard University, Cambridge, Massachusetts.
Sutherland, Donald R.
1974 Excavations at the Spanish Mount Shell Midden, Edisto Island, South Carolina. South Carolina
Antiquities 6(1):25–36
1993 Eglin Air Force Base: Historic Preservation Plan, Technical Synthesis of Cultural Investigations at
Eglin, Santa Rosa, Okaloosa, and Walton Counties, Florida. Report of Investigations No. 192,
submitted to National Park Service (CX5000-2-0497), Southeast Region, Atlanta.
Thompson, Victor D.
2006 Questioning Complexity: The Prehistoric Hunter-Gatherers of Sapelo Island, Georgia. Ph.D.
Dissertation, University of Kentucky, Lexington.
Trinkley, Michael B.
1975 Preliminary Report of Archaeological Excavations at Lighthouse Point Shell Ring, South Carolina.
Southern Indian Studies 27:3–36
1976 A Typology of Thom’s Creek Pottery for the South Carolina Coast. Master’s thesis, Department of
Anthropology, University of North Carolina, Chapel Hill.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
1980a Investigation of the Woodland Period Along the South Carolina Coast. Ph.D. dissertation, Department
of Anthropology, University of North Carolina, Chapel Hill.
1980b A Typology of Thom’s Creek Pottery for the South Carolina Coast. South Carolina Antiquities
12(1):1–35.
1985 The Form and Function of South Carolina’s Early Woodland Shell Rings. In Structure and Process in
Southeastern Archaeology, edited by Roy S. Dickens Jr. and H. Trawick Ward, pp. 102–118.
University of Alabama Press, Tuscaloosa.
1997 The Gradual Accumulation Theory: The Lighthouse Point and Stratton Place Shell Rings. From the
South Carolina Archaeology Week poster, Shell Rings of the Late Archaic. South Carolina Institute of
Archaeology and Anthropology, University of South Carolina, Columbia.
Tuzin, Donald
2001 Social Complexity in the Making: A Case Study Among the Arapesh of New Guinea. Routledge, New
York.
Walker, Karen Jo
1992 The Zooarchaeology of Charlotte Harbor’s Prehistoric Maritime Adaptations: Spatial and Temporal
Perspectives. In Culture and Environment in the Domain in the Calusa, edited by William H.
Marquardt, pp. 265–366. Institute of Archaeology and Paleoenvironmental Studies Monograph No. 1.
University of Florida, Gainesville.
1968b The Earliest Ceramic Remains—B: Fiber-Tempered Pottery and Its Cultural Affiliations on the
Georgia-Carolina Coast. In The Waring Papers: The Collected Works of Antonio J. Waring, Jr., edited
by Stephen Williams, pp. 253–255. Papers of the Peabody Museum of Archaeology and Ethnology 58,
Harvard University, Cambridge.
1968c The Archaic Hunting and Gathering Cultures. The Archaic and Some Shell Rings. In The Waring
Papers: The Collected Works of Antonio J. Waring, Jr., edited by Stephen Williams, pp. 243–246.
Papers of the Peabody Museum of Archaeology and Ethnology 58. Harvard University, Cambridge.
Webb, Clarence H.
1982 The Poverty Point Culture. 2nd ed., rev. Geoscience and Man 17. Louisiana State University, Baton
Rouge.
NPS Form 10-900-a (Rev. Aug. 2002) OMB No. 1024-0018 (Expires 1-31-2009)
Widmer, Randolph J.
1976 Archaeological Investigations at the Palm Tree Site, Berkeley County, South Carolina. Research
Manuscript No. 103. South Carolina Institute of Archaeology and Anthropology, University of South
Carolina, Columbia.
1988 The Evolution of the Calusa: A Nonagricultural Chiefdom on the Southwest Florida Coast. University
of Alabama Press, Tuscaloosa.
Williams, Stephen
1968 Appendix: Radiocarbon Dates from the Georgia Coast. In The Waring Papers: The Collected Works of
Antonio J. Waring, Jr., edited by Stephen Williams, pp. 329–332. Papers of the Peabody Museum of
Archaeology and Ethnology 58. Harvard University, Cambridge.
Wyman, Jeffries
1875 Fresh-Water Shell Mounds of the St. Johns River, Florida. Peabody Academy of Science Memoir 4,
pp. 3–94. Salem, Massachusetts.