The significance of sinistral whelks from Mississippian archaeological sites

MISSING IMAGE

Material Information

Title:
The significance of sinistral whelks from Mississippian archaeological sites
Abstract gives title as:
Marine shells from Mississippian archaeological sites
Physical Description:
ix, 278 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kozuch, Laura
Publication Date:

Subjects

Subjects / Keywords:
Anthropology thesis, Ph.D   ( lcsh )
Dissertations, Academic -- Anthropology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1998.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Laura Kozuch.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 029240784
oclc - 39659878
System ID:
AA00024962:00001

Full Text













THE SIGNIFICANCE OF SINISTRAL WHELKS
FROM MISSISSIPPIAN ARCHAEOLOGICAL SITES















By

LAURA KOZUCH














A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1998














ACKNOWLEDGMENTS



Many people have helped me bring this manuscript to fruition. First, and

foremost, I would like to thank Dr. Elizabeth Wing, who has guided me and allowed me to explore rny own research interests. She has kept me on track, and made many useful suggestions to help make my project much improved. Dr. Wing has been the best advisor I could want, and I am proud to be associated with her. I hope to continue our productive association.

Many people have aided me in rny archaeological collections research. All have been extremely helpful, and have pointed me toward other resources they thought I should be aware of These curatorial staff members include (in no particular order); Julie Droke and Victoria Byre of the Oklahoma Museurn of Natural History, Dennis Peterson of Spiro Mounds Archaeological Park, Thomas E. Emerson of the Illinois Transportation Archeological Research Program, Ronald E. Hollinger of the Anthropology department at the University of Illinois, Ann Hutflies at the Illinois Museum of Natural History, Lori Pendleton and David Hurst Thomas of the American Museum of Natural History, Malinda Blustain and James W. Bradley of the Robert S. Peabody Museum of Archaeology in Andover, MA, Bonnie W. Styles and Terry Martin of the Illinois State Museum, Robert Huffman of the Alabama Archaeological Society,




il









Eugene M. Futato of the Alabama Museum of Natural History, Dan Swan at the Thomas Gilcrease Institute of American History and Art, David Rosenthal and Deborah HullI y (Smithsonian Institution), and Scott
Walski of the National Museum of Natural Histor I ] I

Merritt, George P. Horse Capture and Mark Clark of the National Museum of the American Indian (.Smithsonian Institution). Paul and Lucy Jones of the Moundville Archaeological Park allowed me to stay at their home in Tuscaloosa, AL, and I thank them for their gracious hospitality. Terry Martin and his family (Claire and Tim) twice let me stay with them in Springfield, IL, and together we enjoyed the summer weather of Illinois. I had a great time during all of my collections research, made possible by the generosity and consideration of the people I worked with. Thanks to one and all--I could not have completed this research project without their help.

Kurt Auffenberg has taught me much of what I know about malacology. He is a patient and jovial teacher. Credit goes to him for initially identifying the dwarf olive shells from the Spiro site materials. I shall miss our regular interactions.

My supervisory committee members have all given freely of their time and

advice, and I appreciate all they have done. Their useful suggestions made my work not only better, but easier as well. They have also consoled me when I needed to be consoled. John H. Moore allowed me to contact some Creek and Seminole peoples that he knows in Oklahoma, and made the ethnography possible. Bill Marquardt supported my initial ideas about this project, and gave me the confidence to continue. Larry Crook helped me to understand the role of music and dance within anthropology, and pointed








out the interactive nature of ethnography. Susan Milbrath's expertise in iconography was invaluable, and her views lent a fresh perspective.

Several ethnographic consultants put up with my probing questions about their beliefs. These include; Bertha Tilkens (Seminole), Linda Alexander (Seminole), John Proctor (Creek), Vivian Proctor (Caddo), Felix Brown (Euchee), Simon Harry (Euchee), Sonny Littlebear (Euchee), Jim Brown (Euchee), and Paul Tiger (Euchee). Steve Howe (Muspa) has been very forthcoming with information, and I hope to continue working with him.

Pam Innes helped me find my way around Oklahoma, and introduced me to some important people. Her logistical support was there when I needed it, as was her sense of humor.

Throughout this project, family has been there to support me. My parents, Bruno and Angela Kozuch, have tried their best to cope with my odd schedules, and have been there when I needed them. Marianne Kozuch and Jerzyk Kozuch have let me use their homes as stop-overs in my travels, and I truly appreciate having such a wonderful family.

Many others have helped in subtle and countless ways. Friends and family have held discussions with me to help clarify my thinking and descriptions. Any particular omissions in these acknowledgments are inadvertent.

Though others have contributed to the outcome, this dissertation is fully my responsibility.







iv














TABLE OF CONTENTS


pMe

ACKN OW LED GME N TS ........... .................................................................. I ................. ii

ABSTRACT ......................................................................... I ........................................... viii

CHAPTERS

I IN TRODU CTION ...................................................................................................... 1

2 PREVIOUS RESEARCH ON MISSISSIPPIAN MARINE SHELL ARTIFACTS ... 8

Introduction ................................................................................................................ 8
Shell A rtifacts and the Southeastern Cerem onial Com plex ...................................... 10
Marine Shell Artifacts and Mississippian Cultural Complexity ................................ 14
Ethnohistoric A ccounts and Archaeological Finds .................................................... 16
C u p s ................................................................................. ................................... 1 7
Ethnohistoric accounts .................................................................................. 17
Archaeological accounts ............................................................................... 19
G o rg ets ................................................................................................................. 2 1
Ethnohistoric accounts ................................................................................... 21
Archaeological accounts ................................................................................ 22
Pendants ............................................................................................................... 23
M arine Shell Beads .............................................................................................. 24
Chert m icrodrills ............................................................................................ 24
W hole m arine shell beads ................................... .......................................... 25
D isk beads ...................................................................................................... 26
Colum ella beads ............................................................................................. 26
M iscellaneous ............................................. ........................................................ 29










v








3 M ALACOLOGY AND SHELL AVAILABILITY ..................................................... 30

M aterials and M ethods ............................................................................................... 32
R e su lts ................................. .......... ........................................................................... 3 4
D isc u ssio n .................................................................................................................. 3 7
W h elk S n ails ........................................................................................................ 3 8
,9
Sinistral whelks .............................................................................................. .3
Dextral whelks ............................................................................................... 44
Dextral M arine Snails .......................................................................................... 44
Implications for Archaeological Studies .................................................................... 46

4 ARCHAEOLOGICAL SAM PLES ............................................................................. 49

Site Descriptions .............................. ................................................................ ........ 50
C a h o k ia ...................................... ......................................................................... 5 0
Ram ey Field and M ound 34 Cahokia ................................................ ......... 54
Powell m ound Cahokia ................................................................................ 55
Kunnem ann m ound Cahokia ....................................................................... 56
M ound 72 Cahokia ...................................................................................... 57
E to w a h ................................................................................................................. 5 9
M o u n d v ille ........................................................................................................... 6 1
S p iro ..................................................................................................................... 6 4
M aterials and M ethods ............................................................................................... 68
M ate ria ls ........................................................................................................ ..... 6 8
M e th o d s ................................................................................................................ 6 8
R e su lts ........................................................................................................................ 7 1
Species Used in Artifact M anufacture ................................................................. 79
Columella Beads ........................................................................................ ......... 79
Burned Specim ens ............................ .................................................................. 80
A p ex A n g les ................................... ..................................................................... 83
Engraved Artifacts ............................................................................................... 83
Experim ental Archaeology ........................................................................................ 84
Discussion and Conclusions ........ ............................................................................. 85
S h e ll B ead s ................................................................................. ......................... 8 7
Shell W orking Technology .................................................................................. 88
Pacific Coast Shells at Spiro .............................................. ................................. 89
Other Shell Artifacts ............................................................................................ 95
Im plications for Archaeologists ...................... .................................................. 97


5 M OTIF FREQUENCIES ........................................................................................... 99
M eth o d s .................................................................................................................... 9 9
C ah o k ia ................................................. .............................................................. 10 0
E to w ah ................................................................................................................. 10 1
M o u n d v ille ........................................................................................................... 10 1

vi








S p iro ..................................................................................................................... 1 0 2
R e su lts ........................................................................................................................ 10 2
Cahokia .............................................................. ................................................. 103
Etow ah ................................................................................................................. 103
M oundville ........................ .................................................................................. 103
S p iro .................................... ................................................................................ 1 0 3
D iscussion and Conclusions ...................................................................................... 108


6 THEORY, SOUTHEASTERN ETHNOGRAPHY, AND IDEOLOGY
OF SHELLS ...................................................................................... I ....... ................

In tro d u ctio n ................................................................................................................
Shell Artifacts W orld-W ide ...................................... ................................................ 112
M arine Shell Trade D uring M ississippian Tim es ...................................................... 115
Previous Interpretations of Southeastern Shell and Spiral Ideology ......................... 118
Ethnohistoric and Ethnographic Evidence of Spiral and Shell Ideology ................... 122
Ethnohistoric Reports .......................................................................................... 122
Ethnographic Reports .......................................................................................... 128
Shell cups ....................................................................................................... 128
Current indigenous thoughts on shells .............................. ...................... 128
Circular dance directions ............................................................................... 130
D iscussion and Sum m ary .............................................................. ............................ 131
Spirals As Representations of Circular D irection ................................................ 132
Spirals and M eaning ............................................................................................ 133
Conclusion ................................................................................................................. 13 -3

7 SUM M AR Y AN D CON CLU SION S ............................................... ......................... 136

Sum m ary .................................................................................................................... 136
Trade Routes .............................................................................................................. 139
Im plications for Southeastern Archaeologists ........................................................... 139
Econom y ................................................................... .......................................... 139
Spiral M ounds ............... ...................................................................................... 140
Conclusions .................................. ............................................................................. 146

REFEREN CES CITED ............................................................................................ ..... 147

APPENDICES

I DATA FROM FLMNH MALACOLOGY COLLECTION ........................................ 168

H DATA FROM ARCHAEOLOGICAL SPECIMENS ................................................. 199

BIOGRAPHICAL SKETCH ............................................................................................ 278

vii















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

MARIN SHELLS FROM MISSISSIPPIAN ARCHAEOLOGICAL SITES By

Laura Kozuch

May 1998


Chair: Elizabeth S. Wing
Major Department: Anthropology

Marine shells were widely traded during the Mississippian period in the

Southeast. The overwhelming proportion of shells traded were sinistral lightning whelk shells, as demonstrated by collections research presented hereinafter. I present data on archaeological shells from four Mississippian sites: Cahokia, Etowah, Moundville, and Spiro.

Research on malacological collections at the Florida Museum of Natural History demonstrates that other large marine snails were also readily available to coastal peoples. The largest shell is, and was, the horse conch. H-owvever, horse conch shells comprise a very minor percentage of shells from the archaeological sites. Clearly, some factor other than size or color motivated the shell trade, and this may have been the unique sinistral spiral of whelk shells of the genus Busycon,




Vill








Ethnographic and ethnohistoric resources indicate a special significance or power of spirals and/or shells. I therefore suggest a partial ideational cause for Mississippian shell trade, spurred on by thoughts concerning spirals and shells. This study falls under the rubric of structural archaeology, in conjunction with an economic approach. If spirals were important symbols, we may be able to more accurately interpret other aspects of late Mississippian society. These include dances and spiral mounds that were built in the Southeast. A more ergonomic approach suggests that sinistral shell cups were simply easier to use and hold. The significance of spirals needs to be investigated further.

New findings include Pacific coast shells from the Spiro site in eastern

Oklahoma. This suggests a much larger trade network than previously thought. Another new finding is that Southeastern peoples from Cahokia used fire to aid in artifact manufacture. Many burned col umel las were recovered from a shel I -working area just east of Monks mound at Cahokia.

Marine shells from Mississippian sites are shown to be helpful to archaeologists to illuminate ideology, technology, and economy. A clear understanding of the zoology and natural habitat of the animals whose shells were traded is instrumental to understand trading patterns and coasts) of origin. Additionally, this study shows that research on existing collections can be very fruitful, even if the collections were excavated over 50 years ago.







ix














CHAPTER
INTRODUCTION



This dissertation examines marine shell artifacts from four inland Mississippian sites in North America. My aim is to look at marine shell artifacts in the same way that researchers have analyzed other artifact classes such as galena, lithics, and copper. Of course shellfish are widely consumed, and utilitarian implements are often made from shells. However, shell artifacts that are traded hundreds, or even thousands of infles cannot be considered as just utilitarian items. Some special significance was attached to traded shell artifacts. This introductory chapter is a synopsis of the pertinent research questions concerning marine shell trade, particularly in the Southeast, as well as brief descriptions of each following chapter.

The topics explored are, previous work on Mississippian shell artifacts; worldwide factors of shell choice in artifact manufacture; the availability of large marine snails in the Gulf of Mexico and Atlantic Ocean; the percentages from Southeastern archaeological sites of different artifacts made from marine shells and the proportions of marine shell species used to make these artifacts; and Southeastern ideology concerning spirals and shells. I collected data from I) modem shells, 2) ethnohistoric and ethnographic sources, 3) experimental archaeology, and 4) archaeological shell artifacts. These data shed light on shell sourcing, shell bead manufacture, shell working technology, and ideology of circular directionality among Mississippian peoples.





2


Zooarchaeology may be defined as the study of animal remains from

archaeological sites. According to that definition, this is a zooarchaeological study. Marine shells from an inland site are usually presented in a list of exotic artifacts, perhaps with a note concerning the closest body of marine water. In contrast, my focus is on the species of shells exchanged during Mississippian times, to identify and quantify the shell artifacts and to examine the shell exchange from an archaeo-ethnozoological viewpoint. The ideational aspects presented here are not often addressed in zooarchaeological reports, and the exchange of marine shells to inland sites is rarely studied from a zooarchaeological perspective.

Most studies of shell gorgets, masks, and cups have concentrated on the designs engraved on some of them or on the geographic source of marine shells from which artifacts were made (Brain and Phillips 1996; Claassen and Sigmannm 1993; Hale 1976; Phillips and Brown 1978; Smith and Smith 1989; Strong 1989; Waring and Holder 1945; Williams and Goggin 1956). These studies have added greatly to our understanding of Southeastern archaeology, yet questions remain which I attempt to address. However, my focus is not on shell engravings, but on the shells themselves that were used to create Southeastern artifacts. If marine shells were thought of as a special artistic medium ( apart from ceramics, copper, or stone), they may have different motif frequencies engraved on them.

The four archaeological sites included in my sample are Cahokia (Illinois),

Etowah (Georgia), Moundville (Alabama), and Spiro (Oklahoma). They were chosen. in order to obtain an adequate sample, since marine shell artifacts are common in the









archaeological assemblages of each of these four sites. The sites are similar because they are the remains of large villages with multiple mounds, and date to the Mississippian time period (800-1450 AD).

Many different types of artifacts manufactured from several species of marine snail shells have been found at archaeological sites throughout the Southeastern United States. Cups, gorgets, pendants, beads of several types, and masks all made from marine shells have been found (Baker 1924;- Brain and Phillips 1996; Hale 1976; Holmes 1883; Kneberg 1959; Milanich 1979; Orr 1946; Phillips and Brown 1978). The majority of these artifacts were made from lightning whelk shells (Hale 1976;- Milanich 1979; Ottesen 1979; Phillips and Brown 1978), however artifacts made from other species such as olive, murex, helmet shells and horse conch shells are also found (Brown 1996:4 17; Parmalee 1958). Beads made from the small shells of marginella and olivella snails have been found by the thousands at Spiro, Cahokia, and Ocmulgee sites (Ottesen 1979). Holmes (1883) first noted that the majority of columella (a columella is the inner portion of a snail shell see Chapter 3) beads at Southeastern archaeological sites are from sinistral whelk shells. This observation has withstood the test of time. Other types of shell artifacts are also found, such as earrings, pins, helmet shell lips, and whole shells. Bits of marine shell were used as inlays to represent eyes in wooden masks found at Spiro, Oklahoma (Brown 1982:47 1), and at Key Marco, Florida (Gilliland 1989). Summaries of previous work on marine shell artifacts from inland Mississippian sites are presented in Chapter 2, focusing on the four target archaeological sites.





4


What was the general availability of large marine snail shells along the coast? Which large gastropod shells were most common and/or easiest to obtain? One way to understand shell availability in the archaeological past is to ascertain availability in recent times. Research was conducted using the Malacology collections at the Florida Museum of Natural History to determine the modem availability of large marine snail shells. Only fourteen species were focused upon, chosen by virtue of their size. Infonriation was collected on locality, size, and depth at which each shell was found. Previous work by malacologists shows that different species of sinistral whelk shells have different apex angles, and this difference enables archaeologists to distinguish between the two recognized species of sinistral Busycon. The horse conch (Pieuroploca gigantea), not the lightning whelk, is the largest marine snail occurring in the Gulf of Mexico or Atlantic Ocean. The horse conch gets as large as 48cm in length and has a dextral shell, while sinistral whelks have a maximum length of41cm. A discussion of the confusing taxonomy of sinistral whelks (Busycon spp.) is presented, and data gathered from malacological collections research are summarized in Chapter 3.

What were the proportions of marine shell species found at Mississippian sites, and what are the proportions of each category of shell artifact? Were marine shell columella beads and gorgets from archaeological sites also made from sinistral whelk shells? No one has systematically attempted to tabulate numbers of species, or sinistral versus dextral marine shell artifacts. My analysis of archaeological specimens from the four sites included in my sample addresses these questions. In addition to identifications and tabulations of sinistral versus dextral shell gorgets, cups, pendants, and columella






5


beads from archaeological sites, brief descriptions of each site (Cahokia, Etowah, Moundville, and Spiro) are provided in Chapter 4. The data indicate (using the apex angle of Bus Ycon artifacts) that most sinistral marine shell cups from inland archaeological sites come from- lightning whelks, rather than prickly whelks (both are sinistral), and are probably from the west coast of Florida. It also is possible to identify the genus of shell from which some of the shell gorget specimens are made, and I explain how this is done. Experimental archaeology was employed to reproduce shell cups. The presence of large numbers of burned columellas at the Cahokia site may be explained by the fact that it is easier to work burned shell than unburned shell. Some of the olivella shell beads and an abalone shell recovered from the Spiro site (in what is now eastern Oklahoma) originated on the Pacific coast. This indicates a much larger trade network than heretofore known. The archaeological sample from the four sites indicate that the great majority of artifacts were made from lightning whelk shells, confirming the suggestions of previous researchers.

If marine shell artifacts were conceptualized as inherently sacred objects, we may see differences in frequencies of motifs among the different artistic media from obtained from the four archaeological sites in my sample. Frequencies of motifs on four media are tallied: marine shell, ceramics, copper, and stone. Results of this analysis are presented in Chapter 5.

An analysis of notable world-wide instances of ethnographic and archaeological marine artifact manufacture reveals two inherent qualities important when choosing shells: 1) color or iridescence throughout the shell, and 2) shell shape. Previous work






6


shows that most artifacts in the Southeast were not made from the largest shells available, horse conch shells. In addition, dextral shells of the same family, and of similar shape and size, were rarely exchanged among Southeast Mississippian sites. Lightning whelk shells were the overwhelming choice for manufacture of artifacts, and have no unique color within the shell matrix. Some other factor besides color, shape or size influenced the choice of shell for artifact manufacture, and this was ideology. As Milanich (1979:86, 95) has pointed out, the reason for this may lie in the fact that these shells are sinistral, or that they symbolize spirals. To be sure, marine shells were used in the Southeast as exotic, prestige items, perhaps to legitimate high social status (Brown et al. 1991; Helms 1992). However, if this was the only incentive for obtaining marine shells, then any marine shell would suffice, as long as the shells were large enough to make artifacts such as cups, gorgets and masks. Clearly, a particular kind of shell was targeted, perhaps due to ideas of circular directionality. This type of study falls under the rubric of structural archaeology, but does not, however, exclude a more economic approach. The theoretical underpinnings and ramifications of such a study are considered in Chapter 6.

What is the importance of the study of marine shells within archaeology and

anthropology? If, as I intend to support, sinistral whelk shells were targeted due to the symbolism of the direction of the spiral, then we must reconsider our ideas on past motivations behind the acquisition of sinistral shells in the Southeast. Malacologists everywhere recognize these shells as peculiar because they are sinistral. It seems likely that indegenous Southeastern peoples might also have believed these shells to be peculiar






7


or special. Ideology is most readily accessible through ethnography, ethnohistory, and extant members of the groups of people in question. Data presented in Chapter 6 include spiral dances, modern Creek, Seminole, and Muspa thoughts on shells, and descriptions of the sacred, spiral fire by William Bartramn (1976 [1789], 1928 [1791 ]).

Conclusions and a summary of the significant findings are in Chapter 7. Shells should not be viewed simply as exotic materials to be used as materials on which engravings were made, but rather as legitimate objects of study in themselves. The presence of Pacific coast shells at Spiro, Oklahoma, supports the contention of some Spiro researchers that Spiro was primarily a trading post. That fire was used to heat shell in order to make it easier to work is supported by archaeological, ethnohistorical, and experimental archaeological research. Apex angles on sinistral Busycon shells are identified as one way to resolve at least the general coastal locale where the shells originated. If most lightning whelk shells originated along the Florida west coast, as I support, a number of questions arise concerning Florida economy. If the main motivating factor in obtaining lightning whelk shells was ideology there are implications for both archaeologists and anthropologists. The ideology of circular directionality in the Southeast may be directly associated with the direction of snail or gastropod coiling. Concepts of spirals may help archaeologists to interpret other archaeological remains, such as spiral mounds. The study of shells in archaeology truly has evolved, as Lynn Ceci (1989: 1) stated, to a treatment "of these objects as sensitive markers for social, political, economic, and ideational processes among prehistoric, historic, and living societies."














CHAPTER 2
PREVIOUS RESEARCH ON MISSISSIPPIAN MARINE SHELL ARTIFACTS



Introduction


It is important to understand prior works on Mississippian and early historic marine shell artifacts. This will give the reader a basis for understanding and critical review. A general description of the varieties of shells, kinds of artifacts found, and the geographic extent of marine shell exchange during Mississippian times is presented first, This is intended to give the reader an idea of the types and numbers of marine shell artifacts that have been recovered. Previous attempts to source the coast of origin of marine shells are summarized. Next, information is presented on preceding studies of shell artifact types, including ethnohistoric and modem descriptions of marine shell usage by indigenous people from inland locales. For ease of reading, each category of artifact is examined separately.

Marine shell use is an ancient and persistent practice in Southeastern North

America. Marine shell artifacts have been found from Archaic sites dating to '000 B.C. (Milanich 1979:87). Cups were made from lightning whelk shells at the Archaic sites of Indian Knoll and at Carlson Annis, both in Kentucky. The utilization of marine shell artifacts continues to this day among some tribes, such as the Creek and Seminoles who were forcibly moved to Oklahoma from their homelands in Georgia and Alabama during the "Indian Removal" of the early 1800s (Moore 1994:132). Some modem Creek
8





9


peoples in Oklahoma currently use trumpets made from horse conch shells (and other materials) to call parishioners to begin Sunday church services. I know of no archaeological examples of trumpets made from any species of shell. However, Bartram (1928:400 [1791 ]) explains that at the spring planting, all the people of one "Muscogulge" (Creek) town were called together to the town square "by the sound of a conch shell."

In North America, locally available shells were also used to make utilitarian tools, such as fishhooks, knives, hoes, spoons, tweezers, and so on (Holmes 1883). Artifacts were fashioned from local freshwater mussel shells to make hoes and implements for processing corn, as ethnographic and archaeological studies indicate (Gradwohl 1982; Parnalee 1958; Titterington 1977:9). In and around coastal areas, marine shells were used to make a variety of tools (Marquardt 1992; Webster 1970). Nonetheless, I am concerned here with marine shell artifacts used by people of the Southeast that made their way to inland archaeological sites.

Shel1 beads have been studied more intensely in recent years. The 1986 Shell

Bead Conference indicates a growing interest in this subfield (Hayes et al. 1989). Recent work corroborates the intensification of studies on shell beads (Carlson 1993; Thomas 1996; Trubitt 1996). Beads made from different species of marine shell at Mississippian archaeological sites include rnarginella, olive, and helmet. Disk beads and columella beads (known as barrel or cylindrical. beads) were made from large marine shells such as lightning whelks and horse conchs, though species identification of these types of beads are not possible at this time.





10


Shell Artifacts and the Southeastern. Ceremonial Complex


Marine shell artifacts help to reveal an extensive geographic range of related

artifacts, usually termed the Southern Cult (Muller 1989; Waring and Holder 1945) or the Southeastern Ceremonial Complex (Galloway 1989; Howard 1968). The Southeastern Ceremonial Complex (SCC) was first presented as an heuristic device by Waring and Holder in 1945, defined by artistic motifs and "ceremonial" objects associated with them. James A. Brown (1976:115) reworked the definition to include "an interregional interaction sphere co-extensive with the distribution of complex Mississippian cultural systems." The SCC is a Mississippian culture, with the added accoutrements of ceremonial objects and characteristic motifs on similar materials, including copper, marine shell, lithics, and ceramics. Copper, marine shell, and some lithics were from exotic locales, and were probably acquired by a simple system of exchange, such as down-the-line trade (Muller 1995). This complex of similar artifacts, consisting of copper, galena, shell, and some artistic motifs, refers to late prehistoric times, before Europeans arrived in the Americas. The core area of the SCC is bounded by the Cahokia site in Illinois to the north, the Hollywood site on the Savannah River in Georgia, the Mount Royal site near the east-central Florida coast, and the Spiro site on the Arkansas River in eastern Oklahoma (Howard 1968).

However, the archaeological sites that have marine shell range much further. The northern limits of marine shell artifacts are the northern plains. Beads made from Busvycon spp. shells were found in South Dakota near the Oahe Dam (Lehmer 1954), and in Ontario, Canada at the Huron Kirche site (Pendergast 1989). Whelk shell (species not








mentioned) gorgets and cherts from Illinois are markers for the late prehistoric Devils Lake-Sourisford burial complex in eastern North Dakota and surrounding areas (Vehik and Baugh 1994:258-259). Marine shell beads made from whelk and olive shells were found at the Mississippian site of Aztalan in southern Wisconsin (Barrett 1970). A "shell mask gorget" made from the outer whorl of a lightning whelk was found in Iowa, along the Iowa River near its confluence with the Mississippi River (Collins 1995). The Fort Ancient site of SunWatch near Dayton, Ohio, had a lightning whelk pendant, disk shell beads of an unspecified type, and many marginella beads (Coovert 1988). To the southwest, a sinistral Busycon spp. gorget with a design similar to those found at the Spiro site in Oklahoma was found near Killeen, Texas (Highwater 1983:299). Even farther south, olive shell and sinistral whelk shell beads were found with a marine shell gorget near McAllen, Texas (Hester and Rodgers 1971). The eastern extent of marine shell gorgets is Virginia (Smith and Smith 1989:11). At least one is definitely made from a sinistral whelk shell found in Russell County, Virginia (Maus 1995), and another found in Washington County has the well-known cruciform motif engraved on the concave side (Peck 1995). The latter was found in association with olive shell and marine shell columella beads. Smith and Smith (1989) report two marine shell mask gorgets from Stafford County, Virginia. Based on the motifs of the Southeastern Ceremonial Complex (Waring and Holder 1945), there appears to be little connection to the shell engravings of the Huastecs of Vera Cruz, Mexico (Anton and Dockstader 1968:70), which have as their hallmark intricate fenestrated patterns cut out of shell. Phillips and Brown (1978:166)





12


suggested a Huastecan connection with the Spiro site, however Brown (1996:20) notes a decline in the theory of Mesoamerican influence on Spiro art styles.

For a long time, archaeologists have attempted to source the coast of origin of marine shells found at inland archaeological sites. Most suggest that the shells came frorn the Atlantic Ocean or the Gulf of M6xico (Brown et al. 1990:260, Yerkes 1989:97), but more specific locations are important for establishing prehistoric connections. A definitive provenance would help to settle questions of cultural interaction and complexity, and movement of resources and ideas. Two projects have attempted to 'dentify sources of shells: a master's thesis by H. Stephen Hale (1976), and a more recent work by Claassen and Sigmann (1993). Both conclude that archaeological shells came from the eastern Gulf of M6xico, however, Hale is more specific. He made the assumption that shells came from areas where the species is most abundant. Hale showed, using data from modem collections, that lightning whelks are most common along the west coast of Florida (1976:71). Hale (1976:71) noted that "the area of maximum occurrence [of lightning whelks] is along the beaches in Monroe, Collier and Lee Counties." These counties are in south and southwest Florida. In contradiction, he also claimed (1976:v) that "The most probable source of this shell for trade to Midwestern archaeological sites was the Apalachicola Bay Area."

The method employed by Claassen and Sigmarm (1993) to source marine whelk shells relied on the assumption that whelk snails incorporate the unique chemical signature of their immediate environment into their shells. They asserted that this chemical signature can be used to identify the specific freshwater drainage that the shells





13


lived closest to. This method has great potential, but requires much biological research

in order to show the basic assumption to be true.

Another problem in sourcing shells used in trade is inherent in identification and

nomenclature. This has led Phillips and Brown (1978) to suggest a possible connection

between Spiro with Huastecan shell artifacts. I will quote their work at length to

illuminate the research problem (Phillips and Brown 1978:26-27):

Dr. Pulley was also of the opinion that designation of a subspecies was not important
for our purposes. He stated that most of the undecorated cups in the McDannald
collection are 'what I call Busycon perversum (Linn6), but which many other authors (Hollister 1958; Abbott 1974) call Busycon contrarium (Conrad)' (letter of February 18, 1975). These shells, we understand, could have come from any part of the Gulf
coast from the Florida Keys to the Straits of Yucatan... But Dr. Pulley goes on to say (in the first letter cited) that there are also in the McDannald collection fragments of other species which lead me to believe that they were all collectedfrom the Mexican coast at least as far south as Cabo Rojo. These 'other' fragments are: Pleuroploca
(fon-lerly Fasciolaria) gigantea and Cassis madagascarensis. The latter, he adds,
'could only have come from the Florida Keys or the Veracruz coast.'...Ever since the
first examples of Southeastern figural art in copper and shell became known, the question of Mesoamerican connections has been argued back and forth but never
settled to anyone's satisfaction, the difficulty being the lack of any tangible evidence
of contact. From the standpoint of decorative technology in shell, the place to look for such contacts has always been the Huasteca, where a vigorous tradition of shell
engraving flourished at just about the right time. Nevertheless, statements in the
archaeological literature have concentrated on the South Atlantic and Florida Gulf
coasts as the probable source of raw materials for all Southeastern shellwork.
Pulley's assertion that at least some of the Spiro shell may have come from the
Huasteca (Tamaulipas-northern Veracruz) could be the first break in an impasse
that has prevailed for over a century [emphases added]

This passage underscores the significance of correctly identifying the place of origin of

marine shells. Phillips and Brown (1978) did not realize that horse conchs (Pleuroploca

gigantea) occur regularly along both east and west Florida coasts. Helmet shells (Cassis

madagascariensis) are also found on the Florida west coast all the way to the panhandle

(see Chapter 3).





14


The importance of pinpointing the source of shells as, for example, the southwest Florida coast versus Huastecan Mesoamerica would be in building hypotheses of social complexity for a specific area. A definitive locale may change our ideas about the rise and fall of the stratified societies believed to have occurred in these coastal areas (Marquardt 1987, 1988; Milanich 1994:362). If such a high value was placed on the shells, the coastal peoples who controlled shell harvesting would gain more wealth than they other-wise would if the shells were of no importance,

Two factors support that marine shell artifacts were not copies of Mesoamerican artifacts: 1) marine shell artifacts were used since Archaic times in the Southeast, and 2) engraving styles differ. The antiquity of marine shell use at inland archaeological sites in the Southeast (see above) suggests a long tradition of shell acquisition and modification. This tends to discount the hypothesis that Mississippian peoples learned shell working techniques from the Huastecs (see Phillips and Brown 1978). Moreover, the particular type of fenestrated engraving found in M6xico is rare from Southeastern archaeological sites, tending to negate the Huastecan connection. Although most gorges are engraved (incised), fenestrated "spaghetti style" gorges are uncommon (Brain and Phillips 1996).


Marine Shell Artifacts and Mississippian Cultural Complexily

A debate surrounding the nature of shell beads in a cultural context has arisen among archaeologists. One basic question is whether beads functioned as wealth or status items (Prentice 1987; Thomas 1996). However I shall leave this complicated topic to other workers, since the distinction between wealth and status is problematic even in extant cultures. A related topic is whether or not shell beads functioned as money.





15


Additionally, the topic of craft specialization in marine shell bead production has relevance to the question of social complexity among Mississippian cultures (Muller 1987; Pauketat 1987; Prentice 1983; Yerkes 1981). Increasing interest in the study of shell beads and their production can only enhance our understanding of past peoples.

The exchange of marine shells may not have been an unchanging, consistent

phenomenon. There may have been long periods of time during which very little, if any, marine shell was deposited at Southeastern archaeological sites (Cobb 1991 ). The assumed lack of shell artifacts from some time periods was used by Cobb (199 1) to postulate that shell artifacts, along with other artifacts, were used as symbols to reproduce social systems. Cobb views changes in the frequencies of exotic trade goods as a correlate to changes in cultural complexity. In contrast, Jon Muller (1989:14) states that, "there were few times after Late Archaic when such 'valued' raw materials as native copper or Busycon shells were not moving across the Southeast." The changes in frequencies of exotic trade items may be a result of our incomplete archaeological record, since, "the archaeological ly surviving artifacts marking long-distance interaction may have been only a small part of the goods that were exchanged." [emphasis in the original] (Muller 1995:320). A theory postulated to explain the use of imported materials, the "Prestige Goods Economy" infers that exotic trade items were used to legitimate high social rank (Brown et al. 1990). For a different viewpoint, see Muller (1995), who calls into question the Prestige Goods Economy since long-distance trade may have been of the simplest type. Again, exchange and use of marine shell artifacts becomes important in our understanding of social relations.





16


Ethnohistoric Accounts and Archaeological Finds

I will now present information from ethnohistoric sources concerning marine

shells, and artifactual data on marine shells from Mississippian sites. Emphasis will be placed on the four target sites: Cahokia, Etowah, Moundville, and Spiro. Ethnohistoric accounts can help us better understand the role that marine shell artifacts had in the lives of Mississippian peoples. All ethnohistoric data are from tribes who once inhabited the Southeast. This will, in turn, aid in understanding the archaeological finds.

The use and trade of marine shells was recorded by early observers. Du Pratz (1975:290 [1774]) states that, "The largest of the shell-fish on the coast is the Burgo, .. The shells have long been in request for tobacco-boxes" and (1975:364 [1774]) that "The womens ear-rings are made of the center part of a large shell, called burgo, which is about the thickness of one's l ittle finger. . Dumont de Montigny, writing about southern Louisiana in the early 1700s states

There are found besides, on the shores of the sea, beautiful shells of a spiral shape
called 'burgau.' They are very suitable for the manufacture of pretty tobacco boxes, for they carry their mother-of-pearl with them. It is of these burgau that the Indian women make their earrings. For this purpose they rub the ends of them for a long
time on hard stones and thus give them the shapes of nails with heads, in order that, when they insert them in their ears, they will be stopped by this kind of obstruction.
(In Swanton 1946:486).

John Lawson (1967:203) tells us that shell gorgets were highly valued, the going rate being three or four dressed buckskins. Although Adair does not mention the name of the shell, he does describe the price of a shell bead (Adair 1971:178 [1775]):

Before we supplied them with our European beads, they had great quantities of
wampum;. .. made out of conch-shell, by rubbing them on hard stones, and so they
form them according to their liking. With these they bought and sold at a stated
current rate, without the least variation for circumstances either of time or place;, ...






17


Formerly, four deer-skins was the price of a large conch-shell bead, about the length
and thickness of a man's fore-finger; which they fixed to the crown of their head. ..

He is clearly describing colurnella beads (see below). These short passages give us a clue about the great worth placed on shells, and even pieces of shells. A good deal of effort went into the procurement of a large bead made from one of these marine shells. We have no hard evidence, and can only speculate on what might have been exchanged for whole marine shells.

Cup-s

Ethnohistoric accounts

The ethnohistoric accounts of James Adair, William Bartramn, Jonathan

Dickinson, and a ranger for General Og-lethorpe all attest to the use of marine shell cups as containers for a tea known as cassina, a caffeinated beverage. John Swanton (1928b:503) learned from Creek Chief James Islands that "large conch shells" were still used to take "black drink" in the early 20th century. Adair (1971:49 [17751) describes them being retrieved: "consecrated, large conch-shells, out of a place appropriated for containing the holy things," and as containers for cassina (Adair (197 1:25 [1775]). Dickinson describes the process of making cassina on the east coast of Florida (1985:25 [1699]):

In one part of this house where the fire was kept, was an Indian man, having a pot on the fire wherein he was making a drink of the leaves of a shrub (which we understood afterwards by the Spaniard, is called casseena), boiling the said leaves, after they had
parched them in a pot; then with a gourd having a long neck and at the top of it a
small hole which the top of one's finger could cover, and at the side of it a round hole of two inches diameter, they take the liquor out of the pot and put it into a deep round
bowl, which being almost filled containeth nigh three gallons. With this gourd they
brew the liquor and make it froth very much. It looketh of a deep brown color, In the brewing of this liquor was this noise made which we thought strange; for the pressing of this gourd gently down into the liquor, and the air which it contained being forced





18


out of the little hole at top occasioned a sound; and according to the time and motion
given would be various. This drink when made, and cooled to sup, was in a conchshell first carried to the Casseekey...

A ranger for General Oglethorpe, among the Creek in 1739-1742 (Mereness 1916:220) also writes of marine shells used as containers for cassina:

The head Warriours of the Indians brought us black Drink in Conkshells which they
presented to us and as we were drinking they kept Hooping and Hallowing as a Token of Gladness in seeing us. This Drink is made of a Leaf Called by the English Casena
(and much Resembles the Leaf of Bohea Tea.) It is very Plenty in this country.

William Bartram relates the events at an important ceremony (1976:358 [1789]): "The assembly being now seated, and the house illuminated, two middle aged men,.., come in together at the door, each having very large conch shells full of black drink, and advance..."

Cas sina is well-known to have been ingested for purposes of purification

(Fairbanks 1979). Merrill (1979) gives a very detailed description of the use of cassina among Southeastern groups. It was also poured on the body (Howard 1968:76). The main ingredient of cassina was leaves from the yaupon holly plant (flex vomitoria). This plant is now commonly used as an ornamental shrub in Florida. Swan (in Schoolcraft 1855) states that "The black-drink is a strong decoction of the shrub well known in the Carolinas by the name of Cassina, or the Uupon Tea. . they style it white-drink; but the liquor of itself, which, if strong, is nearly as black as molasses, is by the white people universally called black-drink" (emphasis in the original). Cassina, or the white drink, also called asi, was called the black drink by Europeans (Boyd 1955:25 1; Claiborne 1964:518 [1880]; Gatschet 1969:174, 176; Stiggins in Swanton 1928b:541; Sturtevant 1955:208; see also Swanton 1928b:577, 582, 588, 298, 603, 606, 608). These separate









accounts indicate that cassina was thought of as the white drink among Southeastern peoples (Hudson 1976:226). The reader is referred to Hudson's (1979) excellent book for a description of cassina.

An important aspect of shell cups is the social context in which they were used. The most notable ceremony among pre-contact Southeastern people is the busk, or green corn ceremony, during which cassina was administered, both internally and externally (Hudson 1979). This has been illustrated by Jacques Le Moyne, an artist who was a companion of Ren6 de Laudonni~re in his expedition to Florida in 1564. The text accompanying the illustration said that cassina was drunk in a large shell (Lorant 1946:93). An early 1 800s French account (Claiborne 1964 [1880]) described a funeral custom in which mourners drank "a beverage called the white drink." Another aspect of the funeral tradition was also illustrated by Le Moyne (in Lorant 1946:73, 115), in which shell drinking cups were placed over the graves of the dead owners. Shell cups were used in drinking cassina, however gourd calabashes were also used (Claiborne 1964 [1880]; Milford 1956:139 [1802]; Taitt, in Mereness 1916:502-503). Archaeological accounts

The majority of marine shell cups were made from lightning whelk shells,

Busycon sinistrum (Brown 1996:4 17; Milanich 1979:85; Phillips and Brown 1978:26), and a few were made from helmet and horse conch shells. Cups were made from the outer whorl of a shell; the colurnella and inner whorls were either cut out or hammered out (Phillips and Brown 1978:28). Some cups were engraved and others were plain. A shell cup made from a helmet shell was found at Etowah (van der Schalie and Parmalee





20


1960:44), and al] of the others at Etowah were made from Busycon whelks (Larson 1971:63). None of the shell cups from Etowah were engraved. The greatest quantity of cups were excavated from the Spiro site. At least two marine shell cups have been found from Moundville (Moore 1905:161).

By far, the most cups from any Southeastern, site were found at the Spiro site. At least 1000 engraved shell cups were excavated from Spiro, and about one or two thousand plain shell cups also were found (Brown 1982:469). The sheer numbers of whelk shell cups at Spiro indicate an enormous investment of resources in the acquisition of lightning whelk shells. Add on to that the time and effort it took to engrave so many shells, and we can see that the shell cup artifacts represent a great amount of wealth and/or status. Also at Spiro, most of the shell cups were deliberately broken before interment in the graves (Brown 1982:479). Ottesen (1979:397) tabulated the number of "marine shell vessels", and found that 254 from the Spiro site were made from conch shells of the genus Strombus. However, this data set is probably in error, since no mention was made of Siroinbus vessels in James A. Brown's most recent and complete work on Spiro (1996), nor in any of his previous works. Brown states that "small numbers" of horse conch cups were found, and one was made from a helmet shell (1996:417), in contrast to Phillips and Brown (1978:27), who counted three unengraved helmet shell cups.

No discussion of marine shell cups is complete without mentioning the seminal work of Philip Phillips and James A. Brown (1978). They discussed motifs engraved on marine shell from the Spiro site, and sketched out the history of Spiro site excavations.






21


They included 550 engraved cups, and 115 engraved gorgets in their sample. Only three engraved cup fragments from horse conch shells were reported.

Ceramic effigies of lightning whelk shell cups have also been found at

archaeological sites. Shell cup effigies were found at the Sawmnill Mound and at the Florence Street cemetery from Cahokia (Emerson et al. 1983; Milner 1984:478). They are also found at sites in Missouri (Chapman 1980), Florida (Moore, C. 1922), Arkansas (Holmes 1886:384), and at the Moundville site (Peebles 1979; Steponaitis 1983b-:76). Effigies of freshwater clamshells were found at Moundville (Peebles 1979), as well as other effigy pots of animals such as birds, turtles, fishes, hats, and frogs (Moore 1905; Peebles 197 1). Lightning whelk shell cups are the only exotic object that ceramic effigies were patterned after. This fact signals their cultural importance. Gorgets

Ethnohistoric accounts

Shell gorgets were also used in historic times. Bartramn (1976:299 [1791]), in describing participants in a contest called the ball game, tells us that the "champions likewise were well dressed, painted, and ornamented with silver bracelets, gorgets and wampum..." An early 1700's description of shell gorgets comes from Dumont de Montigny (in Swanton 1946:486), who was familiar with southern. Louisiana:

The savages also wear on their necks plates 3 or 4 inches in diameter made of pieces of this shell, to which they give a round or oval shape by grinding them on stones in
the same manner. They then pierce them near the edge by means of fire and use them
as ornaments.

Hawkins (1938:44 [1848]) describes Creek place names in the late 1700's and early 1800's, and gives the name of one town as Im-mook-fau. Its translation is "a gorget made






22


of a conch." Could imn-mook-fau be the indigenous Creek word for what we call a gorget? Linguistic research may help to clarify the definition of this word. Archaeological accounts

Some shell gorges from archaeological sites are elaborately engraved, while

other are plain. Some shell gorgets are not engraved. Brain and Phillips (1996) provided a thorough description of the motifs. Engraved shell gorgets and other shell artifacts (and the designs on them) from Tennessee were studied in detail by Kneberg (1959). Gorgets are usually engraved on the concave side of the shell, except for shell mask gorgets (see below) and the Cox Mound style of gorget with a bird motif (Brain and Phillips 1996:9). The species of snail shell from which gorgets were made has rarely, if ever, been identified, due to the fragmentary nature of the remains.

Shell masks are usually subsumed under the gorget category, and are called shell mask gorges. They occur from North Dakota to Virginia, and from Ohio to Alabama (Smith and Smith 1989:11). The majority lack engraving, and most are engraved on the convex side of the piece of shell. Shell masks are usually found with male burials (Smith and Smith 1989).

Shell gorgets are made from the outer whorl of a shell, finished in a circular or oval shape. Phillips and Brown (1978) included 115 engraved shell gorgets from the Spiro site. The later work of Brain and Phillips (1996) showed that at least 125 have been found from Spiro. At least eight gorgets were excavated from the Moundville site (Brain and Phillips 1996), and these were found only in the cemetery areas (Peebles 1971:81). Sixty-three gorgets were excavated from the Etowah site, and most were






23


engraved (Brain and Phillips 1996). Only one engraved shell gorget was excavated from Cahokia site (Brain and Phillips 1996), despite the abundance in marine shell-working debitage there (Baker 1929; Mason and Perino 196 1:554). Pendants

Shell pendants have been found at archaeological sites, as well as depictions of them on engraved whelk shell cups and on copper plates (Brown 1982:470; Fundaburk and Foreman 1957). Images of sinistral whelk pendants can sometimes be determined on engraved artifacts from sites in Tennessee, Alabama, Georgia, Oklahoma, and Illinois (Kneberg 1959; Phillips and Brown 1978; Waring and Holder 1945:12). Most shell pendants are made from lightning whelk shells. Depictions of shell pendants are found in the "bird-man theme" (Strong 1989), as well as with "god-animal" representations (Waring and Holder 1945:20),

Pendants have been recovered from all of the four archaeological sites scrutinized here. Many shell pendants were found at the Spiro site (Brown 1996:59 1), though specimen counts vary- Hamilton (1952:56) estimates that about 500 lightning whelk pendants were removed from Spiro previous to the controlled excavations. Ottesen (1979:394) lists 45 marine shell pendants from Spiro. Sixteen pendants were excavated from the Cahokia site (Ottesen 1979:394), some of which were illustrated by Titterington (1977:33). An unspecified number of columella pendants were also found at the Etowah site (Kelly and Larson 1957:44). C. B. Moore (1905:180) found some marine shell pendants at Moundville (Larson 197 1:62).





24


Marine Shell Beads

The study of shell beads from archaeological sites is a subdiscipline that has recently expanded. Lynn Ceci has declared that the study of shell beads within anthropology is now a valid field (Ceci 1989:2). William H. Holmes (1883:219) defined three basic categories of shell beads: 1.) small shells with holes for suspension; 2) disk beads made from bivalves or the outer whorls of snail shells; and 3) marine shell columella beads, known as massive or barrel beads. Great quantities of shell beads have been found in some locations. Uncontrolled excavations at Spiro unearthed between 800 and 1,200 pounds of unspecified shell beads (Hamilton 1952). Each bead type will be discussed separately, in addition to a discussion of chert microdrills, which are believed to have been used to make holes in the beads. Chert microdrills

Chert microdrills used for drilling holes in shell to make beads have been studied most extensively by Richard W. Yerkes (Yerkes 1983, 1991, 1993). Yerkes approached the subject of craft specialization using the production of chert microdrills as evidence for shell bead workshops. First reported by Mason and Perino (1961:554) from the Cahokia site, microdrills were found in association with shell working debitage in the form of "thousands of shell beads and enormous quantities of burned conch columellas and shell scrap." The process of producing chert microdrills involves breaking chert nodules so that smaller pieces can be worked into microdrills or microblades (Yerkes 1991). Yerkes used experimental archaeology and incident light microscopy to confirm that microdrills were used to drill shell (1983).






25


Shell bead production areas, inferred by concentrations of microdrills, have been found at Cahokia and surrounding sites in the American Bottom (Trubitt 1996; Yerkes 1991), Moundville (Peebles and Kus 1977:442), sites in central Tennessee, northern Alabama, western North Carolina, northeastern Georgia, and the Zebree site in northeast Arkansas (Yerkes 1993:237). In spite of the large quantities of shell beads from Spiro, no shell bead workshop in the form of either shell debitage or microdrills has been identified there. Microdrills are solely a Mississippian phenomenon, since they have not been found in contexts dating to earlier time periods (Yerkes 1993:240).

An alternative method of piercing disk beads is by the use of fire. High heat

burns off the organic constituents of shell, leaving a more brittle material (see Chapter 4). This is supported by the ethnohistoric account of Garcilaso de la Vega (1988:330 [1605]), who noted that pearls were made into beads by piercing them, "with hot copper needles."

Whole marine shell beads

Various kinds of smaller, whole marine shells used as ornaments in the Southeast include olive (Oliva spp), dwarf olive (Olivella spp. or Jaspidella spp.), and marginella (Marginella spp.). The only alteration made to these small shells is a hole in the spire for stringing. Some freshwater snail shells were used as well, such as rocksnails (Anculosa spp. and Campeloma spp.) (Trubitt 1996:70).

Between 57,000 and 72,000 marginella beads were found at Cahokia, in

association with burials, and apparently sewn onto bark cloth (Milner 1984:480; Ottesen 1979:398). Some of these are pictured by Titterington (1977:40). The marginella beads






26


from Powell mound (see Chapter 4) were apparently sewn onto bark cloth, since they were recovered in the matrix in neat rows (Ahler and DePuydt 1987). According to Baker (1932:149), Etowah yielded four olive shell (0/iva) beads, and about two dozen marginella shell beads. At least 2,069 marginella beads were excavated from Spiro, (Brown 1996:283). Olive shell beads of the genus Olive//a were numerous at Spiro, totaling 14,937, and olive shell beads of the genus 0/iva total seven (Brown 1996:283). Disk beads

Disk beads are recognized by their thinness in relation to the diameter of the disk. The hole runs perpendicular to the growth lines of the shell, since they were usually made from the outer whorl of a marine snail shell. James Brown (1996:577) believed that disk shell beads are the most common marine shell bead at Mississippian sites. At least 19,835 disk shell beads were recovered from controlled excavations at Spiro (Brown 1996:283). These are made from an unidentified shell, but are assumed to have been made from whelk shells (Brown 1996:577). A high status burial from Mound 72 at Cahokia (see Chapter 4) rested on "a platform of shell beads" (Fowler 1991:3), which consisted of 20,000 disc and columella beads (Iseminger 1996:34). Moore (1905:147) reported that "many shell beads" were found in association with a burial in Mound C at Moundville; these could be disk beads or columella beads. Columnella beads

Beads made from the columellas of marine shells have many forms and have been called tubular, barrel, massive, and a host of cylindrical and spherical terms (Brown 1996; Holmes 1883:223; Moore 1905:154; Ottesen 1979:377). The important defining






27


characteristic is that the hole in these larger beads runs parallel to the growth lines of the columella. Unfortunately, it is difficult to determine the species from which columella shell beads were made. Despite the usual difficulty, sometimes identification of the snail genus from which columella beads were made is possible. Holmes (1883:223) was the first to note this, stating that the beads "often retain the spiral groove as well as other portions of the natural surface." This spiral (or columella) groove slants toward the left or right, depending on the sinistral or dextral shell from which it came. The closest readily recognizable analogy is either the "back slash" or "forward slash" of today's computer keyboards. Sinistral shells have columella grooves with a back slash, while dextral shells have columella grooves with forward slash. Figure 2.1 illustrates how some columella beads retain the columella groove, and why the direction of the slant is important in identification of a sinistral or dextral shell. The direction of the slant does not change when the bead is turned 1800. The only shells in the Atlantic Ocean or Gulf of Mexico from which large sinistral columella beads can be made are of the genus Busycon.

Columella beads have also been recovered in great numbers. At least 43,277 beads made from marine shell columellas were found at Spiro (Brown 1996:283). As noted above, the "principle figure" from Mound 72 at Cahokia was interred on a platform of 20,000 marine shell beads of an unspecified type (Fowler 1991; Iseminger 1996). Another 30,000 disc beads were also buried in Mound 72, Cahokia (Trubitt 1996:266). "Many" tubular shell beads were found in Mound C at Moundville (Baker 1932; Moore 1905:154).





28







0

































Figure 2.1 Bead reduction from sinistral whelk shell. Afier Holmes (1883)





29


Miscellaneous

Other kinds of marine shell artifacts occur in far less numbers from Mississippian sites. These include scallop shells (Pectinidae), sunray venus shells (Macrocallista nimbosa), Atlantic cockle shells (Dinocardium spp.), ark shells (Arca spp.), rangea clam shells (Rangia spp.), helmet shells (Cassis spp.), top shells (Cittarium spp.), horse conch shells (Pleuroploca gigantea), knobbed whelk shells (Busycon carica), channeled whelk shells (Busycotypus canaliculatus), moon shells (Neverita spp.), nutmeg shells (Cancellaria reticulata), dove shells (Columbella spp.), marsh periwinkle shells (Littorina irrorata), scotch bonnet shells (Phalium granulatum), murex shells (Phyllonotus spp.), rock shells (Thais spp.), bubble shells (Bulla spp.), fighting conch shells (Strombus alatus), cone shells (Conus spp.), and tulip shells (Fasciolaria spp.) (Baker 1923,1932:148; Parmalee 1958; Trubitt 1996:86; van der Schalie and Parmalee 1960). The suite of marine molluscan species recovered from Mississippian sites indicates they came from the Gulf of M6xico or Atlantic Ocean (Abbott and Morris 1995), except Busycon carica and Busvcotypus canaliculatus, both of which only occur in the Atlantic Ocean (Abbott and Morris 1995).














CHAPTER 3
MALACOLOGY AND SHELL AVAILABILITY



A thorough knowledge of the shells available is an important step in

understanding shell choice for artifact manufacture. One of the ways to achieve a sense of shell availability in the past is to find out what kinds of shells are currently obtainable. This approach is most applicable for studying shell collections from the recent past. It is also necessary to understand the ecology and habits of the animals being studied. The Florida Museum of Natural History (FLMNH) Malacology Collection houses a large assemblage of specimens, both wet-preserved and dry-preserved. These collections are from many parts of the world, and are considered to be among the finest in the country. I examined shells from 1 3 ) species of marine snails in this collection. My choice of species to examine is discussed below. Brief descriptions of the geographic range and preferred habitats are also provided.

One of the problems encountered while addressing issues of a zoological nature is scientific nomenclature. Disagreements among zoologists concerning the delineation of a species or a subspecies often occur. Additionally, the specific name, generic name, and sometimes the family or order nomenclature require revisions. This sometimes leads to confusion among non-zoologists (or even zoologists themselves). Frequent name changes have hampered attempts by archaeologists to determine which animal's shell



30






31


was used for artifact manufacture. The nomenclature used throughout this dissertation follows the taxonomic scheme presented by the Council of Systemic Malacologists, American Malacological Union, in Common and Scientific Names ofAquatic Invertebrates from the United States and Canada: Mollusks (Turgeon et al. 1988).

The lightning whelk is a good example of the potential confusion caused by

changes in nomenclature. The scientific name for the lightning whelk has changed three times, from Busycon perversum (Linn6 1758) to Busycon contrarium (Conrad 1840) to Busycon sinistrum (Hollister 1958). Some malacologists think that all sinistral whelks should be called by one species name (Abbott 1974:223; Pulley 1959:75), and others think that several sinistral whelks in the family Melongenidae represent different species. Currently, three species of sinistral whelks are recognized by scientists (Turgeon et al. 1988). These are the snow whelk (Busycon laeostomum Kent 1982), the lightning whelk (Busycon sinistrum Hollister 1958), and the prickly whelk (Busycon pulleyi Hollister 1958). These species are believed to be capable of interbreeding, but are recognized as different species based on their geographic separation (allopatric speciation). The difficulties inherent in this nomenclature may have discouraged some archaeologists from assigning scientific names to some shell artifacts.

The topic of snail handedness has implications for both zoological systematics and archaeology. Handedness refers to the coiling direction of the snail, dextral (righthanded) or sinistral (left-handed). Dextral means that with the apex pointing upward, the opening of the shell (aperture) will be on the right side. Most species are normally dextral (Gould 1995; Robertson 1993; Vermeij 1975). Even gastropods without shells






32


(slugs, for example) are usually dextral (Robertson 1993:115). Occasionally, an individual of a species that is normally dextral may have a sinistral shell, and vice versa. Malacologists do not know why the overwhelming majority of species have dextral shells.

Only two genera of marine snails in the western Atlantic Ocean, Gulf of Mexico and Caribbean Sea are normally sinistral. These are Triphora and three currently recognized whelk species of the genus Busycon (Abbott and Morris 1995; Rehder 1996). Triphoras are tiny (6 mm maximum) and live in shallow water on or near sponges, on which they feed (Rehder 1996:446). Triphoras look like ceriths, and are closely related to them. Lightning whelks are easily identified because they are sinistral, and are discussed in detail below.


Materials and Methods


Most of the artifacts were made of large shells, such as lightning whelks. In order to expedite the analysis of shells in the Malacology Collections, I examined only the species that are able to attain a length greater than 17.78 cm (7 inches). Since marine bivalves are rarely, if at all, present in the target archaeological sites, I did not investigate them. Only thirteen marine snail species in the western Atlantic coast and Gulf of Mexico reach at least 17.78 cm in length. According to Abbott (1968; 1974), and Hollister (1958), the following thirteen species fit this parameter:

Strombidae
Queen conch (Sirombus gigas); to 30.5 cm
Cassidae
Cameo helmet (Cassis madagascariensis); to 23 cm
Caribbean helmet (Cassis tuberose); to 23 cm








Ranellidae
Atlantic trumpet triton (Charonia tritonis variegata); to 33 cm
Angular triton (Cymatiumfeminorale); to 21 cm
Tonnidae
Giant tun (Tonna galea); to 18 cm
Melongenidae
Splendid whelk (Busycon candelabrum); to 23 cm
Knobbed whelk (Busycon carica); to 36 cm
Prickly whelk (Busycon pulleyi); to 21 cm
Lightning whelk (Busycon sinistrumn); to 41 cm
Channeled whelk (Busycotypus canaliculatus); to 19 cm
Fasciolariidae
True tulip (Fasciolaria tulipa); to 19 cm
Horse conch (Pleuroploca gigantea); to 58 cm

It is possible that collecting biases of the malacologists have skewed the sample, because large shells require large spaces for storage, large specimens are sometimes not collected. Other biases may be collection of larger specimens by amateurs, a concentration of specimens from a favorite collecting location, or a lack of specimens from another collection. The FLMNH Malacology collection focuses on Florida land and freshwater mollusks, but collections from the Gulf of Mexico, the Atlantic Ocean, and the Caribbean Sea are well-represented (Fred G. Thompson, personal communication). Ecological information on each of the species examined is presented below.

Dry-preserved shells from thirteen different marine snail species were examined. Total length of 1428 specimens was measured (in centimeters). Locality information was also recorded, consisting of country, state, county, town, etc. Specimens were not included that did not have locality data. The depth at which the specimen was found was also recorded, since depth may be an important consideration in shell choice. Depth data was not regularly incorporated into the Malacology database, but malacologists have a





34


general idea of the depths at which various species are found (Abbott 1974). 1 input all of the data into a relational computer database (Microsoft Access).

Figure 3.1 illustrates the terminology concerning different areas of a shell. Some definitions (after Abbott 1968:51) are as follows:

aperture the opening in the last whorl
apex the first-formed, narrow end of the shell, usually of several whorls
columella the solid pillar at the axis of the shell, around which the whorls grow
lip final edge or margin of the body whorl
nodule a round growth on the spire
outer whorl the last and largest coil of a snail shell
shoulder part of the whorl projecting or bulging farthest from the axis
siphonal canal anterior, semitubular, extension of the aperture
spine a sharp, elongated nodule
spire the whorls at the apical end, exclusive of the last whorl Results


Table 3.1 summarizes the information obtained from the Malacology Collection. Appendix I presents all of the data. Size range information is included in order to provide a better comprehension of the completeness of the Malacology collections. As indicated, a large size range for all of the species was in the FLMNH. As might be expected, the largest shells are from the horse conch, Pleuroploca gigantea, since this is the largest snail in the Atlantic Ocean and the Gulf of Mexico. Only specimens of three species show appreciable numbers (over 10%) of shells greater than 20 cm total length; the queen conch (Strombus gigas), the lightning whelk (Busycon sinistrum), and the horse conch (Pleuroploca gigantea). Only lightning whelks and horse conchs have shells larger than 30 cm. Although queen conchs (Strombus gigas), trumpet tritons (Charonia




35


apex nodule spire


lip ) I shoulder


outer
whorl aperture








columella siphonal
canal





Figure 3.1. Shell terminology





36


TABLE 3.1

MALACOLOGY DATABASE SUMMARY

Size Total >20cm >30cm
Species Range (cm) Number Number Number

Strombusgigas 0.5-27.6 231 16% 78 37% 0
Cassis madagascariensis 4.8-29.3 39 3% 13 6% 0
Cassis tuberosa 1.1-27.4 77 5% 3 1% 0
Cymnatium fimorale 2.0-17.3 86 6% 0 0
Charonia variegata 0.8-28.1 61 4% 10 5% 0
Tonna galea 0.9-21.6 116 8% 1 0% 0
Busycon carica 1.7-22.5 131 9% 2 1% 0
Busycon candelabrum 4.3-17.0 15 1% 0 0
Busycon sinistrwnum 1.4-35.9 348 24% 50 23% 7 17%
Busycon pulleyi 3.8-25.1 16 1% 3 1% 0
Busycotypus canaliculatus 2.3-19.9 82 6% 0 0
Pleuroplocagigantea 0.6-49.1 225 16% 53 25% 35 83%

TOTALS 1427 213 42





37


variegala), and knobbed whelks (Busycon carica) attain lengths greater than 30 cm, no specimens of this size is included in the Malacology Collections.


Discussion


The data from the FLMNH Malacology collection show that specimens come from all along the Atlantic seaboard, the Caribbean, and the Gulf of Mexico (Appendix 1). No collecting locations appear to have been overlooked. My study corroborates previous observations of malacologists (see below).

The most important information that my investigation of shells from the FLMNH collection adds to existing knowledge concerns the range in sizes and collection localities of those Species with large shells. If size was the major factor in choice of shells, then queen conchs (Strombus gigas) or particularly horse conchs (Pleuroploca gigantea) would have been harvested for artifact manufacture. Queen conchs contribute 37% of the sample of shells getting larger than 20 cm, while lightning whelks represent 23%. Horse conchs account for 83% of all shells greater than 30 cm, with lightning whelks accounting for the remainder. If shell choice were simply a matter of gathering the largest shells, one would expect to find similar proportions of shells among archaeological artifacts.

Water depth is an important factor to consider when shell availability is being addressed. The splendid whelk (Busycon candelabrum) seems to occur regularly at depths that only scuba divers would normally attempt to reach (Appendix I). Larger lightning whelks are available in shallow water off the Florida coast, but, interestingly,






38


are not common from other areas. Only four of the lightning whelks in my study larger than 17.78 cm are not from Florida waters (Appendix 1), and all of them are from deep water. This study confirms observations of other workers that lightning whelks are most common from the west coast of Florida. Large prickly whelks in my sample also seem to occur in deeper water, in agreement with comments by Abbott (1974:222-223). Both of the helmet snails (Cassis spp.) included in my study also seem to be more prevalent in deeper, tropical waters. The trumpet triton and the angular triton are also species that usually require a dive to obtain. Only one angular triton in my study was larger than 17 cm. Giant tun shells were regularly found at depths of greater than 20 meters. Horse conchs are available from shallow water (less than I meter).

In the following discussion, particular attention is paid to the sinistral whelks, since it has long been known that these shells were commonly used for artifact manufacture (Milanich 1979). The descriptions are arranged first by whelk snails, followed by information on other snails. Data are from Abbott (1974), Rehder (1996), and Abbott and Morris (1995) unless otherwise noted. Whelk Snails

An excellent ecological paper by Hulda Magalhaes (1948) discusses whelk snails living near Beaufort, North Carolina. She showed that in her study area, the ratio of lightning whelks to channeled whelks to knobbed whelks is 1:2:33. Lightning whelks in this area live near the northern limit of their range, and are not thought to breed (Magalhaes 1948:406). Also, lightning whelks do not exceed 26 cm total length and average 14 cm in this area. Whelk shells exhibit characteristic breakage patterns after





39


the live snail has been eaten by a stone crab (Menippe mercenaria). In order to get at the meat, a stone crab breaks the outer edge of the siphonal canal and aperture, the body whorl is broken, and sometimes the columella is broken off (Magalhaes 1948:400). Sea gulls (Larus spp.) are other predators of small snails. They expose the soft body parts by dropping the snails from the air onto a hard surface (Magalhaes 1948:401). All whelks are carnivores. Both Magalhaes (1948:399) and Paine (1962) showed that species of Busycon prey on at least 21 species of bivalves. Age and growth studies have not been attempted, and Magalhaes (1948:397) claimed that these types of studies "cannot be applied with any accuracy to Busycons." Whelks are mobile, migrating both seasonally and diurnally; seeking prey, and avoiding predators, excessive heat, cold, or dehydration. The maximum distance covered in one day is 40 meters (Magalhaes 1948:394). Sinistral whelks

Nomenclature for sinistral whelks continues to be in a state of flux, and has been since the time of Linnaeus. Genetic studies may resolve their taxonomic affiliation. More detailed descriptions of the shell characteristics, ecology, and distribution of sinistral whelks are presented below (see Figures 3.2 and 3.3). All of the sinistral whelks of the genus Busycon have been placed into a separate subgenus, Sinistrofulgur (Hollister 1958:84). This is based on the sinistral shell alone, in order that other characters used to place dextral whelks into specific groups may also be used for sinistral whelks.

Hollister recognized four species of Sin istrofulgur; B. aspinosum offshore from Sarasota, FL, B. perversum from Campeche Bay, Mexico, B. pulleyi from the north Mexican coast to Louisiana, and B. sinistrum from Cape Hatteras, NC to west Florida.





40


Busycon aspinosumin and B. perversum are not currently recognized. The prickly whelk (Busycon pulleyi) breeds with the lightning whelk (Busycon sinistrum) in the Gulf of Mexico where their ranges overlap "giving rise to the possibility that this complex is one species, namely perversum Linne" (Abbott 1974:223).

Busycon sinistrum Hollister 1958. Historically, the nomenclature has changed frequently for this animal. The synonomy of Busycon sinistrum is as follows (after Hollister 1958:85): Pyrula perversa (Linn6) Lamarck 1816, Fulgurperversus Gill 1867, Busycon perversumn Chenu 1859, Fulgur perversus "Linn6" Tryon 1881, Fulgur perversum "Linn6" Dall 1890, Busycon perversum "Linn6" Johnson 1934, and Busycon contrarium (Conrad) B. Smith 1939. Hollister thought that Murexperversa Linn6 1758 was synonymous with Busycon perversum (see above), rather than with B. sinistrum.

The lightning whelk is most common on the west Florida coast (Abbott 1974:

222; Magalhaes 1948:406), and ranges from Cape Hatteras to Mobile, Alabama, and also occurs off the Yucatan coast. It has not been reported from Campeche Bay to Mobile Bay (Hollister 1958). Lightning whelks are found in warmer water than knobbed whelks (Magalhaes 1948), and seem to migrate to shallow water in the winter months in Florida (Paine 1963:64). A gracile form with smooth, round shoulders is found today around Longboat Key, Sarasota, Florida. Specimens have been found off Yucatan that have large spines curving toward the aperture. One dextral specimen is known from Fort Myers Beach, Florida (Hollister 1958:87), and another from Sanibel Island, Florida (Shiflett 1972). The foot and mantle are almost always deep black except in young and





41

















Busycon sinistrum Busycon laeostomum
Campeche Bay, Mex. Avalon, NJ
258mm total length


Busycon sinistrum
Dry Tortugas, FL 285mm total length













Busycon pulleyi Busycon perversum
Corpus Christi, TX Campeche Bay, Mex.
98mm total length (juv.)






Figure 3.2. Variants of sinistral whelk shells.





4 2


Busycon
laeostomum


















Busycon
Pulleyi


Busycon
sinistrum










Figure 3.3. Sinistral Nvhelk species geographic ranges.





43


very old individuals (Hollister 1958:86; Magalhaes 1948:403). Young shells are striped parallel with the growth lines, with white and reddish brown to olive stripes edged with yellow on one side. Shells become lighter with increasing age, very old shells being white. The angle of the apex is about 105' (Hollister 1958:85). In Alligator Harbor, Florida, lightning whelks "appear to be highly mobile" (Paine 1962:517). This species prefers to eat bivalves.

Busycon pulleyi Hollister 1958. Former species names for the prickly whelk (in Hollister 1958:85) include Murexperversus Linn6 (1758), Pyrulaperversa Kiener (1840), Pyrula Kieneri Philippi (1848), Bulgur gibbosum Conrad (1853), Busycon gibbosum Conrad (1862), Busycon kieneri Philippi (1867), Busycon perversum kieneri Philippi (1934), and Busyconperversum Linn6 (1939). The geographic distribution of this species is in the Gulf of Mexico from Louisiana and to Campeche Bay Mexico. The spire is somewhat turreted, and the columella is sinuous (Hollister 1958:89-90). Abbott (1974:222-223) reports that they are dredged from 7.3 to 12.2 meters. The prickly whelk has a spire angle of about 90 degrees (Abbott 1974:222).

Busycon laeostomum Kent 1982. The snow whelk has only recently been named, and occurs only off the coast from New Jersey to northern Virginia, offshore (Kent 1982). Abbott (1974:222) was aware of the existence of these rare, deeper-water specimens (gathered by scuba divers), but did not consider them to constitute a distinct species. The snow whelk grows to be large, at least 25.8 cm. The apex angle ranges from 920 to 1260, with an average of 1050 (Kent 1982:10 1). The head and foot are black.





44


Dextral whelks

Busycon carica (Gmelin 1791). The knobbed whelk ranges from Massachusetts to Cape Canaveral, Florida in shallow sandy water. It lives in very shallow water to at least 5 meters. The foot and mantle of knobbed whelks are gray to almost black (Magalhaes 1948:403).

Busycon candelabrum (Lamarck 1816). The splendid whelk occurs at depths from 37 to 64 meters and ranges in the Gulf of Mexico from Texas to Yucatan, Mexico. Scientists did not even know of the existence of dextral species of the genus Busycon in the Gulf of Mexico until the 1950's when, "shrimping activities in Yucatan waters brought to light a number of shells that were thought to be new" (Hollister 1958:79). The depth at which it is commonly found reduces the chances of being captured for artifact making.

Busycotypus canaliculatus (Linn6 1758). The channeled whelk is common in shallow sandy areas from Massachusetts to the Atlantic coast of eastern Florida. This species occurs on sand or mud intertidally to just below the low-tide level. It is abundant in the shallow bays of southern New England. It barely reaches the minimum of 17.78 cm to be included in my study. The shell is thinner than other whelk species, and the spire is turreted (Magalhaes 1948:385). The foot and mantle are pale grayish white. Dextral Marine Snails

Strombus gigas (Linn6 1758). The queen conch, also known as the pink conch, ranges from the West Indies to Southeastern Florida and the Florida Keys. It only rarely occurs on the southwest coast of Florida, and can be found in one to 12.2 meters of





45


water, usually near eel-grass. This shell was commonly used for artifact manufacture in the Caribbean before Europeans arrived, particularly from the columella and outer lip of the shell (Keegan 1984). Conch meat is a popular ingredient in stews and chowders where it occurs. Queen conchs eat algae.

Cassis tuberosa (Linn6 1758). The Caribbean helmet ranges from North Carolina to Brazil in sandy shallow water. It is rare in Florida and northward, but common in the Caribbean Sea. Both of the helmet snails listed here prey on sea urchins.

Cassis madagascariensis (Lamarck 1816). The cameo helmet ranges from

North Carolina to the West Indies, and grows larger than the Caribbean helmet. It can be found on sand in shallow water to nine meters deep, and is rare in Florida.

Cymatiumfemorale (Linn6 1758). The angular triton is also a shallow water

species, ranging from southeast Florida to Brazil. It is most commonly found in the West Indies in eel-grass, and is only rarely found in the lower Florida Keys.

Charonia tritonis variegata (Lamarck 1816). The Atlantic trumpet triton can be found in moderately shallow water from southeast Florida to Brazil. It is rare in Florida and moderately common in the West Indies near reefs.

Tonna galea (Linn6 1758. The giant tun ranges from North Carolina to Brazil. It lives in moderately deep water to 33 meters, and it is an uncommon species. The shell is thin, but sturdy.

Fasciolaria tulipa (Linn6 1758). The true tulip ranges from North Carolina to the West Indies and Brazil. It can be found on sand and mud, intertidally to nine meters






46


in depth. It just barely grows to be larger than my minimum requirement of 17.78 cm total length. Tulip snails are carnivorous, feeding upon other gastropods.

Pleuroploca gigantea (Kiener 1840). The horse conch ranges from North

Carolina to Florida and to the western Gulf of Mexico. It can be found from low-tide level to six meters deep. Horse conchs may migrate into shallow water in summer on the Florida coast (Paine 1963:65). This is the largest snail in its range, and one of the largest in the world. It is carnivorous, feeding mostly on marine gastropods (53 of 66 items ingested are gastropods) (Paine 1963:66). The brown penostracumn of these animals flakes off easily once the shell is dry, leaving behind a white shell. Implications for Archaeological Studies

What importance does all of this information have for archaeologists working with marine shell artifacts in eastern North America? There are several points that should be emphasized. One is that the knobbed whelk (Busycon carica) does not occur in the Gulf of Mexico. Therefore, any occurrence of the shell of this species at an inland archaeological site infers contact with people on the Atlantic coast. This species should not be confused with the splendid whelk (Busycon candelabrum), since the presence of a shell from the splendid whelk implies contact with the Gulf of Mexico.

A very important piece of information that may be obtained from archaeological sinistral whelk shell artifacts is the apex angle. This may be viewed as a distinguishing characteristic between lightning whelks and prickly whelks. Lightning whelks have an apex angle of about 105', while prickly whelks have an apex angle of about 90'. If there is a large enough sample size from an archaeological site, we can hypothesize that the






47


shells are from one or the other species. The apex angle of snow whelks is also about 105', but it is not likely that archaeological shell artifacts came from this species for the following reasons:

there have been few specimens found, only 13 as of 1982
it is not certain that the snow whelk is a distinct species, though it is recognized
by Turgeon et a]., 1988
they occur in deeper waters
they occur offshore

Depth is another factor to consider. Lightning whelks, queen conchs, knobbed whelks, and horse conchs may be easily found in shallow water. Prickly and snow whelks are found in deeper water than lightning whelks. Abbott (1974:222-223) states that the perverse whelk is dredged from 7.3 to 12.2 meters, making it hard to obtain.

The build of the shell is important. Some lightning whelk shells are extremely gracile, while those of the prickly whelk are much more robust. The colurnellas of lightning whelks are less heavy and sinuous than those of prickly whelks.

Archaeologists must also be aware that shells may sometimes show signs of regular breakage patterns that are not the result of human activities. This is more pertinent for archaeologists working at coastal sites. As mentioned above, animals such as stone crabs break whelk shells in a habitual manner (Magalhaes 1948:400). Other animals also make shell alterations in prey species. For example, octopuses sometimes make holes in marine snail shells in order to get at the soft tissues inside. These holes look very much like drill holes that humans may make. Archaeologists must first rule out any type of animal-generated breaks in shells before attributing them to humans.

Some archaeologists have assumed that marine shells were sought simply for their size. If it was true that size was the main factor (other things being equal), then





48


similar proportions of shells should be found at the target archaeological sites. If size was the only reason for the use of marine snail shells for artifacts, then horse conch shells would be the most numerous marine shell used since they are the largest. Table 3.1 shows that of all of the shells, horse conchs comprise 83% of all shells larger than 30 cm. Assuming that the specimens in the FLMNH collection reflect the availability of large marine snails, then horse conch shells are expected to comprise 25%, queen conchs 37%, and lightning whelks only 23% of the shells longer than 20 cm in the archaeological deposits. This is not the case, as the next chapter shows.














CHAPTER 4
ARCHAEOLOGICAL SAMPLES



Marine shell artifacts are abundant in the four archaeological sites chosen for investigation. A total of 17,645 shell artifacts, representing 20 different molluscan genera were examined from these four sites. Two of these sites, Cahokia and Spiro, have far larger samples than Etowah and Moundville. The contexts with which shell artifacts are associated are important, since interpretations of artifact function and type are sometimes dependent upon provenience. Descriptions of each of the four sites and excavation strategies provide information on the contexts with which shell artifacts are associated. This is followed by an account of the methods used here to identify the taxon and record data from archaeological specimens made of marine shell. The procedure for determining whether a shell gorget was made from a dextral or sinistral shell is also given. Next, I supply summary information of the data. Ratios of dextral versus sinistral columella beads and gorgets are tabulated, as well as identification of the species from which shell pendants and cups were made. An analysis of apex angles from sinistral whelk shell artifacts reveals that most are from lightning whdlks, rather than prickly whelks. Artifacts from the Cahokia site show that Mississippians probably used fire (at times) to remove the columella from the rest of the shell.






49






50


In order to obtain a large sample of marine shell artifacts, I chose sites that are known to have produced quantities of them. Additionally, my sample is restricted to Mississippian sites (800-1450 A.D.), since any hypotheses concerning ideology drawn from the ethnographic present become less viable as more time passes. Connections among the four sites (Cahokia, Etowah, Moundville, and Spiro) are indicated by exotic artfacts, described below. Spiro has long been known to have had large quantities of marine shells, in fact, more marine shell artifacts have been found there than at any other inland North American site. Some marine shell artifacts from Spiro originated on the Pacific coast. Debitage from a shell-bead production area at Cahokia also provided an opportunity to test whether columella beads were made from sinistral or dextral shells. A map of all the site locations is presented in Figure 4. 1. 1 had intended to include the Ocmulgee site as well. Unfortunately, the time constraints involved in obtaining proper authorization, from both the National Park Service and the Creek Nation in Oklahoma, to view the collections have prevented me from analyzing the Ocmulgee material for this dissertation. It is hoped that an investigation of Ocmulgee artifacts will be possible in the near future.


Site Descriptions


Cahokia

Cahokia is located in the American Bottom, an area defined as part of the

Mississippi River valley bounded by the mouth of the Illinois River and the mouth of the Kaskaskia River (Fowler 1969:1-5). The appellation Cahokia is presumably derived







51






Wabash




River














RRiver






4. r,,,,1 erI I














0 0 20Gulf Of Mlexico
Kivemters





















Figure 4.1. Site location map.





52


from a village of the Illinois tribe that lived nearby, which du Pratz called "Caouquias" (du Pratz 1774:300-301). The site straddles a county line; Madison County at the north end and St. Clair County to the south (Fowler 1989:8). Cahokia is the largest Mississippian site in North America. It has at least 120 mounds (Iseminger 1996:32), including Monks mound, the largest mound north of the Teotihuacdn site in Mexico. The Cahokia site, owned by the state of Illinois, now comprises the Cahokia Mounds State Historic Site. Cahokia has provided the materials for a wealth of archaeological research on the development of complex societies, population estimates, exchange of exotic goods, and craft specialization. The locations of most of the archaeological structures at Cahokia mentioned in the text can be seen in Figure 4.2. The mounds are arranged around many plazas (Fowler 1989:11). Monks mound is 30.5 meters (100 feet) high, and covers an area of land 316.1 by 240.8 meters (1037 by 790 feet) (Reed 1969). Cahokia covers about eight square kilometers (5 square miles). The site had been in use since Paleoindian times (8,000 B.C.), and occupation was continuous through the Mississippian period. It was abandoned by about 1400 A.D. Population estimates of Cahokia at its zenith, about 1100 A.D., range from 8,000 to 40,000 (Fowler 1989:7; Iseminger 1996), the most cited figure being from 30,000 to 40,000 residents (Milner 1990:11). An area known as "downtown Cahokia" is defined by an 83 hectare (200 acre) palisaded enclosure known as the "stockade" (Anderson 1969:89-99; Iseminger 1996:32). Remains of a shell bead storage facility with 1,960 disk beads were found 1.5 kilometers west of Monks mound (Titterington 1977:9). About 914.4 meters (3,000 feet) west of Monks mound in Tract 15A are the remains of a circular construction of wooden cedar







53





00


0
a 00


00 c n

0

E, 0



00 0 0


0~~ 0i'a00 00
0 C 0



00


0


oDc

0

ElE U0 5E
0C
_5 CDC
0 c0

0






Figure~~~~ ~ ~ ~ ~ ~ 4.2 Caol iemp fe olrI99fgr -)





54


posts known as an "American Woodhenge." This structure was about 125 meters in diameter (410 feet), and had alignments with the summer and winter solstices, and the equinoxes (Fowler 1989:32; Wittry 1969). The mound numbering system used in the following descriptions is that of Fowler (1989).

Cahokia was first systematically excavated by Warren K. Moorehead in the

1920's (1922, 1923, 1929). He focused on describing and mapping the mounds, although he also performed minor excavations, including Sawmill mound (#39), the James Ramey Mound (#33), and Kunnemann mound (#10 and 11). He found numerous burials, potsherds, bone artifacts, shell artifacts, and lithics. Ramey Field and Mound 34 Cahokia

In 1956 Gregory Perino, of the Gilcrease Foundation in Tulsa, OK, excavated portions of an area east of Monks mound known as the Ramey Field, and Mound 34 inside of Ramey Field (Fowler 1989; Mason and Perino 1961; Perino 1959). These areas may date from 1250 to 1500 A.D. (Brain and Phillips 1996:267), however Pauketat (1993:148) argued for earlier dates of 1150 to 1250 A.D. Mound 34 was a conical mound east of Monks mound, and has been mostly destroyed (Fowler 1989:88). Perino excavated a trench through Mound 34, as well as a refuse pit to the north of the mound. Sherds from pottery originating in the Caddo area, and some that came from the lower Mississippi River valley were found (Perino 1959). Lithics from Arkansas were also recovered (Fowler 1989:88; Perino 1959). Shell-working debitage and a few chert microdrills were found at Ramey Field (Mason and Perino 1961:554). Brain and Phillips





55


(1996:267) pointed out that all of the artifacts from Cahokia associated with the Southeastern Ceremonial Complex are associated with Ramey Field and Mound 34. Powell mound Cahokia

Powell mound was 2.6 kilometers west of Monks mound, about 13 meters high, and 94.5 by 54.9 meters at the base (Ahler and DePuydt 1987). It was due north of Mound 84 (now obliterated) which was due west of the southwest corner of Monks mound (Ahler and DePuydt 1987:4). In 1930, the owners of the Powell Mound (#86 of Fowler 1989:162) began demolishing it. Concomitant with this demolition, some interesting materials were recovered by Arthur R. Kelley of the University of Illinois, who witnessed the destruction (Fowler 1989:165; Titterington 1977:15). Chert microdrills at the base of the mound indicate a shell artifact workshop (Yerkes 1989:97). Powell mound was constructed in at least two stages, the first being a truncated pyramid mound, and the latter stage gave it the final appearance of a rounded linear ridge-top mound (Ahler and DePuydt 1987:3-4). Two burial pits were sandwiched between these two mound building stages. Although one burial pit had already been destroyed by the time Kelley got there, the other contained remains of 20 to 30 human burials, accompanied by "a very large number of disc shell or Marginella beads, but in no instance were both types of beads found with the same burial" (Titterington 1977:2). Some of these marginella beads, still embedded in the matrix in neat rows, are now curated at the University of Illinois Museum of Natural History, and the Lab of Anthropology, Department of Anthropology, University of Illinois, Urbana-Champaign.






56


Controlled excavations of Powell mound began in 1931 under the direction of

Thorne Deuel of the University of Chicago (Ahler and DePuydt 1987:5). Little remained of the mound, and a detailed analysis of the material remains was not published until 1987 (Ahler and DePuydt 1987:5). Ceramics indicate that Powell mound was constructed between 900 and 1150 A.D. (Ahler and DePuydt 1987). Much locally produced pottery was found, but exotic ceramics were also found. One sherd was tentatively identified as originating from the Caddo area, and two more (Nodena Whitefilmed and Red-and-White Painted) came from the lower Mississippi River valley (Abler and DePuydt 1987:23). Chert artifacts were recovered, 44 in the form of microdrills, indicating a shell bead workshop (Yerkes 1989:97). Kunnemann mound Cahokia

After Moorehead's excavations in 1921, Preston Holder partially excavated the Kunnemann mound (#10 and 11) in 1955 and 1956. Holder's reports are not formal publications, but rather presented papers, correspondence, etc. A publication of Holder's work was prepared by Tim Pauketat from Holder's field notes, correspondence, and reports on file (Pauketat 1993). The Kunnemann mound was 10.64 meters high, and is located in what is termed the Kunnemann Tract (Fowler 1989), about 1.5 kilometers north of Monks mound, and across the channel of Cahokia Creek (Pauketat 1993:8). The Kunnemann mound had two parts, a conical mound conjoined on top of a lower terrace. Structural remains, such as post pits and wall trenches were recovered, in addition to hearth features and human burials. The mound dates from 1000 to 1200 A.D. (Pauketat 1993:5). Carbonized fabric was recovered, and a quantity of charcoal suggests that the





57


building on the mound had burned down (Pauketat 1993:43). Sixty-eight chert microblades, a sandstone saw, and unfinished marine shell beads are evidence of a shell bead workshop.

Mound 72 Cahokia

Mound 72 was a small ridge-top mound 2.5 or 3 meters high (Fowler 1989-148). Its location is due south of the same point on the southwest corner of Monks mound that is due east of Mound 84 (see above), and forms a 90' angle with this line (Ahler and Depuydt 1987:4). Fowler (1991) discussed possible alignments of Mound 72 with other mounds at Cahokia. Most of the features date to around 1000 A.D. (Fowler 1991:3). Mound 72 entombed a high status human burial (often referred to as the "Beaded Burial") that was placed on a platform of marine shell beads (see Chapter 2), however, other burials from Mound 72 indicate individuals of both high and low status (Fowler 1991:3). Since some artifacts included in my sample are from various locations within Mound 72, these features excavated from Mound 72 are mapped in Figure 4.3.

Exotic materials found in the mound include lithic arrowheads originating in

northeast Oklahoma, some from Wisconsin, and others from southern Illinois. Ceramics indicate connections with the Lower Mississippi River valley (Fowler 1991:14, 23). One of the human skeletons from Mound 72 has had isotopic analysis performed by Stanley H. Ambrose from the University of Illinois. This individual's isotopic signature suggests that the person came from the Gulf coast or Caribbean area (Ambrose, personal communication 1996). If true, this may indicate more direct connections to coastal peoples. Ambrose's isotopic analyses of the Mound 72 burials are not yet published.









58




















z






cc





0 CI4



U

00 C)
Lj



E

-0
C 10 LO
D CD
0 0
E



CD
LLLL






CN









10 CI)
CN


CN

ca
ce
D LU
0
Uj U
J) ::*
C>




















Figure43. Mound 72, Cahokla site. After Fowler (1991).





59


Etowah

The Etowah site is located above the fall line in northeast Georgia on the Etowah River, a tributary of the Coosa and Alabama Rivers. It was first excavated by Cyrus Thomas in 1884 (Thomas 1985:292 [1894]). Locations of the major features of Etowah are shown in Figure 4.4. The name of the site is probably an anglicization of the Muskogean word etulwa (also spelled talwa [Wright 1986:231 and Otilwa [Bell 1991:333]), meaning town or township (Moore 1995). Kelly and Larson (1957:44) thought that at the time of European contact the site was occupied by Creek tribal ancestors. The Etowah site had at least seven mounds, with a canal or moat surrounding the center of the site and running into the Etowah River (Kelly and Larson 1957:41; Moorehead 1932). Mound A was about 21 meters high (70 feet) (Moorehead 1932:4), Mound B about 10 meters high (35 feet) (Thomas 1985:295 (1894]), and Mound C was about 5 or 6 meters high (18 feet) (Larson 197 1). Mound C had many burials, some with exotic grave goods. A total of 350 human burials have been excavated from the whole site, and most of the burials date frorn 950 to 1440 A.D. (Larson 1971:61).

Warren K. Moorehead of the Phillips Academy in Andover, MA, excavated

various locations at Etowah beginning in 1925 and continuing through 1927 (Moorehead 1932). He unearthed burials in Mound C and from the Etowah Village immediately to the north of the large mounds, and provided a description of the kinds of artifacts recovered with each burial. Copper, lithic, shell, ceramic, textile, and pearl bead artifacts were found, His excavations indicate that individuals buried in Mound C had greater amounts, and more exotic, artifacts interred with them than those from the Village area.






60









Canal
G63 Village F 0
EE







































Figure 4.4. Etowah site map. After Cyrus Thomas (1985:299 r[ 894]).






61


However, the presence of all ages and both sexes seems to infer prescribed status (Larson 1971:66), in other words, a person was probably born with high or low status. Lewis H. Larson excavated at Etowah from 1954 through 1958 (Stuart 1991:58). In addition to house remains and middens, Larson found a total of 210 burials, some with elaborate grave goods (shell, copper, lithics, wood, cloth, tortoise shell, and ceramics) (Kelly and Larson 1957:4 1). The burials were in simple pits or log tombs.

Indications of exotic trade include limestone-tempered pot sherds, possibly from Tennessee (Wauchope 1966:256), marine shell beads, Busycon spp. and Cassis madagascariensis bowls (van der Schalie and Pannalee 1960:45), Dover flint from Tennessee or southern Illinois, shark teeth, and sea turtle shell (Larson 1971:66). Brain and Phillips (1996:173) suggest a direct connection with the Lake Jackson site in Florida, based on shell gorget designs. Copper may have been mined locally in northeast Georgia (Larson 1971:64). However, the design styles on two copper headdresses indicate a connection with the Moundville site (Brain and Phillips 1996:158), and the style of the copper-covered wooden masks are similar to Spiro site materials (Brain and Phillips 1996:159).

Moundville

The Moundville site in west-central Alabama has remains of over 20 mounds

arranged around a plaza (Figure 4.5), dating from 1050 to 1550 A.D. (Steponaitis 1983b). It is on a bluff of the Black Warrior River, about 2.5 kilometers south of the city of Tuscaloosa (Steponaitis 1983b:3). The largest mound, Mound B, is about 17 meters high and 100 meters squared at the base. Clarence B. Moore excavated at Moundville in 1905





62





N




bluff II








scD















0 250 500
Rt I "








FEET HO M0











Figure 4.5. Moundville site map. After Moore (1905:129).





63


and 1906, unearthing 800 human burials. Various federal governmental agencies excavated from 1929 to 194 1, and these efforts revealed over 2000 burials (Steponaitis 1983b:7). Hallmarks of the Moundville remains are pyramidal platform mounds, animal effigy ceramics, and shell-tempered pottery, some of which had a "black filmed" surface. A presum-ed shell bead workshop was detected from an area east of Mound E (Peebles and Kus 1977:442; Welch 1991:170), consisting of unworked shell and finished beads. An oft-cited publication on some of the archaeological correlates of chiefdoms was produced from archaeological materials at Moundville (Peebles and Kus 1977). This study suggested that Moundville was indeed a chiefdom-level society using the following criteria: 1) ascribed ranking of persons, 2) hierarchy of settlement types and sizes, 3) a high degree of local subsistence sufficiency, 4) an organization of productive activities that transcends the basic household group, and, 5) a society-wide buffering system against environmental perturbations (Peebles and Kus 1977:431-433).

Exotic goods recovered from the site include ceramics from the central

Mississippi valley (Parkin Punctated, Bell Plain, and Nodena Red and White), sherds from Tennessee (Nashville Negative Painted), and from the Gulf coast (Pensacola Incised and Lake Jackson Decorated) (Steponaitis 1983b:221-225). Exotic lithics include; Dover chert from Tennessee, Ft. Payne chert from northern Alabama, Tallahatta. quartzite from southern Alabama, and mica sheets from several hundred kilometers to the east and northeast (Welch 1991:161-162, 169). Mill Creek chert hoes from southwestern Illinois (Welch 1991:173), copper from the Great Lakes area, and galena from Missouri add to the inventory of foreign materials (Peebles and Kus 1977). Other exotic items are marine






64


shellI artifacts, which Peebles and Kus (1 977T443) state are "by far the most abundant non-local itemn found at Moundville." Brain and Phillips' (1996:298-299) study of shell gorgets revealed that two engraved shell cup fragments from Moundville show a close stylistic affinity with engravings from the Spiro site. Spiro

The Spiro site (34Lf40) is located in Leflore County, eastern Oklahoma, on the south bank of the Arkansas River, about 45 kilometers downstream of the confluence with the Canadian River. It consists of a series of burial mounds, known as the Craig mound, and a series of habitation mounds (Brown 1996) (Figure 4.6). The site dates from 1000 to 1450 A.D., with the burials from the Great Mortuary in the Craig mound dating to 1400 A.D. (Brown 1996:153-161). The Great Mortuary is the central feature of Craig mound, and is a mounded charnel house from which most of the grave goods were obtained. Other burials contributed to the sample. A total of 377 individual burials were recovered (Brown 1996:2 16).

The excavation history at Spiro is among the best well-known in North America. In 1933, a group of men began legally looting the Craig mound for its salable riches. They reportedly unearthed artifacts such as copper axes, copper needles, pearl beads, stone effigy pipes, textiles, wooden artifacts, quartz crystals, about 453.6 kilograms (1,000 pounds) of galena, engraved and unengraved sinistral whelk shell cups, and shell beads (Hamilton 1952). This commercial venture continued until November 1935. A joint excavation by the federal Works Projects Administration (WPTA) and the University of Oklahoma was performed from 1936 to 194 1, under the direction of Forrest E.






65






Copple
Mounds
30



iQ Brown

20 6 0 Mound

60
40
N5 Craig

Mound




0 100 200 Word Mound 2
i 0
METERS Ward Mound I


























Figure 4.6. Spiro site map. After Brown (1996:15).






66


Clements (Orr 1946:228). However, a detailed account of these scientific excavations awaited publication by James A. Brown in 1996.

Numerous non-local materials were excavated from Spiro. Exotic lithics include: Mill Creek chert from southwestern Illinois; marble from the southern Appalachians in eastern Tennessee and northwest Georgia; "greenstone" (chlorite schist) and phyllite from the southern Appalachians (in northern Georgia?); copper from either the Lake Superior region or from the southern Appalachians; galena from the Potosi formation near Cahokia; mica from western North Carolina; Burlington chert from Missouri that had been fashioned into a Cahokia Notched point, Dover chert from either the Cahokia area or Tennessee; Kaolin chert from southern Illinois; and Alibates dolomite from the panhandle of Texas; Smoky Hill jasper from northern Kansas or southern Nebraska (Brown 1996:639-649; Phillips and Brown 1978:19). Ceramics that indicate connections with a wide region of the southeast are: Smithport Plain from the Arkansas River valley in Louisiana and Arkansas; Coles Creek Polished Plain from southwestern Arkansas; Pennington Punctate-Incised from the Mississippi valley area of southern Arkansas; Undesignated Irregular Zigzag from southern Mississippi; Powell Plain from the American Bottom; Parkin Punctated from the central Mississippi valley; Hiwassee Island Red-on-Buff from western Tennessee; and Nashville Negative Painted from central Tennessee (Brown 1996).

Shell engraving stylistic affiliations with other areas have also been studied. The Braden shell engraving style at Spiro shows a connection with ceramic incision styles from Cahokia and the Mississippi Valley region (Brown 1989). Another assertion is that





67


an engraved shell gorget found at Spiro is intrusive from the Etowah area (Brown 1989:187; Phillips and Brown 1978:126-127). Phillips and Brown (1978:166) suggested a Huastecan connection with the Spiro site based on motifs, however James Brown (1996:32) has recently commented that "it is highly unlikely that Spiro was part of a bridgehead of any importance to the Southwest." Brown (1996:20) notes a decline in theories of Mesoamerican influence on Spiro art styles.

The famous Craig mound at Spiro had most of the burials and exotic artifacts

found at the site. The main cone-shaped mound was about 10 meters (33 feet) in height. Three other mounds were built to the southeast, 6.4, 6.4, and 1.4 meters high, consecutively (Brown 1996:53). James Brown (1996:65) believes part of the Craig mound consisted of "the remnants of a berrn surrounding a central roofed structure." This structure afforded the excellent preservation of some textiles, wood, and other fragile artifacts. All of the burials in Craig mound date to the Spiro phase, 1000 to 1450 A.D. (Brown 1996:16 1). Excavations from the Craig mound were designated by the WPA as either LfCrl or LfCrIl (Leflore County, Craig mound), and only the main mound was excavated by the WPA crew (Brown 1996:54). Burials were numbered consecutively, B 1, B2, etc. Artifacts not directly associated with burials were designated consecutively, A I, A2, etc. (Brown 1996:60). A "dump" east of the mound consisting of the spoils of the excavations yielded some artifacts, which were numbered D I, D2, etc. The matrix was not screened.






68


Materials and Methods


Materials

Collections from the four sites included in the samples I studied are curated at institutions throughout the eastern United States. Positive identifications of the marine shell species used for artifact manufacture usually cannot be made from an illustration: additionally, most of these shell artifacts are not illustrated. It was therefore necessary to travel to various institutions to build a database of the artifacts and the marine shell species from which they were made. I included as many marine shell artifacts as possible from each of the four sites. The institutions I visited are listed in the legend of Appendix 2.

Methods

Infon-nation for each specimen was recorded on data sheets (see Figure 4.7). A specimen that could not be identified at least as sinistral or dextral was not included. Disk beads made from the outer whorl of a gastropod shell were not included, since no taxonomic identification is possible (at this time) for these artifacts. The following data were recorded: type of artifact (e.g., cup, bead, bead blank, pendant, gorget, pin, debitage, and whole shell), catalog number, the shell taxon from which it was made, whether it was from a sinistral or dextral shell (especially for columella beads), length, the apex angle (when present, solely for species determination of sinistral whelk shells), whether the artifact was engraved, and whether it was associated with a burial. An attempt was made to collect information on the age and sex of the individuals associated with shell artifacts, but this information is usually not available. Any pertinent






69


observations were also recorded in the "comments" section. All measurements were in millimeters. The data was input into a relational computer database (Microsoft Access).

Identification of some shell artifacts is easy. However, the more fragmentary remains of shell gorgets and masks make shell species identification difficult or impossible. Most masks and gorgets are sanded, sometimes obliterating the distinguishing characters necessary for identification. It is possible to identify some of them to either a dextral or sinistral shell, and if sinistral, whether it is from a gracile or robust shell.

Growth lines and shell curvature combine to allow identification of masks and gorgets made from dextral or sinistral shells. The artifact must be in a state of preservation that allows the detection of growth lines on the shell. Sometimes the direction of shell growth may be determined by examining the edge of the gorget. In whelk shells, growth lines converge toward the anterior of the shell, toward the anterior of the shell indicated by the siphonal canal (Figure 4.8). If growth lines are not visible, taxonomic identification is not possible. Secondly, a large enough section of the outer whorl of the shell must be present in order to detect shell curvature. Curvature is more flat at the lip and more pronounced towards the columella (Figure 4.8). Only specimens that were able to be identified using the above procedures were included in the sample I studied.

Shell beads made from gastropod shell columellas may also be sufficiently

distinctive to determine whether they were made from dextral or sinistral shells. These





70


INSTITUTION:

SITE: SITE#:

CAT.#: ENGRAVED?:

PROVENIENCE: BURIAL?

MALE/FEMALE: ADULT/SUB/JUV:

ARTIFACT TYPE: LENGTH:

WIDTH: APEX ANGLE: SHOULDER ANGLE:

SINISTRAL/DEXTRAL: TAXON:

COMMENTS:







CAT.#: ENGRAVED?:

PROVENIENCE: BURIAL?

MALE/FEMALE: ADULT/SUB/JUV:

ARTIFACT TYPE: LENGTH:

WIDTH: APEX ANGLE: SHOULDER ANGLE:

SINISTRAL/DEXTRAL: TAXON:

COMMENTS:








Figure 4.7. Sample data sheet.





71


beads are referred to in the archaeological literature as barrel, tubular, massive, columella, cylindrical, etc. (see Chapter 2). Some columella beads may be identified to species based on their large sizes, either to horse conch or lightning whelk. Of course, this depends upon the presence of the columella groove. Only columella beads that have a colurnella groove clearly slanting to one direction or the other were included in my sample. In a few specimens, the groove was parallel to the hole, and these were not included. William H. Holmes (1 883) used these same criteria to distinguish between dextral and sinistral columella beads (see Chapter 2 and Figure 2. 1). "Bead blanks" were unfinished columella beads that have evidence of being worked; either cut or scored. Columellas, without direct evidence of being worked were classified as debitage.

Categories were, by necessity, imposed upon some of the artifacts. It should be noted that some of the specimens are artifact fragments. Artifact type was sometimes inferred from context and sometimes was not known. Debitage fragments were generally broken pieces of shell having no particular shape. Shell "pins" may have been earrings, hairpins, or some kind of device to help fasten clothing. A pin consists of a columella that has been worked to a long, tapered, cylinder with a larger piece of the columella remaining intact to serve as a stopper. Pendants have been wellI-documented, therefore this category is more reliable; the same is true of cups. Some artifacts could not be classified according to function.


Results

The sample of marine shell artifacts is large with a total of 17,645 specimens examined: 8,318 from Cahokia, 1,257 from Etowah, 40 from Moundville, and 8,030





72



















































Figure 4.8. Outer whorl of sinistral whelk shell. Note growth lines used for
identification.






73


from Spiro. A summary of the raw data on artifact types is presented in Table 4. 1. Appendix 2 is the complete database.

It is immediately apparent that most of the specimens examined are beads. Most of these beads are made from small whole shells of two types; marginella and olive shells, and because they are usually not fragmented, they are much easier to identify. From Cahokia, 97% (7,735) of the 7,949 beads were from small whole shells of two types olive and marginella shells (of the genera Oliva and Marginella). Of the 1,209 beads from Etowah, 92% (1,1 18) were from either olive or marginella shells; 27% (six) of the 23 beads from Moundville were from the genus Oliva, and 87% (6,501) of the 7,451 beads from Spiro were from small whole shells representing the genera Oliva, Olivella, Marginella, and Thais.

The bead count represents skewed data. I housands of columella and disk beads made from large marine gastropod shells have been recovered from these sites (see Chapter 2). Relatively few of these beads could be identified to any taxonomic level, and for that reason they were not included in my sample. This biased my sample toward small, whole shell beads, which are much easier to identify. Therefore, the bead count should not regarded as a comprehensive numeration of all bead types.

If we focus upon percentages of artifacts other than shell beads and bead blanks, a different picture emerges. Calculations of the percentages of different artifact types excluding shell bead totals are presented in Table 4.2. Shell cups and pendants comprise most of the sample, while other artifact types contribute only small percentages. Cups





74


TABLE 4.1
ARCHAEOLOGICAL SPECIMENS BY ARTIFACT TYPE


Cahokia Etowah Moundville Spiro
Artifact Type Number Percent Number Percent Number Percent NumberPercent whole shell 17 0% 1 0% 2 5% 4 0%

cup 2 0% 8 1% 5 12.5% 374 5%

bead 7949 96% 1209 96% 23 57.5% 7451 93%

bead blank 28 0%

debitage 314 4% 9 1% 3 7.5% 19 0%

gorget 2 0% 12 1% 5 12.5% 13 0%

pin 17 1%

spoon 1 0%

pendant 5 0% 2 5% 132 2%

unclassified 1 0% 37 0%

TOTAL 8318 100% 1257 100% 401 100% 1 8030 100%





75


TABLE 4.2
ARTIFACT TYPES EXCLUDING SHELL BEADS


Cahokia Etowah Moundville Spiro
Artifact Type Number Percent Number Percent Number Percent Number Percent whole shell 17 5 1 2 2 12 4 1

cup 2 1 8 17 5 29 374 65

debitage 314 92 9 19 3 18 19 3

gorget 2 1 12 25 5 29 13 2

pin 17 35

spoon 1 2

pendant 5 1 2 12 132 23

unclassified 1 0 37 6

TOTAL 341 100 48 100 17 100 579 100






76


TABLE 4.3
NUMBER OF ARTIFACTS BY SPECIES
Taxon Cahokia Etowah Moundville Spiro Total
Gastropods # 0% # % # % ft % #t %
Busycon sinistrwn 316 4 79 6 29 72.5 1249 16 1673 10
Busycon cf sin/strum? 41 0 5 0 1 2.5 1 47 0
Busycon spp. 1 0 1 0
Busycotypus spiral us 8 0 8 0
Busycotypus cf spiratus 1 0 1 0
Busycon cf carica 17 1 17 0
cf Busycon cari .ca 2 0 2 0
Fasciolaria iuhipa 1 0 1 0 2 0
F7asciolaria spp. 2 0 1 2.5 3 0
Pleuroploca gigantea 10 0 6 0 1 2.5 71 1 88 1
cf. Fleuropioca gigantea 1 0 1 0 2 0
Fasciolariidae 1 0 1 0
Str~mbus a/at us 3 0 3 0
S troinbus qf costa/us 1 0 1 0
Littorina irrorata 1 0 1 0
0liva spp. 126 2 1 0 6 15 3 0 136 1
Oliva cf litterata 1 0 1 0
0liva cf. sayana 1 1 0 1 0 12 0
Olive/la b/pl/cata 33 0 33 0
Olive//a spp. 6051 75 6051 34
Marg/nella cf. apicina 82 1 82 0
Marg/neila spp. 7517 90 1117 89 400 5 9034 51
Tha/s haemostorna 2 0 2 0
cf. Thais 14 0 14 0
Conus cf spur/us 2 0 2 0
Cass/s madagascar/ens/s 6 0 6 0
Cass/s spp. 1 0 7 0 8 0
Hal/ot/s cracherodi/ 1 0 1 0
Phy/lonotus pon urn 2 0 2 0
Pha//um ef granulatum 1 0 1 0
Unclassified Beads 188 2 29 2 2 5 184 2 403 2
Bivalves
Mercenaria spp. 1 0 1 0
D/nocard/um rob ustum 2 0 2 0
Crassosirea virgin/ca 2 0 2 0
Anadara transversa 1 0 1 0
Macrocall/sta ninbosa 1 0 1, 0
TOTAL 8318 1257 40 8030 17645





77


TABLE 4.4
ARTIFACTS BY SPECIES
EXCLUDfNG OLIVE AND MARGINELLA BEADS


Taxon Cahokia Etowah Moundville Spiro Total
Gastropods # % # % # % N % # %
Busyconsinistrum 316 54 79 57 29 85 1249 81 1673 73
Busycon cf sinistrum 41 7 5 4 1 3 47 2
Busycon spp. 1 0 1 0
Busycolypusspiralus, 8 1 8 0
Busycotypus cf spiratus 1 0 1 0
Busycon cf carica 17 12 17 1
cf Busycon carica 2 1 2 0
Fasciolaria lulipa 1 0 1 0 2 0
Fasciolaria spp. 2 0 1 3 3 0
Pleuroploca giganfea 10 2 6 4 1 3 71 5 88 4
cf Pleuroploca 1 0 1 1 2 0
gigantea
Fasciolariidae 1 0 1 0
Strombus alatus 3 1 3 0
Strombus cf costatus 1 0 1 0
Litforina irrorata 1 0 1 0
Thais haemostoma 2 0 2 0
cf Thais 14 1 14 1
Conus cf spurius 2 0 2 0
Cassis madagascariensis 6 0 6 0
Cassis spp. 1 0 7 1 8 0
Haliotis cracherodii 1 0 1 0
Phyllonotus pomum 2 0 2 0
Phalium cf granulatum 1 0 1 0
Unclassified Beads 188 32 29 21 2 6 184 12 403 18
Bivalves
Mercenaria spp. 1 0 1 0
Dinocardium robustum 2 0 2 0
Crassostrea virginica 2 0 2 0
Anadara fransversa 1 0 1 0
Macrocallista nimbosa 1 0 1 0
TOTAL 581 139 3 4 1542 2296





78




TABLE 4.5
RATIOS OF DEXTRAL AND SINISTRAL COLUMELLA SHELL BEADS



Cahokia % Etowah % Moundville % Spiro % Total %

Sinistral 200 93 46 51 16 94 877 92 1377 88

Dextral 14 7 45 49 1 6 73 8 195 12

TOTAL 214 91 17 950 1572













TABLE 4.6
SINISTRAL WHELK APEX ANGLES



Sample Range Standard
Site Size Low High Average Mode Deviation

Cahokia 63 85 125 111 115 9.22

Etowah 3 105 125 117 10.41

Moundville 2 100 125 113 17.68

Spiro 181 94 140 120 120 7.93






79


were more abundant at Spiro, while very uncommon at Cahokia. Gorgets were more frequent in collections from Etowah and Moundville than at either Spiro or Cahokia. Species Used in Artifact Manufacture

As with percentages of artifact types, percentages of species used in artifact

manufacture are biased toward small, whole shell beads (see above), such as dwarf olive and especially marginella shells (Table 4.3). Again, this is a result of method: small whole shell beads are easily identifiable, while disk beads and most columella beads were not included in my sample because the shells from which they were made cannot be identified. If we remove this bias from the sample by calculating percentages without olive and marginella beads, lightning whelk specimens become most common (Table

4.4). Lightning whelk and horse conch now contribute 73% and 4%, respectively, to the specimen count. Unclassified columella beads contribute most of the rest (18%), and will be discussed below.

Columella Beads

One of the goals is to gather data on sinistral versus dextral columella beads using the method described above. Only about one-third of the columella beads I encountered had columella grooves that allowed designation of sinistral or dextral shell. However a large enough sample could be identified (n=1,572), and it can be relied upon as a valid subsample of columella beads (Table 4.5). Most (88%) columella beads from all four sites were made from sinistral shells, and the remaining 12% were made from dextral shells. The columella beads found at Cahokia, Moundville, and Spiro were far more





80


abundant (92-94%) than dextral columella shell beads. The columella beads at Etowah were fairly evenly divided between sinistral (51 %) and dextral (49%) shells.

Columella bead blanks were only found in collections from Cahokia. Twentyeight columella bead blanks were examined, and all were from sinistral whelk shells. Some of them had clear evidence of being worked, either cut, scored, or burned. None had drill-holes. Nine bead blanks were burned, while 19 were not burned. I found remains of a columella bead reduction sequence, starting with the unworked columella (Figures 4.9 and 4.10) and ending with a finished bead. Burned Specimens

Many of the shell specimens, particularly those from Cahokia, were burned. A total of 157 specimens including beads, bead blanks, and debitage fragments from Cahokia were burned. The majority of these (125) were whole shell beads made from olive shells (Oliva spp.), which had burned portions. The remaining 32 fragments were from sinistral whelk shells, of which 88% (28) were burned columellas, 9% (3) were inner whorl fragments, and 3% (1) was an apex fragment. The large percentage of burned columellas and inner whorls at Cahokia suggests a shell working technique that will be discussed below. Spiro had three burned specimens, consisting of two burned sinistral columella beads, and one debitage fragment of a sinistral whelk shell (A6-13). The debitage fragment was sanded or ground on the outer whorl, and had an intact apex, spire, columella, and inner and outer whorls.





81























































Figure 4.9. Cahokia columrnella bead blanks from Ramey Field (ISM #1958-91).
























































Figure 4.10. Cahokia, finished beads from Mound 72.





83


Apex Angles

In an effort to differentiate sinistral whelk shell species, the angle of the apex was recorded on all sinistral whelk artifacts possible. The apex angle (in degrees) is one of the defining characteristics between lightning whelks and prickly whelks. Table 4.6 summarizes apex angle data from the four sites. Unfortunately, Etowah and Moundville both had very small sample sizes (three and two, respectively), but Cahokia and Spiro had enough to comprise adequate samples. The average apex angle from sinistral Busycon artifacts from Cahokia is 1110, and 1200 from Spiro. Lightning whelks have apex angles of about 1050, while prickly whelks have apex angles of about 900 (see Chapter 3).

Engraved Artifacts

Only cups, debitage fragments, pendants, and gorges were engraved. The great majority are from sinistral whelk shell artifacts. Only one artifact from Cahokia (made of a sinistral lightning whelk shell) was engraved: a gorget from Ramey Field with a cross design. Twelve engraved gorgets were included in my sample from Etowah and all are from sinistral shells; eleven from lightning whelk shells, and one probably being made from a lightning whelk shell. Seven engraved fragments were from Moundville and included four cups and three gorgets, all made from lightning whelk shells.

The 144 engraved specimens from Spiro included 128 cups, two debitage

fragments, twelve gorges, one pendant, and one artifact of unknown type. Ninety-seven percent (140) of engraved specimens from Spiro were from lightning whelk shells. One gorget made from a horse conch (A448746) was engraved on both concave and convex






84


sides of the outer whorl. A debitage fragment (18/9080) consisting of the lip of a helmet shell (Cassis madagascari ensi .s) was engraved in a scalloping pattern on the edge of the lip. Two cup fragments made from horse conch shells were engraved (A423288 and A26-60).


Experimental Archaeology


In an attempt to understand better the methods of making a shell cup without using modern high-speed electric-driven metal tools, I made four shells cups. I purchased lightning whelk shells that presumably came from the Florida west coast. The first one I made was from a small shell about 18 cm long. The second was from a slightly larger shell about 22 cm long, but very gracile and thin-shelled. The third was from a very large and gracile shell about 37 cm long. The last was from a shell about 26 cm long. Bowing to modem tools, I used a small ball-peen hammer to remove unwanted portions of the outer and inner whorls. This was a substitute for the lithic hammers that were available to Mississippian peoples.

Removal of the outer and inner whorls was relatively easy. I removed the entire columella from all but the last specimen mentioned above. However, after the whorls were taken out, removal of the columella required more precise hammering. On the gracile shell of 22 cm, I broke the apex out while trying to remove the columella in the spire area. With the smaller specimens, I was able to remove all of the columella by hammering. On the very large shell I found I could not remove the columella by hammering, except by breaking the outer whorl, thus ruining the cup.





85


1 spoke with a crafts instructor (Ray Ferguson, personal communication) at the University of Florida, and he suggested that I try to use fire to help make the shell more friable so that it could be more easily worked. Shell that has been exposed to heat is more brittle and chalky, since the organic constituents have been burned away. At first I used small sticks taken from a fire, but that method did not work. What was needed was an intense focused flame, I followed the suggestion to use a blow torch, though I was hesitant at first since Mississippian peoples did not have blow torches. However, Mississippians did have means of using directed fire, such as pitch pine or small torches. The blow torch worked well, as long as the heat was not too intense. If the heat was kept focused for too long, the shell started popping and cracking, threatening to explode, I kept moving the small flame evenly over the area that I wanted to break. The fire worked, after about an hour of heating the desired area on the shell. Using the ball-peen hammer, I was now able to hammer the burned columella off the rest of the shell without breaking the outer whorl into pieces. The resulting burned columella is pictured in Figure 4.11.


Discussion and Conclusions


Connections among the four sites are indicated not only by the presence of similar artistic motifs, which have been so thoroughly studied (Brain and Phillips 1996; Emerson 1989; Waring and Holder 1945; Williams and Goggin 1956), but by exotic materials and artifacts. The uneven distribution of natural resources among these four sites predisposed them to engage in some kind of trade or exchange, but not the long-distance





86

























































Figure 4.11. Experimental archaeological burned columella.






87


trade that apparently existed. Certain substances were desirable for their utilitarian attributes, such as Mill Creek chert from southwestern Illinois. Other materials were sought for purely ideological reasons, such as copper, which had more value as ceremonial objects rather than utilitarian items. Marine shells were used to make ceremonial and/or status items. Ethnographic and ethnohistoric sources clearly place marine shell artifacts in the ceremonial category. Shell Beads

A surprising number of small whole shell beads (from olive and marginella

shells) were found in the collections. Large quantities of this type of bead have not been widely reported in the archaeological literature, and the large percentages of them in my sample do not take into account the large numbers of disk and columella beads that were also recovered. Many archaeologists presume that most of shell disk and columella beads were made from sinistral whelk shells, and this underscores the bias in my sample toward olive and marginella beads. Additionally, small snail shells are easily gathered, sometimes by the hundreds (Kurt Auffenburg, personal communication), in the areas where they occur. They represent less of an investment in time and energy, since they are more easily transported, and much easier to make into beads.

Most columella beads (88%) were made from sinistral whelk shells (probably lightning whelk shells), confirming previous assumptions by archaeologists. If we examine artifact percentages without shell beads as presented in Table 4.2, we see an emphasis on bead-making debitage at Cahokia, from clear bead-producing areas such as Ramey Field, and mounds 33 and 34 (all east of Monks mound). Production of shell





88


beads was emphasized at Cahokia, almost to the exclusion of cups, gorgets and other shell artifacts. This may shed light on the debate concerning specialized production of shell beads at Cahokia. If shell beads at Cahokia were sought after so intensely, then some type of control may have existed over the creation of such a desirable product. Shell Working Technology

At Cahokia, the large proportion of burned columellas indicates that Cahokians occasionally used fire to help them separate the columella from the rest of the shell. Garcilaso de la Vega (1988:330 [1605]) mentioned in his account of DeSoto's expedition that pearls were pierced using hot copper needles, supporting that heat was a known method of working with shell-like material. Another early record (among the Natchez) from Dumont de Montigny (in Swanton 1946:486), stated that shell gorgets were pierced by means of fire. My own attempts to remove a columella from a large lightning whelk shell supports the importance of heat in shell-working technology. The resulting burned columella from my experiment looks remarkably similar to the archaeological specimens.

Shell cups from Spiro had clear evidence of having the columella cut out,

probably by a chert implement. A photograph of a shell cup from Spiro shows clear "saw" marks where the columella used to be (Figure 4.12). This contrasts with the burned columella method probably employed at Cahokia. The choice of technique, either burning or cutting, may depend on the desired outcome, either production of beads or cups. Differences in shell working technology could have been the result of local traditions.





89


Pacific Coast Shells at Spiro

An important result of the study of marine shell from the Spiro site is the presence of beads made from dwarf olive shells, and one unmodified black abalone shell. Although all. of the marine shell artifacts are from distant coastal areas, these artifacts are the first from Spiro to be identified as originating from the Pacific coast. Finding Pacific coast shells at Spiro supports the idea of Frank F. Schambach (1993) that the people who lived at Spiro conducted long-distance trade as their primary method of economic subsistence.

Kurt Auffenberg, Collections Manager of the Malacology Range at the Florida Museum of Natural History, made the identifications. Identification of the dwarf olive shells from Spiro (Olive/la biplicala, curated at the Oklahoma Museum of Natural History, Lf40/953c) was based on the large size of the shells, in addition to the pink and purple coloration that is still visible (Figure 4.13). The shell attains a length of 3.2cm (1.25 inches). This species is found only in the Pacific Ocean, from Vancouver Island, British Columbia, to Baja, California. It lives on sand substrates from low tide to 46 meters deep (Rehder 1996:585). This is the same species that was targeted by tribes from southern California for bead artifact manufacture from 900 to 1600 A.D. (Arnold and Munns 1994). On the Channel Islands off the California coast, whole shell beads, as well as disk beads made from the outer whorl, were made from these dwarf olive shells.

Most olive shells of the genus Olive//a were not identified to the species level, since specific identifications of this genus are very difficult. The majority of Olive//a shell beads at Mississippian sites have been assumed to originate from Atlantic or Gulf of





90




















































Figure 4.12. Sinistral shell cup from Spiro (Gilcrease #9025.1723, LfCrll).
Note the columella was cut out.





91


Mexico waters. Perhaps these beads should be re-examined to better determine the coast of origin.

The black abalone shell from Spiro (Haliotis cracherodii, A448776) also came from the Pacific coast. This species ranges from Coos Bay, Oregon. to Baja, California, and inhabits shallow rocky shores from the intertidal zone to 6 meters deep (Rehder 1996:343). It is the most common abalone on the Pacific coast.

Dwarf olive beads and abalone shells were widely exchanged in the Southwest. These may have been valued due to their purplish/pink color and shine. Pueblo people associated these shells with war, and on rare occasions made the long trip to the Pacific to collect shells themselves (Safer and Gill 1982). Pre-European trade routes were known to have existed that would have allowed transportation of Pacific coast shells to New Mexico (Brand 1938; Tanner 1989) and then to the Spiro site. Figure 4.14 is a map loosely based on Brand (193 8:8) showing how Pacific shells might have reached the Spiro site. Brand (1938) discussed possible trade routes for Pacific marine shells, especially Olive/la biplicata and Haliotis spp., using data from 132 archaeological sites. The most important procurement area for Olive/la biplicata beads on the California coast was the Santa Barbara Channel Islands (Ann Munns, personal communication), and it is probable that the dwarf olive shells originated there.

Potable water is especially important in the dry Southwest, and it is likely that travelers sought routes with reliable water sources. The Canadian River is saline for most of the year (Geraghty et a]. 1973), but is not saline during the spring, when snow is melting (Moore 1987:154). The term "saline" is defined as "dissolved solids, which may