... The Irene/Altamaha Transition on Sapelo Island, Georgia by Elizabeth Straub Department of Anthropology University of Indianapolis 1400 E Hanna Ave Indianapolis, IN 46227 May 2017 University of Indianapolis Material Culture Lab Technical Report Series No. 4 The Irene/Altamaha Transition on Sapelo Island, Georgia By Elizabeth Straub Dr. Christopher R. Moore Dr. Richard W. Jefferies Dr. Christopher W. Schmidt Director of Graduate Studies The Irene/Altamaha Transition on Sapelo Island, Georgia by Elizabeth Straub Department of Anthropology Material Culture Lab Technical Report Series No. 4 May 2017 ii Abstract The nature and timing of the transition between Late Prehistoric Irene ceramics and Mission period Altamaha ceramics are not well understood. Archaeologists in the Southeast debate over the what caused coastal Guale groups to adopt new motifs and decorative styles, but these changes have often been attributed to Spanish interaction. I have conducted excavations and analysis of Irene, Altamaha, and possible transitional contexts at Site 9Mc23 on Sapelo Island, in hopes of better understanding these changes. These comparisons suggest that there is something unique about the possibly transitional context. Chi-square tests indicate that it may be possible to identify transitional assemblages based on land and groove width, and surface treatment. No significant patterns were revealed through thin section analysis and comparisons of temper. I used Scanning Electron Microscope analysis to evaluate changes in slip technologies between these periods and revealed that Altamaha red filmed sherds are the only sherds that contain a true slip. This suggests a change in technology that is highly visible and may be associated with the Spanish. Unfortunate- ly, sample sizes for transitional sherds were small. The excavation of more possible transitional contexts is needed to fully understand their significance, and more radiocarbon dates for Irene and possible transitional contexts are needed to fully understand the timing of this transition. Acknowledgements Now that this project has come to an end, there are some people that I would like to thank. This project would not have been possible without the interest and support of Drs. Christopher Moore and Richard Jefferies. Thank you for your guidance and for making the field work for this project possible. Most of all, thank you for bringing me along to Sapelo. I owe Dr. Moore an extra thanks for pushing me to choose a thesis topic in the summer after my freshman year and for badgering me about it until I agreed to this one. Thank you for answering my endless questions and for dipping my thesis drafts in red ink. I would also like to thank Dr. Christopher Schmidt, not only for agreeing to serve on my iii committee, but for granting me access to the SEM and teaching me to use it. Thanks to Arysa Gonzalez for helping in this process, as well. I would like to thank Dr. Nadjib Bouzar for assisting me in finding the best statistical methods for this study. I would like to give a huge thank you to the University of Indianapolis and University of Kentucky students who contributed to the fieldwork portion of this project. Thanks to my good friend, Elizabeth Kunz, for taking the time to drive all the way to Georgia to donate free labor to this project, as well. It was a pleasure playing in the dirt with all of you. I owe an extra thanks to the University of Indianapolis students who assisted me in processing the floats for this project (including the 2016 Field Methods class). Thank you to the Georgia Department of Natural Resources, especially to Brian Tucker and Fred Haye for allowing us access to the R.J. Reynolds wildlife preserve and assisting us with the logistics of doing field work on Sapelo Island. Thank you to the InQuery Collaborative for the funding that made the lab and field work for this project possible. Of course, I would also like to thank my parents for their support throughout my time at UIndy. None of this would have been possible without you. Your encouragement and faith in my abilities has meant the world to me. Thank you for always picking up the phone when I needed you. Most of all, thank you for encouraging me to pursue archaeology, even when you werent sure that it was practical. iv Table of Contents List of Figures ............................................................................................................................ vi List of Tables ............................................................................................................................. ix Introduction ................................................................................................................................ 1 Background ................................................................................................................................ 4 Debates Over Spanish Influence12 Understanding Changes in Technical and Decorative Style..16 Field and Lab Methods .............................................................................................................. 17 Excavation and Ceramics Attributes Analysis .............................................................. 17 Thin Section Analysis ................................................................................................... 20 Scanning Electron Microscope Analysis ...................................................................... 22 Results of Investigations ............................................................................................................ 23 Excavations .................................................................................................................... 23 Units 42 and 43 .................................................................................................. 23 Units 46 and 50 .................................................................................................. 32 Unit 47 ............................................................................................................... 43 Units 48 and 51 .................................................................................................. 46 Ceramics Attributes Analysis ........................................................................................ 58 Statistical Comparisons .................................................................................................. 69 Thin Section Analysis .................................................................................................... 79 Scanning Electron Microscope Analysis ....................................................................... 85 Conclusions ................................................................................................................................ 90 References .................................................................................................................................. 92 v List of Figures Figure 1. Map Showing the Location of Sapelo Island .............................................................. 2 Figure 2. An Example of St. Simons Plain Fiber Tempered Pottery (UK6956) ........................ 6 Figure 3. An Example of Refuge/Deptford Check Stamped Pottery (UK6957) ....................... 6 Figure 4. An Example of Savannah Check Stamped Pottery (UK6829.1) ................................ 6 Figure 5. An Example of an Irene Rim Fillet with Whole Cane Punctations (UK6817.2). ...... 8 Figure 6. The Filfot Cross Motif ................................................................................................ 8 Figure 7. A Map of the Georgia Coast Depicting the Locations of Spanish Missions during the Mid-17th Century ................................................................................................. 9 Figure 8. Line Block Motif ........................................................................................................ 11 Figure 9. An Example of Cross Simple Stamping (UK6266) ................................................ 11 Figure 10. Altamaha Rim Fold/Strip Decorated with Fingernail Punctations (UK6263.3) ...... 11 Figure 11. Locations of Excavated Units ................................................................................... 18 Figure 12. Some Altamaha Varieties ......................................................................................... 20 Figure 13. Plan View of Units 42 and 43 ................................................................................... 23 Figure 14. Unit 42 West Profile ................................................................................................. 24 Figure 15. Unit 43, South Portion of Feature 50 ........................................................................ 24 Figure 16. Glass Beads from Feature 50 in Unit 42................................................................... 25 Figure 17. Glass Beads from Unit 42 ......................................................................................... 25 Figure 18. Coarse Earthenware from Unit 42 ............................................................................ 26 Figure 19. Porcelain from Unit 42 ............................................................................................. 26 Figure 20. Majolica Pottery from Units 42, 46, and 50 ............................................................. 27 Figure 21. Lead Shot from Unit 42 ............................................................................................ 27 Figure 22. Composite Plan View of Unit 46 .............................................................................. 33 Figure 23. Plan Views of Unit 50 .............................................................................................. 33 Figure 24. Unit 46 West Profile ................................................................................................. 34 Figure 25. Unit 50 West Profile ................................................................................................. 34 Figure 26. Feature 60 in Unit 46 ................................................................................................ 35 Figure 27. Feature 60 Profile ..................................................................................................... 35 Figure 28. Feature 60 Post in Profile ........................................................................................ 36 Figure 29. Feature 61 at 40 cm BD ............................................................................................ 36 Figure 30. Feature 62 at 40 cm BD ............................................................................................ 37 Figure 31. Anthropomorphic Pipe from Feature 60 ................................................................... 37 Figure 32. Biface from Feature 60 (UK6721) ............................................................................ 38 Figure 33. Coarse Earthenware from Units 46, 28, and 51 ........................................................ 38 vi Figure 34. Unit 47 Shell Midden East Profile ............................................................................ 43 Figure 35. Unit 47 East Profile................................................................................................... 44 Figure 36. Unit 47 Plan View at D = 50 cm BD ........................................................................ 44 Figure 37. Possible Transitional Sherd (UK6100) Recovered from STP 346 ................................... 46 Figure 38. Unit 48 in Plan View ................................................................................................ 48 Figure 39. Unit 48 Plan View Map ............................................................................................ 49 Figure 40. West Profile of Units 48 and 51................................................................................ 50 Figure 41. East Profile of Units 48 and 51 ................................................................................. 50 Figure 42. Feature 63 West Profile ............................................................................................ 51 Figure 43. Feature 64 West Profile Photo .................................................................................. 51 Figure 44. Feature 64 West Profile Drawing ............................................................................. 52 Figure 45. Feature 67 at 61 cm BD ............................................................................................ 52 Figure 46. Feature 68 North Profile Photo ................................................................................. 53 Figure 47. Feature 68 North Profile Drawing ............................................................................ 53 Figure 48. Feature 68 at 50 cm BD ............................................................................................ 54 Figure 49. Feature 66 Photo ....................................................................................................... 54 Figure 51. Pie Chart of Ceramics by Series ............................................................................... 58 Figure 52. A Sample of Uncommon Sherd Types ..................................................................... 59 Figure 53. Savannah Sherds ....................................................................................................... 59 Figure 54. Savannah Check Stamped Jar ................................................................................... 61 Figure 55. Irene Sherds .............................................................................................................. 61 Figure 56. Altamaha Sherd with an Unidentified Incised Design Over Overstamping ............. 64 Figure 57. Base Sherd with Two Painted Dots .......................................................................... 64 Figure 58. Altamaha Rims and Decorated Sherds ..................................................................... 65 Figure 59. A Sample of Altamaha Complicated Stamped Sherds ............................................. 65 Figure 60. Sample of Altamaha Overstamped Sherds ............................................................... 66 Figure 61. Complicated Stamped Altamaha Rim....................................................................... 66 Figure 62. Sherds with Identifiable Vessel Form ....................................................................... 67 Figure 63. Colonoware with Identifiable Vessel Forms ........................................................... 67 Figure 64. A Sample of Irene/Altamaha Sherds ........................................................................ 68 Figure 65. Boxplot of Maximum Body Thickness by Series ..................................................... 70 Figure 66. Savannah Sherds in Thin Section ............................................................................. 80 Figure 67. Irene Sherds in Thin Section ..................................................................................... 81 Figure 68. Altamaha Sherds in Thin Section ............................................................................. 82 Figure 69. Possible Transitional Irene/Altamaha Sherds in Thin Section ................................. 84 Figure 70. SEM Photo of an Altamaha Red Filmed Sherd ........................................................ 87 Figure 71. SEM Photo of an Altamaha Sherd Originally Coded as Having an Orange Slip ..... 87 vii Figure 72. SEM Photo of an Altamaha Sherd Originally Coded as Having a White/Tan Slip ..88 Figure 73. SEM Photo of an Altamaha Sherd with a Blackened Interior ..................................88 Figure 74. SEM Photo of an Irene Sherd Originally Coded as Having a White/Tan Slip .........89 Figure 75. SEM Photo of an Irene Sherd with a Blackened Interior ..........................................89 viii List of Tables Table 1. All Artifacts from Unit 42 ............................................................................................ 29 Table 2. All Ceramics from Unit 42 ........................................................................................... 30 Table 3. All Artifacts from Unit 43 ............................................................................................ 31 Table 4. All Ceramics from Unit 43 ........................................................................................... 31 Table 5. All Artifacts from Unit 46.. ......................................................................39 Table 6. All Ceramics from Unit 46... ....................................................................40 Table 7. All Artifacts from Unit 50.. ......................................................................41 Table 8. All Ceramics from Unit 50... ....................................................................42 Table 9. All Artifacts from Unit 47... .................................................45 Table 10. All Ceramics from Unit 47.. ...................................................................45 Table 11. AMS Dates from Features 50 and 66.. ...................................................48 Table 12. All Artifacts from Unit 48.. ....................................................................55 Table 13. All Ceramics from Unit 48.. ...................................................................56 Table 14. All Artifacts from Unit 51.. ....................................................................57 Table 15. All Ceramics from Unit 51.. ...................................................................57 Table 16. Table of Irene Temper... .....................................................................60 Table 17. Table of Irene Surface Treatments.. .......................................................60 Table 18. Table of Altamaha Surface Treatment.. .................................................62 Table 19. Table of Altamaha Temper.. ...................................................................63 Table 20. SPSS Output for Surface Treatment Chi-Square Test.. ..........................73 Table 21. SPSS Output for Surface Treatment Crosstab with Residuals.. .............73 Table 22. SPSS Output for Land and Groove Width Chi-Square Test.. ................75 Table 23. SPSS Output for Land and Groove Width Crosstab with Residuals.. ....75 Table 24. SPSS Output for Temper Chi-Square Test.. ...........................................78 Table 25. SPSS Output for Temper Crosstab with Residuals.. .............................. 78 Table 26. Sherds Used for Thin Section Analysis.. ................................................85 ix Introduction Between what southeastern archaeologists have called the Irene and Altamaha periods, rim forms, temper recipes, surface treatments, and decorative techniques and motifs all change (Saunders 2000). The timing and cause of these changes are still debated by archaeologists working in the Southeast. Often they have been associated with changes in social organization and religious views related to the rise of the Spanish Mission period around 1580 around the same time (Deagan and Thomas 2009; Saunders 2000, 2009). However, studies of this transition have yielded somewhat conflicting results and its nature and timing are poorly understood. The Sapelo Island Mission Period Archaeology Project (SIMPAP) has been investigating Mission period activities on at the Sapelo Shell Ring Complex (Site 9Mc23) since 2003 when University of Kentucky archaeologists discovered the sites Mission period component north of the monumental rings. SIMPAPs goal has been to locate Mission San Joseph de Sapala and the associated Guale town (Jefferies and Moore 2009). To date, excavations at Site 9Mc23 on Sapelo Is- land (Figure 1), have led to the discovery of a number of sherds with traits of both Irene and Altamaha ceramics. As part of this project, I have conducted excavations targeting Irene, Altamaha, and possible transitional context in order to further our knowledge about these changes and their cause. Macroscopic, thin section, and scanning electron microscope (SEM) analyses, along with statistical comparisons, have been used to better understand this mixed trait assemblage and the Irene-Altamaha transition. This project had three primary goals. The first was to determine how the technical and decorative style of Guale ceramics change between the Irene and Altamaha phases. Particular attention was be paid to which elements change and which may be temporally diagnostic. This investigation was be done through detailed attributes analysis and statistical comparison. Aside from often cited and diagnostic traits, this study revealed only two notable differences between Irene and Altamaha. The first was an increase in the diversity of temper recipes among Altamaha ceramics, which was noted through thin section analysis. The other was related to changes in slip technologies. SEM analysis indicates that red filming, a trait that is diagnostic of Altamaha, represents the only case of 1 N Figure 1. Map of Sapelo Island. 2 a true slip on ceramics from this area. Red filming represents a change in technology, rather than a change in style. The second goal was to determine when the Irene/Altamaha transition occurred on Sapelo Island. The timing of the transition varies along the coast, with earlier dates in the north and later dates to the south (Worth 2009). Saunders (2009) notes that the transition was extremely rapid, but we do not know when the transition occurred on Sapelo Island. Radiocarbon dates from Irene and Altamaha contexts at Site 9Mc23 will contribute to a greater understanding of the spread of Altamaha along the Georgia coast. However, most of the datable material recovered in these excavations came from Altamaha contexts. More dates from Irene and possibly transitional contexts are needed to accomplish this goal. The final goal of this project was to determine whether mixed trait sherds from Sapelo Island represent a transitional period. This was investigated through the excavation of a possibly transitional context and analysis of mixed-trait sherds. This could contribute to a better understanding of what this period of transition looked like and how quickly it occurred. Based on the results of the statistical comparison of temper and surface treatment, this research suggests that it is possible to identify transitional contexts at the assemblage level. This is further supported by thin section analysis, which indicates an increase in temper recipes during the Altamaha period. This diversity was not observed in Irene contexts or those thought to be transitional. 3 Background Sapelo Island is a barrier island located just off the Georgia coast in McIntosh County (Figure 1). Unlike many of Georgias other barrier islands, Sapelo is not a remnant Pleistocene shoreline, but a Holocene formation, which was deposited on top of Pleistocene mud as sea levels fell (Truck and Alexander 2013). McMichael (1980) has divided the island into five zones based on soil type, vegetation, and elevation. The zone that he named Stratum A is most likely to contain sites and consists of two ridges, one each on the east and west side of the island. These areas have high elevations compared to the rest of the island; soils range from poorly to excessively drained. Vegetation consists of maritime live oak forest. The concentration of settlements in this zone is noted on other coastal islands, as well. The island as a whole has a relatively low elevation, at about three meters (10 feet) (USCD 2017). Site 9Mc23 is located on the western edge of the island, just inside the islands maritime forest, with direct access to the salt marsh. This location would allow its residents easy access to what have been called the two most productive resource zones for coastal Georgia hunter-gatherers (Thomas 2008a). Resources on the island are abundant and diverse, with a number of productive terrestrial and marine foraging possibilities available to its residents. While many different foods were consumed by the residents of Site 9Mc23, common foods included a variety of fish, deer, clam and oysters which were (Jefferies and Moore 2009; Thomas 2008a). Sapelo may have been inhabited almost immediately after its formation in 4500 B.P., but these populations likely would have been small and their settlements temporary, leaving little evidence of their occupation (Truck and Alexander 2013:186-187). Excavations at Site 9Mc23 confirm that the site has been occupied at least since the Late Archaic (cal. 3000-1000 B.C.), when the most famous component of the site, three monumental shell rings were constructed and occupied (Thompson 2006, 2007; Thompson and Andrus 2011; Thompson et. al 2004). Today, the largest of these rings remains more or less intact, and the second is still partially visible, though damaged by erosion and mid 20th century excavations. The third and smallest ring is now visible only as a scatter of shell. 4 Investigations of the interior of the Shell Ring II have yielded very few prehistoric artifacts. However, investigations of the areas outside of the rings yielded evidence of Late Archaic occupation, including features that have been interpreted as structures. These excavations suggest that the shell rings were likely formed by domestic midden deposits (Simpkins 1980). This interpretation is supported by resistivity studies, which indicate that most of the shell deposits accumulated slowly over time. Based on the presence of a few clean shell deposits, it is likely that some feasting events occurred at the site, but they account for only a small portion of the deposits. Seasonality studies of oyster deposits in the shell rings indicate that the site was occupied year-round (Thompson and Andrus 2011). While Sapelos inhabitants were constructing the shell rings, they were producing what archeologists have called St. Simons pottery (Jefferies and Moore 2009; Sassaman 1993) (Figure 2). This pottery is characterized by molded or modeled construction and plain or incised surface treatments, often decorated with punctations (DePratter 1991). In order to produce pottery, non-plastic materials are often added to the clay in a process called tempering. St. Simons pottery tends to be tempered with organic material, as is most early pottery worldwide (Sassaman 1993). Specifically, St. Simons pottery is tempered with Spanish moss, which is abundant in coastal Georgia (Simpkins and Allard 1986). On Sapelo, the St. Simons series is followed by Refuge/Deptford and Wilmington (Figure 3), dating to the Woodland period (about cal. 1000 B.C. to A.D. 800) (Thompson 2006). This is when the coil-building technique replaced earlier molding and modeling techniques and grit/sand and sand tempers became common (DePratter 1991). The Savannah phase replaces Refuge/Deptford and Wilmington at the beginning of the early Late Prehistoric (about cal. A.D. 1000 - 1300) (Thompson 2006). During this period, there appears to be an increase in population size and complexity of social organization, evidenced by excavations at Bourbon Field, a village and mound site that is also located on Sapelo Island. The Savannah component of the site consists of heavier occupation zones when compared to the preceding Deptford period. These occupation zones are equally spaced and organized in a linear pattern, 5 suggesting more formal organization of space (Crook 1984). Excavations at Kenan Field have revealed evidence of what may have been the residence of a high-ranking official on a low platform mound. A structure that was likely the site of community gatherings was also excavated at the site (Crook 1980). During the Savannah phase, potters produced thin, coil-built vessels, averaging around 6.5 mm thick at Site 9Mc23 (Moore 2009:113). Savannah Figure 2. An Example of St. Simons Plain Fiber Tempered Pottery (UK6956). pastes tend to be fine, with grit/sand tempers (DePratter 1991). Check stamping is the most common surface treatment among Savannah sherds (Figure 4) recorded at Site 9Mc23, representing 75% of analyzed Savannah sherds as of 2009 (Moore 2009). Burnished plain, cord marked, and complicated stamped surface treatments have been recorded as well. St. Catherines Series ceramics, which are tempered with finer grog, also date to this period (DePratter 1991), but are not often differentiated Figure 3. An Example of Refuge/Deptford Check Stamped Pottery (UK6957). from Savannah by SIMPAP because there is an overlap between the radiocarbon dates for the two periods (Thomas 2008b) and both are found in the same contexts at Site 9Mc23 (Moore 2009) The Savannah series often exhibits curvilinear complicated stamping, similar to the stamping described on later Irene sherds, though these are uncommon at Site 9Mc23. These patterns were proFigure 4. An Example of Savannah Check Stamped Pottery (UK6829.1). 6 duced by carving curved designs into wooden paddles. The paddles were then pressed into cethe ramics surface before firing, leaving impressed designs on the surfaces of the pots (Saunders 2000). Savannah Complicated Stamped motifs include figure eights, figure nines, and concentric circles, which may include crosses in the center. Burnished interiors are common (DePratter 1991). The Irene phase begins around A.D. 1300 to 1350 and is commonly associated with the late Late Prehistoric occupation of the coast (about cal. A.D. 1300-1580) (Saunders 2000; Thomas 2008b). During the Irene phase, there is a noteworthy change in social organization. Between 1937 andn 1940, the WPA excavated a mound complex dating to the Savannah and Irene phases at the Irene Mound site, near Savannah Georgia. The first seven stages of the mound, which dated to the Savannah phase, were platform mounds typical of that time, where a chiefs residence would have been constructed. The final stage of the mound, which dated to the Irene phase, was rounded, rather than flat. The mound was converted from a chiefly residence to a shared burial space. The inhabitants also constructed a council house at the site during this period (Caldwell and McCann 1941). These changes have been interpreted as evidence of a shift towards a more egalitarian society, where decisions were made by a group, rather than an individual, and the chiefs position was somewhat less elevated, both literally and figuratively (DePratter 1991). Thompson (2009) argues that, while changes in mound construction and architecture at the site did result in more inclusive social spaces, this may be a reflection of changes in how leaders chose to display and reinforce their power, rather than a qualitative change in social organization. The structures at the top of the Savannah period mounds would have served to hide the actions of those in power, excluding others. Rituals and feasts held in council houses during the Irene period would have been for performances that served to reinforce power, just like exclusion. In both periods, the space is arranged to reinforce the positions of those in power, first through exclusion and later through performance. Thompson (2009) suggests that this change in methods of reinforcing power may have been triggered by a decline in agricultural productivity, based on the evidence of decreased maize consumption indicated by stable isotope analysis of individuals at the Irene site. These social changes were accompanied by changes in ceramic technology and design. Ce- 7 ramics became thicker than they were during the Savannah phase (averaging 7.3 mm at Site 9Mc23) and are grit or coarse sand tempered (Moore 2009:119; Saunders 2000). Changes in design elements during the Irene period include changes in complicated stamped motifs. During the Irene phase complicated stamped designs include curved lines, which are often part of the Filfot Cross motif (Figure 5). This motif consists of a raised central dot with lines radiating in four directions and ending in a curvilinear scroll (Saunders 2000). The design tends to be composed of lands and grooves that are narrow relative to those found on later Altama- ha ceramics. These complicated stamped motifs characterized 72.4% of analyzed Irene sherds at Site 9Mc23 as of 2009. Plain and incised sherds also were recorded at the site (Moore 2009:114). Irene jar rims tend to be decorated with a rim fillet, which is produced by applying a strip of clay below the vessels lip (Figure 6). This strip is often decorated with cane punctations (DePratter 1991). It was during this period that the Spanish began their initial explorations of what is now the east coast of the United States and when they first came into contact with the Guale. After a number of exploratory expeditions, the Spanish established a permanent settlement at St. Augustine in 1565 (Jefferies and Moore 2009). The northernmost Spanish town of Santa Elena was established in the following year on Parris Island, off the coast of South Carolina (DePratter 2009). Soon after, Figure 5. An Example of an Irene Rim Fillet with Whole Cane Punctations (UK6817.2). Figure 6. The Filfot Cross Motif. Adapted from Saunders (2000). 8 Figure 7. A Map of the Georgia Coast Depicting the Locations of Spanish Missions during the Mid-17th Century. Adapted from Worth (2007:11) 9 the Jesuits established missions in Guale towns along the coast (Worth 2007). At this time the Guale were organized in three stratified chiefdoms, located in three different provinces. The leadership of these chiefdoms alternated between micos in different towns in the province, who held the highest positions of authority. Guale society was matrilineal and there were many instances of female leadership in the 17th century (Thomas 2008c). Guale towns were occupied year-round, and their residents relied on agriculture, as well as marine resources, for subsistence. This combination of foraging and farming meant that the Guale would occasionally leave their towns temporarily in order to exploit certain resources. This may explain conflicting reports from Europeans regarding Guale mobility (Thomas 2008a, 2008c). Reliance on agriculture increased as the Guale were incorporated into the mission system. It is likely that the sabana system, which required the Guale to grow crops for local consumption as well as export, left them with less time to fish and forage and resulted in a more sedentary lifestyle (Thomas 2008d). The stress caused by the mission systems labor draft, as well as epidemics, caused satellite villages to relocate to larger towns, resulting in larger, more centralized towns (Worth 2007). Eventually, during the later Altamaha period, external pressures caused the mission system to contract. Raids conducted by the English-allied Westo beginning in the 1660s forced the Spanish and Guale to relocate to the barrier islands to the east of their initial settlements, creating blended communities that included refugees from the interior, the Spanish, and displaced Guale. Shortly after, in 1683, raids conducted by English and French pirates forced the Spanish and Guale to move yet again, this time retreating to the south, with all groups eventually settling on Amelia Island and the adjacent area around St. Augustine (Worth 2009). Near the beginning of the Mission period, around A.D. 1580, a new ceramic series called Altamaha develops (Saunders 2000). During the Altamaha period, vessel walls become slightly thicker, with an average maximum body thickness of 8.2 mm reported by SIMPAP at Site 9Mc23. Grit/sand is the most common temper, but this series does exhibit a greater diversity of temper recipes (Moore 2009:142). 10 Altamaha sherds are characterized by a rectilinear variation on the Filfot Cross, known as Line Block (Figure 8). This motif consists of a raised central dot surrounded by straight, rather than curved, lines radiating in four directions. These lands and grooves tend to be relatively wide, compared to complicated stamped Irene sherds (Moore 2009; Saunders 2000). Complicated stamped and overstamped sherds both have been Figure 8. Line Block Motif. described in the Altamaha assemblage from Site 9Mc23 (Moore 2009). Overstamped ceramics are produced by stamping a design into wet clay, rotating the paddle, and stamping over the design, obscuring the original motif. This is often done with paddles carved with complicated designs, which may or may not completely obscure the original design. However, cross simple stamping, where the design is completely obliterated, may also be Figure 9. An Example of Cross Simple Stamping (UK6266). produced using paddles carved with a series of strait lines oriented in the same direction. In addition to complicated and overstamping, there is a resurgence of check stamping during the Altamaha phase, which is uncommon during the previous Irene phase (Saunders 2000). Other surface treatments include plain, incised, and red filmed (Moore 2009). Rims are often finished with Figure 10. Altamaha Rim Fold/Strip Decorated with Fingernail Punctations (UK6263.3). a rim fold/strip. This rim form may be the result of 11 the same process that creates the Irene rim fillets, or it may be the result of the potter building up the vessel wall and folding it over at the top. The bottom of the fold is often decorated with fingernail or cane punctations (Saunders 2000). It is unclear which of the changes between Irene and Altamaha occurred first. Some sherds displaying traits of both Irene and Altamaha ceramics, which are thought to be transitional, have been recovered in previous excavations at Site 9Mc23. Sherds identified as possibly transitional often have mixed wide and narrow lands and grooves (an Altamaha trait) in combination with cur- vilinear stamping (an Irene trait). We recovered sherds that have Irene motifs on Altamaha pastes. The study of these sherds may lead to a better understanding of which changes in the series occurred first. A comparison of traits such as temper and construction techniques could allow for a better understanding of other less commonly investigated differences between the two series. It may also contribute to a better understanding of how the transition occurred and possibly what caused changes in Guale ceramic production. Debates Over Spanish Influence The arrival of the Spanish undoubtedly had an impact on the Guale of coastal Georgia, but the importance of that impact and the degree to which it influenced ceramic technology and design is a matter of some debate. Assumptions of Spanish dominance in their interactions with the Guale have often been extended into questions about changing ceramic technology. Saunderss (2000) comparison of one Late Irene and two Mission period sites suggests that the level of interaction with the Spanish and the degree of pottery change are related. She has gone so far as to suggest that the Spanish friars may have been actively involved in the process, directing native producers to change certain elements. Later, she suggests that the Guale altered their methods to make their pottery more appealing to the Spanish, an interpretation that attributes change to market forces (Saunders 2009, 2012). The replacement of the Filfot Cross with the Line Block motif is one of the changes that has been interpreted as the result of Spanish interaction. The Filfot Cross is symbolic of the Guale worldview, and Saunders (2000) has states that changes in the Guale worldview caused by their 12 incorporation into the Spanish mission system led them to alter this design. She has also proposed that the retention of the four fields design, albeit in a different form, was an act of resistance on the part of the Guale. If the Spanish friars recognized the Filfot Cross as a cosmological symbol and disapproved of its use, the Guale may have found ways to alter it, while still retaining its meaning (Saunders 2009). Either way, the change is not interpreted as originating with the Guale or other indigenous groups, but with the Spanish colonizers. However, it should be noted that there were Guale groups who were not immediately incor- porated into the mission system (Saunders 2000) and that non-mission Guale sites have received little attention from archaeologists (Deagan and Thomas 2009). Saunderss (2000) assertion that pottery change is related to interaction with the Spanish is based on a comparison of one late Irene site and two Altamaha mission sites. Without knowing what kind of pottery was being produced outside of the mission system at the same time, it is impossible to say that the two factors are directly related (Moore and Jeffereies 2014; Worth 2009). However, investigations of this kind are somewhat complicated by the fact that mission period sites are primarily identified by the presence of European artifacts (Deagan and Thomas 2009). The assumption that the Spanish would have had such a heavy influence on the Guale has been contested. In fact, in many instances, it seems that the opposite was true. In many of the areas colonized by the Spanish there is evidence of heavy indigenous influence in architecture and town planning (Voss 2008). It also seems that the Spanish adopted a primarily indigenous diet in many colonies, relying on locally hunted, gathered, and cultivated resources rather than European domesticates, a pattern that archaeologists have observed on both Sapelo Island and St. Catherines (Moore and Jefferies 2014; Reitz and Dukes 2008). Archaeological investigations have shown that the Spanish used Native American ceramics, not just as utilitarian wares, but in public spaces, as well (Deagan 1983; Voss 2008). Due to the fact that Spanish St. Augustine, and the rest of La Florida, existed as a periphery in the world system of the day, it was likely difficult for the residents of this region to obtain the Spanish goods to which they were accustomed, including pottery. To supplement what they were 13 able to get from non-Spanish European countries, they would have turned to what was locally available (Hughes 2012; Voss 2008). In some ways, this likely influenced pottery production, as in the case of colonowares, which clearly imitate Spanish vessel forms (Saunders 2000; Moore and Jefferies 2014), and Saunders (2009) states that the increased diversity in decoration from Irene to Altamaha may have been an attempt to make Guale pottery more appealing to the Spanish. However, it is clearly the case that the Spanish had to adopt many aspects of Guale culture and that the Guale may have had a greater cultural influence on the Spanish living in La Florida than the Span- ish had on the Guale (Moore and Jefferies 2014). It is especially difficult to imagine that the Spanish could have had such a profound influence on the Guale at the time of the development of the Altamaha series. The new ceramic style developed after the Spanish abandoned their early settlement at Santa Elena, which suggests that the Spanish were likely not a significant influence at the time of its initial adoption (Moore and Jefferies 2014; Worth 2009). Worth (2009) proposes that Altamaha pottery developed in situ in the northern region of the Spanish mission system before or very soon after the arrival of the Spanish, when their influence would not have been at its peak. Based on the absence of Altamaha ceramics at Santa Elena, and its presence at Santa Catalina on St. Catherines Island, Saunders (2009) suggests that this transition occurred quickly, within a single generation of potters. The earliest dates for Altamaha ceramics occur in the northern portions of the region, with later dates at sites farther to the south, such as at St. Augustine (Worth 2009). For this reason Worth (2009) argues that the Altamaha series was carried southward by Guale migration. However, a combination of ethnohistoric and archaeological investigations show that the dates for the adoption of Altamaha ceramics in the southern regions of the mission system do not correspond with the dates of the arrival of the southward migration of the Guale and Spanish (Worth 2009). It also seems unlikely that the Spanish actively promoted Altamaha ceramics. The presence of multiple ceramic types within the Spanish mission system suggests that the Spanish were not directly promoting any single pottery style. While it is true that Altamaha spread southward along the coast, replacing other ceramic types as it moved through the mission system, it did not spread 14 to the western portion of the Spanish mission system in Florida. The western portion of the mission system experienced a similar phenomenon, with Jefferson ceramics coming to dominate the regions ceramics assemblages in the same way that Altamaha dominated ceramics assemblages in the east (Worth 2009). The fact that Altamaha spread only through part of the mission system suggests that the Spanish did not promote the new ceramic style. The concept of interaction spheres that extend along the coast also has been used to explain the pattern of the spread of Altamaha from north to south, accounting for the prevalence of Altama- ha ceramics along the eastern coast of Florida and Jefferson along the west coast (Worth 2009). This explanation suggests that interactions between indigenous groups, rather than contact with the Spanish, are primarily responsible for the development and spread of the Altamaha series, though the presence of the Spanish likely had some influence. So far, these explanations suggest that the Spanish were responsible for the development and spread of Altamaha ceramics, either through direct promotion or relocation and influence on Guale worldviews. Another more satisfying explanation for changes in pottery production is that it was part of a process that was in motion before the arrival of the Spanish. It is important to remember that Guale culture was not static before Spanish arrival and that the Spanish form only one part of the historical context in which these changes occurred (Moore and Jefferies 2014). Guale society was in a state of change before the arrival of the Spanish, with changes in political organization during the Irene period illustrated by the capping of the Irene mound (DePratter 1991; Thompson 2009) and a general trend toward increased sedentism, social complexity, and population growth throughout the history of the occupants of the Georgia coast (Crook 1984). Guale society was already in a state of change when the Spanish arrived. The development of Altamaha pottery can be seen as a continuation of these processes. When Altamaha ceramics are compared to other Lamar ceramics from the interior, it becomes clear that many of the traits that are new to Guale ceramics could have come from interactions with interior groups that were not a part of the Spanish mission system. Both Altamaha-like rim forms and rectilinear stamping are found at interior sites. Many other traits that are considered new to Altama- 15 ha can be found on earlier ceramics and appear to be revivals of old techniques rather than innovations. Red filming, for example, is commonly associated with mission sites and is considered a trait that is new to Altamaha and may indicate Spanish influence. However, red filming has been described at Woodland and Late Prehistoric sites in Georgia prior to Spanish contact. Based on these similarities and revival of old techniques, it seems more likely that Altamaha developed out of interactions with interior groups than as a result of Spanish influence. Further investigations focused on sourcing would allow for a better understanding of the significance of interactions between coastal and interior groups (Moore and Jefferies 2014). There is a resurgence in check stamping during the Altamaha period as well, but this trait is not new either. It was a fairly common surface treatment during the Savannah period and is present on non-Guale ceramics along the coast during the Irene period. SIMPAP has reported low frequencies of check stamped Irene sherds, but this is not common along the coast (Semon 2015). The revival of this surface treatment may be the result of interactions with other coastal groups who practiced check stamping (Jefferies and Moore 2009; Moore and Jefferies 2014). Understanding Changes in Technical and Decorative Style There are still a lot of questions surrounding the transition from Irene to Altamaha, especially with regards to the cause of the changes in technology and design. Investigations that involve the excavation of non-mission sites dating to the Altamaha period and sourcing studies that focus on identifying interactions between coastal and interior groups could shed some light on these questions, but a detailed attributes analysis may also be helpful. Technical style, as well as decorative style, has important cultural meanings, and the study of both can reveal the structures behind patterns of human behavior. The act of pottery production would have communicated cultural values, in both its technical and decorative aspects (Lechtman 1977), suggesting that changes in this process would likely reflect changing Guale worldviews, as proposed by Saunders (2000). It is likely that if change in the technical or decorative aspects of pottery production were encouraged by the Spanish, these changes would be to the traits that are either highly visible, likely 16 causing them to look more Spanish, or that improve the function of the vessels in some way. The Spanish would have had little incentive to actively change Guale pottery unless it benefitted them in some way. They might have made changes that made the pottery more visually appealing to them or altered the symbolic meaning of the pottery. Or they may have introduced methods that made the pottery more functional, such as higher firing temperatures (Saunders 2000, 2009). These elements might include slip, decoration, vessel form, and/or surface treatment. However, changes in less visible traits, with little impact on performance, would seem to indicate a lack of Spanish involvement, as the Spanish would have little incentive to change these elements. Less visible elements might include temper recipes, thickness, and perhaps subtle changes in surface treatment, such as land and groove width. A better understanding of the nature and timing of this transition may help to clarify some of the issues surrounding the cause of Guale ceramic change, as well as possibly allowing for the refinement of the ceramic chronology for the Irene/Altamaha transition. 17 Field and Lab Methods Excavation and Ceramic Attributes Analysis All excavations were performed following procedures developed for SIMPAP (Jefferies and Moore 2009). Units were primarily placed as part of SIMPAPs sampling strategy, based on the location of positive shovel tests during a systematic site-wide survey. Unit 42 was a sample unit placed over a soil core anomaly that indicated the presence of a shell-filled pit. Unit 43 was an extension of this unit, placed to excavate the south half of a second shell-filled pit. The location of Unit 46 was determined based on high concentrations of Altamaha pottery and a single sherd of redware recovered during shovel test survey. Excavation of this unit revealed what is thought to be an Irene structure floor and Unit 50 was placed to further investigate this feature. Unit 48 was excavated to test a shell midden pile located at the southeastern edge of the site. The location of Unit 48 was chosen to investigate a shell midden pile located during shovel test survey, which yielded a large sherd with both Irene and Altamaha traits. Unit 51 was an extension of this unit, which was excavated in order to remove a Mission period pig burial that was partially located within Unit 48. Most excavations were carried out in 10 cm levels, except where features were encountered and excavated separately. Features were excavated as a single element unless there were differences in soil color or texture that justified treating them as separate zones. The exception to this rule was Unit 51, an extension of Unit 48, which was excavated in cultural strata to allow for the quick removal of pig skeletal remains that had to be left partially exposed while the upper levels of the new unit were removed. This also provided a stratigraphic control for dating the ceramics from the shell midden. When possible postmolds were encountered, they were bisected to reveal a profile that would allow excavators to confirm their function. Floatation samples were taken from 25 x 25 cm columns at the corner of each unit, and bulk float samples were taken from features that excavators thought were likely to contain botanicals. All artifacts, including ceramics, collected from these excavations were carefully cleaned and analyzed following SIMPAP protocols (Jefferies and Moore 2009). Ceramics were first sorted by size, which determined the amount of data collected for each sherd. For sherd size 1 (x < 1 cm2) 18 N Figure 11. Locations of Excavated Units. 19 and sherd size 2 (1 < x < 4 cm2) only counts and weights were recorded. Surface treatment, decoration, temper, and thicknesses were recorded for size 3 (x > 4 cm2) sherds. Multiple measurements were taken with digital calipers to determine the maximum thickness for each rim and body sherd. For both size 3 (x > 4 cm2) and partial vessels, vessel form also was recorded. Sherds size 3 and larger with both Irene and Altamaha traits were noted as possible transitional sherds. Unlike a traditional Southeastern typological analysis, SIMPAP analysis codes the ceramic series as an attribute, in order to allow for investigation of within-series variation. In addition to the types commonly used in this region, a number of varieties within the Altamaha series were defined in order to conduct a more detailed study of the diversity of temper recipes that developed during that period (Figure 12). Eventually we hope to determine whether there is a temporal element to these varieties. The most common of these varieties is Altamaha, which is characterized by grit/sand temper and fits the standard definition of the type (DePratter 2009). The Green variety has the same paste, but includes only sherds with a fine check stamped surface treatment. The Belle Marsh variety is defined by shell temper (Figure 12a) and the Darien variety has a finer, sandier paste than Altamaha (Figure 12b).The definition of the Mud River variety is based on the inclusion of large pieces of quartz with iron oxide films (Figure 12c). The Sapelo variety is defined by clay temper (Figure 12d). The definition of the Caldwell variety is based on vessel form and decoration and includes bellshaped bowls with paired hemiconical punctates. The Lamar variety is based on similar criteria and consists of cazuela bowls with well-executed Lamar motifs. If these vessels have fine pastes, they are considered to be part of the McIntosh variety. Thin Section Analysis The investigation of temper was carried out through macroscopic analysis of the broken edges of sherds and the analysis of a sample of thin sections. Thin sectioning allows for a more detailed study of temper and construction techniques. The thin sections were studied microscopically, using the point counting method to allow for quantitative analysis (Stoltman 2001). Data were recorded at points along a 1mm x 1mm grid and included temper, inclusions, paste, and voids. This was not a 20 true petrographic analysis, using various spectrums of light to identify the make up of the sherds, so it was not possible to identify most inclusions that were smaller than of a millimeter. For this reason, anything smaller than of a millimeter was considered to be part of the paste. It is unlikely that materials of this size were included intentionally, so their exclusion was not a major loss, as this analysis is more concerned with intentional changes in temper recipes. Sand and grit also were differentiated based on size and in keeping with SIMPAP protocols, where sand is composed of quartz grains of less than 1 mm and anything larger is grit. Clay was identified when its color did not match Shell A B Clay Quartz C D Figure 12. Some Altamaha Varieties. A) Belle Marsh, B) Darien, C) Mud River, D) Sapelo. Temper is too fine to be labeled for the Darien variety. 21 the majority of the sherds paste, and sherd inclusions were identified based on their angular shape or the inclusion of temper. Percentages were calculated based on this point counting sample. Twenty sherds were selected for this process: six Irene, seven Altamaha, and six possible transitional sherds, as well as two Savannah sherds for comparative purposes. Sherds selected for this analysis included at least one rimsherd from each ceramic type, in order to compare rim construction techniques between the types. However, there was no observable difference in this aspect of the thin sections. Sherds selected from the Altamaha series include a sampling of different varie- ties. This method was used to test for temper variability within types, as well as to test the varieties described above. Scanning Electron Microscope Analysis This study used the SEM to identify sherds that have true slips and compare slip technologies from different series. Slips can be identified using the SEM because they have finer particles than the body of the sherd, resulting in what appears to be a thin, smooth layer resting on top of a much larger and rougher one. The sherds selected for SEM analysis were embedded in epoxy and cut into thick sections to provide samples small enough to place in the SEM. The use of epoxy ensured that the possible slips came all the way up to the edge of the sherd. Cutting the thick section on both sides also ensured a more standardized orientation. Sherds were imaged at multiple magnifications in order to ensure that slips of varying thicknesses would not be missed. Sherds with red filming, a known slip, were examined under the SEM in order to determine that the slip could be differentiated from the sherd body and to confirm that it did not resemble burnished or plain surfaces. Other possible slipped sherds were then examined in order to determine whether they were truly slipped or merely discolored by oxidizing or reducing conditions during firing. One blackened and one white/tan slipped sherd were selected from both the Irene and Altamaha series for comparison. An additional Altamaha sherd with a possible orange slip, which was not present among Irene sherds, also was selected. Unfortunately, no possible transitional sherds in this assemblage had possible slips, so they could not be included in the sample. This allows for the identification of true slips and possibly the identification of changes in slip technologies through time. 22 Results of Investigations Excavations Units 42 and 43 Unit 42 was a 2 x 2 m sample unit located near a block excavation carried out by SIMPAP from 2006 to 2011. The purpose of this block excavation was to investigate what is believed to be a Spanish structure dating to the Mission period (Moore and Jefferies 2016). When excavated, Unit 42 revealed two pits, one filled with shell (Feature 50) and one filled with midden (Feature 51), and two possible post features (Features 54 and 55). Two features were likely either posts or pits (Features 52 and 53). In order to excavate all of Feature 50, Unit 43, a 0.5 x 1 m unit, was added to the south. The shell-filled pits discovered during this block excavation resemble those described at other Spanish colonial sites and are likely the result of Spanish waste disposal (Moore and Jefferies 2016). This is supported by the presence of glass beads (Figures 16 and 17), coarse earthenware (Figure 18), porcelain (Figure 19), majolica pottery (Figure 20), and lead shot (Figure 21), along with faunal remains in Feature 50. The ceramics from these features provide a sample of Altamaha ceramics associated with the Mission period. AMS dating of a maize specimen recovered from Feature 50 returned a date of 160 +/- 25 B.P. and a Bayesian modeled date of 1660 to 1685 (Table 11). Figure 13. Plan View of Units 42 and 43. Created by Dr. Christopher Moore. 23 Figure 14. Unit 42 West Profile. Photo taken by Dr. Richard Jefferies. Figure 15. Unit 43, South Portion of Feature 50. Photo taken by Dr. Richard Jefferies. 24 Figure 16. Glass Beads from Feature 50 in Unit 42. Photo taken by Dr. Richard Jefferies. Figure 17. Glass Beads from Unit 42. Photo taken by Dr. Richard Jefferies. 25 B A C Figure 18. Coarse Earthenware from Unit 42. A: Level 1, B: Level 2, C: Feature 50. Photo taken by Dr. Richard Jefferies. E B C A D Figure 19. Porcelain from Unit 42. A: Level 1, B:-D: Level 2, E: Level 3. Photo taken by Dr. Richard Jefferies. 26 A C B D E F Figure 20. Majolica Pottery from Units 42, 46, and 50. A: Unit 42 Level 1, B: Unit 42 Level 2, C: Unit 42 Level 4, D: Unit 46 Feature 60, E: Unit 50 Level 2, F: Unit 42 Feature 50. Photo taken by Dr. Richard Jefferies. Figure 21. Lead Shot from Unit 42. Photo taken by Dr. Richard Jefferies. 27 28 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Feature 51 Feature 52 Feature 53 Feature 54 Feature 55 Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Feature 51 Feature 52 Feature 53 Feature 54 Feature 55 Total 0 0 1 0 0 0 0 0 0 0 0 1 11 37 28 6 1 13 0 0 0 0 0 96 9 73 142 76 7 326 4 9 11 6 6 669 Groundstone Tool Fragment 23 60 66 78 11 58 5 24 3 12 2 342 0 3 12 0 0 335 49 0 1 0 0 400 Bone Glass Bead Bivalve Antler Table 1. All Artifacts from Unit 42 1 1 3 0 0 1 0 0 0 0 0 6 Lead Shot 7 0 7 0 0 8 0 0 0 11 9 42 Charcoal 2 2 0 1 0 2 0 0 0 0 0 7 Majolica 2 1 0 0 0 2 2 0 0 0 0 7 Coarse Earthenware 33 64 16 2 0 2 0 0 0 0 0 117 Metal Fragment 0 0 1 0 0 0 0 0 0 0 0 1 Core 1 1 2 0 0 4 0 0 1 0 0 9 Otolith 0 0 0 0 0 4 0 0 0 0 0 4 2 1 2 0 0 5 0 2 0 0 0 12 1 3 1 0 0 0 0 0 0 0 0 5 0 1 1 0 0 1 0 0 0 0 0 3 Flake 1 3 1 1 0 41 0 1 2 0 0 50 Gastropod 57 145 79 64 16 62 6 11 6 6 0 452 85 170 94 30 19 125 7 11 14 1 0 556 Prehistoric Prehistoric Ceramics Ceramics (< 1 cm2) (1 < x < 4 cm2) Fired Clay/ Daub Porcelain Crustacean 29 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Feature 51 Feature 52 Feature 53 Feature 54 Feature 55 Total Rock 0 0 0 0 0 2 0 0 0 0 0 2 Prehistoric Ceramics (> 4 cm2) 14 41 50 27 2 72 6 1 2 1 1 217 Table 1. Continued. All Artifacts from Unit 42 0 0 1 0 0 0 0 0 0 0 0 1 Shell Bead 0 5 15 4 0 0 0 0 0 0 0 24 Tooth 0 8 9 7 0 0 0 0 0 0 0 24 Turtle Shell 1 8 4 3 0 0 0 0 0 0 0 16 Whelk Shell 0 1 0 1 1 2 0 0 0 0 0 5 UID Botanical 1 0 0 0 0 0 0 0 0 0 0 1 UID Glass 2 0 0 0 0 0 0 0 0 0 0 2 UID 30 Irene Complicated UID UID Altamaha Stamped Lamar Altamaha Plain 0 0 3 0 2 0 4 1 0 1 3 3 1 0 1 1 0 0 0 0 0 0 13 1 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 4 1 25 7 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Feature 51 Feature 52 Feature 53 Feature 54 Feature 55 Total Refuge 0 0 1 0 0 0 0 0 0 0 0 1 Altamaha Burnished Plain 0 1 2 0 0 0 0 0 0 0 0 3 Refuge/Deptford Plain 0 0 0 2 0 0 0 0 0 0 0 2 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Feature 51 Feature 52 Feature 53 Feature 54 Feature 55 Total St. Simons Incised 0 0 1 0 0 0 0 0 0 0 0 1 St. Simons Plain 0 0 0 2 0 0 0 0 0 0 0 2 Table 2. All Ceramics from Unit 42. Wilmington 0 0 1 0 0 0 0 0 0 0 0 1 Savannah Burnished Plain 0 0 0 1 0 0 1 0 0 0 0 2 Altamaha Check Stamped 0 1 0 0 0 1 0 0 1 0 0 3 Savannah Check Stamped 0 1 1 0 0 1 0 0 0 0 0 3 Altamaha Complicated Altamaha Altamaha Stamped Overstamped Red Filmed 3 4 1 12 16 1 6 14 2 2 2 1 2 0 0 27 36 4 1 1 0 0 1 0 0 0 1 0 0 1 0 0 0 53 74 11 Deptford Check Stamped 0 0 1 0 0 0 0 0 0 0 0 1 UID 3 2 4 0 0 2 2 0 0 0 0 13 31 Deptford Cordmarked Level 1 (0-20 cm BD) 0 Level 2 (20-30 cm BD) 2 Level 3 (30-40 cm BD) 0 Feature 50 0 Total 2 St. Johns Punctate 0 0 0 1 1 Irene Complicated Stamped 0 0 0 1 1 Otolith 0 0 0 0 0 7 7 Tooth 0 1 1 0 0 4 6 Nutshell 0 1 0 0 0 0 1 Altamaha Red Filmed 0 0 1 3 4 UID 1 0 0 0 1 Whelk Shell 0 1 0 0 0 7 8 Metal Fragment 0 0 1 0 0 1 2 Turtle Shell 0 1 0 0 0 12 13 Gastropod 0 0 0 0 0 12 12 Altamaha UID Complicated Altamaha Altamaha Stamped Overstamped 0 0 2 1 0 0 0 0 0 3 17 8 4 17 10 Slag 0 1 0 0 0 0 1 Crustacean 0 0 0 0 0 3 3 Shell Bead 0 0 0 0 0 1 1 Coarse 2 1 0 0 0 0 3 Prehistoric Ceramics (> 4 cm2) 3 2 1 0 0 3 9 Charcoal 3 3 0 0 3 3 12 Prehistoric Prehistoric Ceramics Ceramics (< 1 cm2) (1 < x < 4 cm2) 4 8 33 19 16 11 3 1 0 1 4 8 60 48 Table 4. All Ceramics from Unit 43. Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 50 Total Bone 307 0 31 23 3 1 365 Bivalve 2 7 2 2 0 0 13 Antler 0 8 54 0 0 2 64 Table 3. All Artifacts from Unit 43. Units 46 and 50 Unit 46 was placed as part of SIMPAPs sampling strategy. Excavations began in February 2015, but were not completed until the following August. During the excavation of Unit 46, we identified a large feature that continued into the units north wall. We halted excavations until a 2 x 2 meter extension, Unit 50, could be placed to the north. Majolica (Figure 20) and some coarse earthenware (Figure 33), along with Altamaha ceramics were found in the upper levels of the unit, but the most interesting items dated to the Irene period. In Unit 46, excavators identified Feature 60, a basin-like depression in the north half of the unit (Figures 16-18). It was identified primarily based on dark soil coloration, which, at the time, resulted in well-defined feature boundaries. Artifact density was rather low, but artifacts recovered included primarily Irene sherds, whelk, a chert flake, and an anthropomorphic pipe (Figure 19). Several concentrations of unfired clay were noted in this portion of the feature, as well. A biface also was found in this area (Figure 32). When we expanded the excavation in July, our crew faced much warmer and drier soil conditions, which made the identification of Feature 60s northern boundaries in the sandy soil nearly impossible. However, we noted areas of dark and mottled soil with slight increases in artifact densities in the southern half of the unit that likely represent the northern portion of Feature 60. Along with the basin-like feature, we identified several posts (Feature 60 Post and Features 61, 65, and possibly 62). One of these (Feature 60 Post) was situated directly in the southern half of Feature 60, and another is likely located within or just outside of the features amorphous northern boundary. Three additional features that may be posts were identified in the units western profile, which seems to lie within the feature. Based on the shape of the feature and the increased, but relatively low, density of artifacts within Feature 60, it may represent a living surface or structure floor. It seems likely that at least some of the posts are associated with the feature, which would indicate that it may have been the location of a structure. The presence of a structure may account for the unfired clay that was recovered from the southern portion. The ceramics recovered from this unit indicate that the possible structure dates to 32 the Irene Figure 22. Composite Plan View of Unit 46. A clear outline of the southern portion of Feature 60 is visible. period. Figure 23. Plan Views of Unit 50. Soil conditions were poor during this excavation, making feature boundaries difficult to identify. In these two maps, drawn at 35 and 40 cm BD, the possible location of the northern portion of Feature 60 is visible as areas of very dark brown soil in the southern portion of the unit. 33 Figure 24. Unit 46 West Profile. Figure 25. Unit 50 West Profile. 34 Figure 26. Feature 60 in Unit 46. Photo by Richard Jefferies. Note that north is down. Figure 27. Feature 60 Profile. 35 Figure 28. Feature 60 Post in Profile Figure 29. Feature 61 at D = 40 cm BD. 36 Figure 30. Feature 62 at D = 40 cm BD. Figure 31. Anthropomorphic Pipe from Feature 60. 37 Figure 32. Biface from Feature 60 (UK 6721). Photo taken by Dr. Richard Jefferies. B A C Figure 33. Coarse Earthenware from Units 46, 48, and 51. A: Unit 46 Level 2, B: Unit 48 Level 2, C: Unit 51 Level 2. Photo taken by Dr. Richard Jefferies. 38 39 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 60 Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 60 Total 0 0 1 0 0 0 1 Otolith 0 0 0 0 0 1 1 Biface Table 5. All Artifacts from Unit 46. 58 49 46 5 0 4 162 Prehistoric Ceramics (< 1 cm2) 46 60 27 1 2 3 139 Bivalve 2 15 1 0 0 10 28 120 116 53 7 2 7 305 0 1 0 0 0 0 1 10 25 17 4 3 3 62 0 0 0 0 0 1 1 Cut Bone 0 9 0 0 0 0 9 Turtle Shell Coarse Earthenware Prehistoric Ceramics (> 4 cm2) Charcoal Prehistoric Ceramics (1 < x < 4 cm2) 6 4 9 1 0 10 30 Bone 6 12 4 0 0 2 24 0 0 0 0 0 1 1 0 0 0 0 0 1 1 1 0 0 0 0 0 1 Metal Fragment 0 0 9 1 0 0 10 Possible Red Ocher Majolica UID Botanical 0 1 0 0 0 1 2 Gastropod Whelk Shell 0 1 0 0 0 1 2 Flake 40 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 60 Total Savannah Plain 1 1 0 0 0 0 2 Altamaha Complicated Stamped 3 12 6 2 2 0 25 St. Simons Incised 0 0 0 0 1 1 2 Altamaha Irene Complicated Irene UID Burnished Stamped Incised Altamaha Plain 3 0 0 0 7 0 1 1 5 1 0 2 0 0 0 0 0 0 0 0 1 0 0 0 16 1 1 3 St. Simons Punctated 0 0 1 0 0 0 1 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 60 Total St. Simons Plain 0 1 2 2 2 0 7 UID St. Simons 0 0 3 2 0 1 6 Table 6. All Ceramics from Unit 46. Altamaha Overstamped 4 16 14 2 0 0 36 Savannah Burnished Plain 0 2 0 0 0 0 2 Altamaha Red Filmed 1 1 2 1 0 0 5 Altamaha Check Stamped 0 1 0 0 0 0 1 UID 3 2 4 0 0 0 9 Savannah Check Stamped UID Irene Irene Plain 2 0 0 5 2 1 4 0 0 0 0 0 0 0 0 0 0 0 11 2 1 41 1 2 0 0 0 3 Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Total Otolith Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Total Biface 0 0 1 0 0 1 Table 7. All Artifacts from Unit 50. 1 59 3 17 28 10 2 316 18 30 143 123 1 77 3 3 30 40 Bivalve Bone Charcoal Flake Flake Tool 7 1 8 0 0 44 19 0 1 0 46 7 44 1 0 4 4 1 1 1 2 1 13 1 0 103 32 66 4 1 Prehistoric Ceramics Prehistoric Ceramics Prehistoric Ceramics (< 1 cm2) (1 < x < 4 cm2) (> 4 cm2) 0 1 0 0 1 0 Rock 0 1 1 0 0 0 Shatter Gastropod 1 0 0 0 0 1 0 3 0 0 1 2 Tooth 1 1 0 0 0 0 UID Olive Jar 2 1 0 0 0 3 42 Level 1 (0-21 cm BD) Level 2 (20-31 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-63 cm BD) Feature 61 Feature 62 Total Level 1 (0-21 cm BD) Level 2 (20-31 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-63 cm BD) Feature 61 Feature 62 Total UID St. Simons St. Simons Plain Refuge 0 0 0 0 0 0 0 1 1 0 0 1 1 0 0 0 0 0 0 0 0 1 1 2 Irene Complicated UID Altamaha stamped Altamaha Plain 0 1 0 2 2 1 8 4 0 0 0 0 0 0 0 0 0 0 0 0 0 10 7 1 Table 8. All Ceramics from Unit 50. Savannah Savannah Savannah Savannah Plain Burnished Plain Complicated Stamped Check Stamped 0 0 0 0 0 0 0 1 2 1 1 2 1 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 3 2 1 4 Altamaha Complicated Altamaha Altamaha Irene/ Stamped Overstamped Incised Altamaha UID 0 1 0 0 1 11 7 0 0 5 4 9 1 5 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 16 17 1 5 7 UID Irene 0 1 0 0 0 0 0 1 Unit 47 Unit 47 was placed in a 40 cm thick linear midden deposit, located on the southeastern periphery of the site. Along with clam, oyster, and whelk, Levels 1 and 2 contained mostly Irene ceramics. Shell and bone deposits reach their peak in Level 3, which contains only Savannah ceramics. A decrease in the amount of shell, bone, and ceramics begins in Level 4. Ceramics from levels 3 and 4 were consistent with what would be expected in a Savannah context and suggest that the area was heavily utilized during the Savannah period. With the exception of a small area of surface bioturbation, no other features were recorded in this unit. Figure 34. Unit 47 Shell Midden East Profile. 43 Figure 35. Unit 47 East Profile Figure 36. Unit 47 Plan View at D = 50 cm BD. 44 45 8 243 1 145 53 450 Bone 13 22 16 2 0 53 0 1 0 0 1 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Total UID Savannah 3 1 1 1 0 0 0 0 18 1 19 0 0 0 3 0 3 1 0 1 0 0 Savannah Burnished Plain 5 8 6 2 0 21 0 1 0 0 0 1 Tooth 0 1 0 0 0 1 Flake 5 0 5 0 0 Savannah Check Stamped Crustacean Prehistoric Ceramics (> 4 cm2) Charcoal Savannah Plain Prehistoric Ceramics (1 < x < 4 cm2) 0 3 0 0 0 3 Table 10. All Ceramics from Unit 47. Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Total Bivalve Table 9. All Artifacts from Unit 47. 1 0 0 1 0 2 0 0 0 2 Savannah Cob UID Impressed Irene 0 39 0 15 0 54 Turtle Shell 0 0 0 14 0 14 Gastropod 1 0 0 0 1 Irene Plain 0 7 0 5 0 12 Otolith 3 0 0 3 0 1 0 0 1 0 Irene Complicated Stamped Irene Incised 2 1 0 3 1 7 Prehistoric Ceramics (< 1 cm2) 3 1 0 0 2 UID Units 48 and 51 Units 48 and 51 were placed to investigate one of SIMPAPs shovel test probes. Excavation of STP 346 led to the recovery of a particularly interesting sherd with both Irene and Altamaha traits, which seemed to be located in a shell midden. The slightly flared jar is complicated stamped with mostly rectilinear designs, with only a single curvilinear element visible at the top right corner of the sherd (Figure 37). The sherd has narrow lands and grooves and stamping on the lip, which are Irene traits. It has a grit/clay temper and a heavily burnished interior, traits that are common among Altamaha ceramics from the site. Other mixed trait sherds have been recovered at the site - Figure 37. Possible Transitional Sherd Recovered (UK6100) from STP 346. this one is simply the largest example to date. We placed a 2 x 2 meter unit near the location of the probe in order to further investigate what we hoped might be a transitional context. The discovery of a pig burial forced us to expand our excavations with the addition of Unit 51, a 1 x 2 meter unit to the south. The upper levels of these excavations caught the edge of a shell midden pile that was identified in the previously mentioned shovel probe. The upper levels of the shell midden contained primarily Altamaha ceramics, with a couple of sherds that were identified as possibly transitional. The presence of Altamaha Red Filmed sherds and a single olive jar sherd suggest that this portion of the midden was deposited during the Mission period. The lower levels contained a mixture of Irene and Altamaha ceramics, as well as some possible transitional sherds. The presence of Irene, Altamaha, and transitional sherds suggests that the midden was in use beginning sometime during a transitional period and continued to be used during the Mission period. 46 Below the shell midden was a dark non-shell midden, which contained mostly Altamaha and a few Irene ceramics. It also contained the largest numbers of possible transitional sherds. Irene and Altamaha ceramics were found in the same contexts, though the frequencies of Irene decreased in the upper levels. This suggests that there was some overlap in the production of these two ceramic types, with the production of Irene ceramics decreasing over time, rather than a sudden replacement. Below the midden we identified several features, including two pit features and a series of three post features (Features 64, 67, and 68). The postmolds originated just below the non-shell mid- den and are arranged in a semi-circular shape. One of these post features, Feature 64, seems to be three separate, but related, posts. Two of these are set a small distance apart, suggesting that they may represent paired posts. The centermost post appears to have been dug at a diagonal angle, suggesting that it may have been placed as a repair or replacement. In the northwest corner of the unit, we identified a shallow pit feature, Feature 63, which contained oyster, clam, charcoal, and a few Altamaha ceramics. Excavators were unable to determine whether the pit originated below or within the shell midden. There were dark linear stains originating at the pit feature, seemingly extending away from it at a right angle. All of the post features were positioned within this area of staining, suggesting that it might represent a trench related to their construction. The posts contained only Altamaha ceramics, so they probably date to the Mission period. We identified a second pit feature, Feature 66, along the southern wall of our original 2 x 2 meter unit. The outline of the pit was clear in plan view at lower levels, but it was difficult to distinguish from the dark shell midden above. However, it was obvious that the pit intruded into the shell midden or originated just below it. Much to our surprise, excavations revealed the front foot and head of an articulated pig skeleton. In order to excavate the rest of the feature and remove the skeleton, we expanded our excavation with a second 1 x 2 meter unit. Once complete, our excavation revealed an entire juvenile pig skeleton in excellent condition. The bone shows no sign of butchering of any kind and all elements were present. The burial pit is clearly intentional and, with the exception of a few sherds, the pit contained no refuse. The few ceramics that were included in the feature fill date to the Mission period. A small sample of the pig 47 remains was used for radiocarbon dating, the results of which indicate that the burial occurred between 220 +/- 25 B.P, with a Bayesian modeled date of A.D. 1645 and 1680 (Table 11). The pig is currently undergoing ancient DNA analysis, which we hope will give us an idea of where it originated (Richard Jefferies, personal communication 2016). These excavations suggest that there was a transitional period in which Irene, Altamaha, and mixed-trait ceramics were produced simultaneously. They also provided us with a number of possible transitional sherds for analysis and comparison with Irene and Altamaha ceramics. Table 11. AMS dates from Features 50 and 66. -8.8 14C Age, Years BP 160 25 -14.6 220 25 Context UGAMS# Material 13C, Feature 50 25052 Maize Feature 66 25048 Pig Bone Figure 38. Unit 48 in Plan View. Photo taken by Dr. Richard Jefferies. Note that North is Down 48 Modeled Age 1660-1685 A.D. 1645-1680 A.D. Figure 39. Unit 48 Plan View Map. 49 Figure 40. West Profile of Units 48 and 51. Figure 41. East Profile of Units 48 and 51. 50 Figure 42. Feature 63 West Profile. Figure 43. Feature 64 West Profile Photo. 51 Figure 44. Feature 64 West Profile Drawing. Figure 45. Feature 67 at D = 61 cm BD. 52 Figur46. Feature 68 North Profile Photo. Figure 47. Feature 68 North Profile Drawing. 53 Figure 48. Feature 68 at D = 50 cm BD. Figure 49. Feature 66 Photo. D = 50 cm BD. 54 55 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 63 Feature 64 Feature 66 Feature 68 Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 63 Feature 64 Feature 66 Feature 68 Total 35 53 49 37 1 4 1 44 0 224 Bone 90 87 59 6 0 7 0 1 0 250 20 34 36 2 1 4 0 2 1 100 0 1 2 0 0 0 0 1 0 4 0 0 0 0 0 0 0 1 0 1 0 0 1 0 0 0 0 0 0 1 Lead Shot 1 0 1 0 0 0 0 0 0 2 .22 Caliber Cartridge 12 14 10 4 0 2 1 0 0 43 Gastropod Whelk Shell 12 14 10 4 0 0 0 0 0 40 Fired Clay/ Daub Tooth 0 17 55 36 25 2 5 0 0 140 Charcoal Prehistoric Prehistoric Ceramics Ceramics (1 < x < 4 cm2) (> 4 cm2) 211 343 196 29 6 17 28 1 0 831 Bivalve Table 12. All Artifacts from Unit 48. 2 1 0 0 0 0 0 0 0 3 Olive Jar 1 2 0 0 0 0 0 0 0 3 Otolith 54 86 42 6 0 9 0 0 0 197 Prehistoric Ceramics (< 1 cm2) 56 Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-61 cm BD) Feature 63 Feature 64 Feature 66 Feature 68 Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-61 cm BD) Feature 63 Feature 64 Feature 66 Feature 68 Total Savannah Check Stamped 0 0 2 0 0 0 0 0 0 2 Irene Complicated UID Altamaha Altamaha Complicated UID Irene Stamped Altamaha Plain Stamped Altamaha Overstamped 0 0 6 0 6 7 1 4 2 0 10 9 0 2 4 1 11 9 0 1 0 0 0 1 0 0 0 0 0 1 0 0 1 0 3 0 0 0 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 1 1 7 13 1 32 28 Altamaha Red Altamaha Check Altamaha Incised Filmed Stamped UID 1 0 0 0 1 2 0 3 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 2 2 1 6 Table 13. All Ceramics from Unit 48. 57 Level 1 (0-27 cm BD) Level 2 (24-35 cm BD) Level 3 (33-39 cm BD) Level 4 (38-49 cm BD) Feature 66 Total Savannah 0 0 0 1 0 1 Savannah Check Stamped 0 0 1 0 0 1 1 8 4 3 0 1 17 UID Altamaha 3 4 1 0 1 9 10 1 1 0 0 0 12 Gastropod Irene Complicated Stamped 1 2 1 0 0 4 0 0 0 0 1 0 1 UID Whelk Shell Tooth 1 1 0 1 0 0 3 0 0 22 44 59 12 137 Charcoal 14 9 16 3 1 38 81 Bone 136 119 52 9 2 0 318 Table 15. All Ceramics from Unit 51. Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 66 Total Level 1 (0-20 cm BD) Level 2 (20-30 cm BD) Level 3 (30-40 cm BD) Level 4 (40-50 cm BD) Level 5 (50-60 cm BD) Feature 66 Total Bivalve Table 14. All Artifacts from Unit 51. Altamaha Plain 1 1 0 0 0 2 0 1 0 0 0 0 1 Olive Jar Altamaha Burnished Plain 0 1 0 0 0 1 23 26 16 2 1 1 69 Prehistoric Ceramics (< 1 cm2) Altamaha Complicated Stamped 3 2 3 0 1 9 49 47 17 6 2 3 124 0 1 0 0 0 0 1 Shell Bead Altamaha Possibly Overstamped Transitional 1 0 1 0 2 1 0 0 0 0 4 1 9 11 9 1 0 2 32 Prehistoric Prehistoric Ceramics Ceramics (1 < x < 4 cm2) (> 4 cm2) Ceramics Attributes Analysis The focus of this analysis is on traits belonging to Irene, Altamaha, and possible transitional sherds. For this reason, only sherds larger than 3 cm2, which can be assigned to a series, are included. Of the 3414 sherds recovered in these excavations, 584 of them were large enough to assign to a ceramic series. This assemblage contains ceramics from all time periods represented at Site 9Mc23, but was primarily composed of Irene and Ceramics Series Altamaha, which make up 10.7% and 67.4% of the typ- 9% able sherds, respectively. St. Simons 11% sherds 72% (Figure 52 a-c) made up 4% of the typable assemblage and were mostly recovered St. Simons Refuge Deptford Wilmington Savannah Irene Altamaha St. Johns Lamar Possibly Transitional Figure 51. Pie Chart of Ceramics by Series. from the lower levels of Unit 46 (Table 6). St. Simons sherds were mostly plain. Three were incised and one had semi-circular punctations (Figure 52 b-c). Five Refuge sherds made up 0.9% of the typeable assemblage (Figure 52 d-e). Surface treatments for all of these sherds were plain or unidentified. Deptford ceramics were also relatively uncommon, at 0.4% (Figure 52 f-h, j-k). One of these was a rather large cordmarked sherd Figure 46f), and the other was check stamped (Figure 52k). A single Wilmington sherd with an overstamped surface treatment was recovered. Savannah sherds were common by comparison, making up 8.2% of the typable assemblage (Figure 53-54). Most of these were check stamped, but plain or burnished plain surface treatments were also common. There was a single example of a cob-impressed sherd (Figure 53c). Most of the Savannah sherds were grit/sand tempered, but there were a few examples of grit/clay and sherd/grit temper, as well as one each of sand/grit, sand/fiber, grit/sherd, and grit/shell. While most of the 58 C B A D E I G F H J K L Figure 52. A Sample of Uncommon Sherd Types. A: St. Simons Plain (UK7031), B: St. Simons Punctate (UK7036), C: St. Simons Incised (UK7055), D-E: Refuge Plain (UK6969, UK6957), F: Deptford Cordmarked (UK7074), G-H: UID Refuge/Deptford (UK6243.1, UK6233), I: Lamar Incised (UK6231), J: UID Refuge/Deptford (UK6243.2), K: Refuge/Deptford Check Stamped (UK6229), L: St. Johns Punctate (UK7088). B C A D F E G Figure 53. Savannah Sherds. A: Savannah Burnished plain bowl (UK6832), B: unidentified Savannah rim (UK6935), C: Savannah Cob-impressed rim (UK6824), D: Savannah Check Stamped rim (UK7025.1), E-G: Savannah Check Stamped (UK6946, UK7025.3, UK6217). 59 sherds were undecorated, there was a single burnished sherd with a pellet node. Twenty-two of the forty-five sherds had burnished or heavily burnished interiors. Vessel form was identified for two Savannah sherds, one jar (Figure 54) and one bowl (Figure 53a) . Irene ceramics account for 10.7% of the typeable assemblage (Figure 55). The majority of these (88.1%) were complicated stamped, mostly with the Filfot Cross motif (62.7%). Just over half of the complicated stamped sherds were overstamped. Of the complicated stamped sherds, 61.5% had the narrow lands and grooves that should be expected among Irene ceramics. Wide and mixed lands and grooves also were present in the Irene assemblage, at 19.2% each. There were two instances each of plain (Figure 55f) and incised (Figure 55d-e) Irene sherds. There were five additional decorated Irene sherds. Three of these had cane punctations along a rim fillet, (Figure 55 a-c) and there was one example each of trailing, partial cane punctations, and a stamped lip. Twenty-nine (49.1%) Irene sherds had burnished or heavily burnished interiors. Of the five Irene rims that were recovered, two had unmodified rims (Figure 55d) and three had rim fillets (Figure 55a-c). Four of the lips were classified as flat-flat/rounded. A single lip was classified as exterior beveled. There was on ceramic disk (Figure 55g). The majority of Irene sherds (72.9%) were grit/sand tempered. Grit/clay was also a common temper combination at 18.6%. Two Irene sherds had grit/shell temper, and there was one example each of sand, sherd/grit, and clay/sand temper. Irene sherds with at least one inclusion, meaning that there was only a single fragment of that material in the sherd, make up 18.6% of the Irene assemblage. Nearly half of these inclusions were Table 16. Table of Irene Temper. Temper Count Sand 1 Grit/Sand 43 Grit/Clay 11 Grit/Shell 2 Sherd/Grit 1 Clay/Sand 1 Total 59 Percentage 1.7% 72.9% 18.6% 3.4% 1.7% 1.7% 100% 60 Table 17. Table of Irene Surface Treatments. Surface Treatment Count Percentage UID Rectilinear 3 5.1% Plain 3 5.1% UID Complicated Stamped 4 6.8% Filfot Cross Complicated Stamped 17 28.8% Line Block Complicated Stamped 1 1.7% Other Curvilinear Complicated Stamped 4 6.8% Overstamped Filfot Cross 20 33.9% Other Curvilinear Overstamped 2 3.4% Other Rectilinear Overstamped 1 1.7% Unidentified 4 6.8% Total 59 100% Figure 54. Savannah Check Stamped Jar (UK6862). A D C B G F E J I K H Figure 55. Irene Sherds. A-C: UID Irene cane punctated rim fillets (UK7017.1, UK6984.2, UK6920.1), D: Irene Incised rim (UK7026), E: Irene Incised body sherd (UK6826), F: Irene Plain (UK7009), G: Irene Complicated Stamped ceramic disc (UK7061), H-K: Irene Complicated Stamped sherds (UK6947.1, UK6882.1, UK6882.2, UK6865.2). 61 fiber and the other half were clay. Only one Table 18. Table of Altamaha Surface Treatment. sherd contained a second inclusion, which was Surface Treatment Count UID Rectilinear 29 Plain 20 Burnished Plain 1 Heavily Burnished Plain 2 Burnished Plain Interior/ Exterior 2 Heavily Burnished Plain Interior/Exterior 1 UID Complicated Stamped 2 Filfot Cross Complicated Stamped 1 Line Block Complicated Stamped 26 Chevron Complicated Stamped 1 Other Curvilinear Complicated Stamped 2 Other Rectilinear Complicated Stamped 6 Overstamped 2 "Cross Simple" 140 Overstamped Filfot Cross 4 Overstamped Line Block 87 Other Curvilinear Overstamped 1 Other Rectilinear Overstamped 6 Checkstamped 4 Eroded/Spalled Exterior 2 Unidentified 31 Total 370 shell. The Altamaha portion of the assemblage was the most diverse, but this may be the result of its larger size. Altamaha ceramics made up 67.4% of the assemblage, at 370 sherds. The majority of Altamaha sherds were classified as complicated stamped (36.2%) or overstamped/cross simple (37%). As expected, the majority (88.4%) of stamped sherds had wide lands and grooves. Narrow and mixed lands and grooves were present in smaller percentages, at 5.3% and 6.3% respectively. After stamping, red filmed (4.6%) and plain (3%) were the most common surface treatments. All other surface treatments were identified in frequencies lower than 1%, including burnished plain, incised, and check Percentage 7.8% 5.3% 0.3% 0.5% 0.5% 0.3% 0.5% 0.3% 1.1% 0.3% 0.5% 1.6% 0.5% 37.8% 1.1% 23.5% 0.3% 1.6% 1.1% 0.5% 8.4% 100% stamped (Table 18). Over half (58.1%) of the Altamaha sherds had burnished or heavily burnished interiors. Despite the larger sample size, Altamaha sherds were slightly less likely to be decorated than Irene sherds, with 8.6% of sherds being decorated, compared to 11.8% for Irene. Decoration on Altamaha sherds consisted mostly of various kinds of punctations applied below the lip of the vessel or along the bottom of the rim fold/strip, including partial cane (n = 6), fingernail (n = 5), paired hemiconical (n = 4), whole cane (n = 3), and one each of triangular, curved shell, vertical rectangle, and unidentified (Figure 58). All paired hemiconical punctates were found with trailing. Other deco- 62 rations include incising (n = 6), trailing (n = 4), and painting (n = 2). Incising and trailing usually occur on sherds that otherwise appear to be plain. However, one sherd in this assemblage shows signs of incising over overstamping. Much of the design is missing, but it may be the letter C and a portion of another letter (Figure 56). One of the painted sherds is an overstamped rim with dots and a wide line on the interior. The other is a base sherd with two dots on the interior (Figure 57). As expected, the Altamaha series shows the most diversity in temper recipes (Table 9). Grit/sand was by Table 19. Counts and Percentages of Altamaha Tempers. Temper Count Sand 6 Sand/Grit 3 Grit/Sand 287 Grit/Sherd 1 Grit/Clay 39 Grit/Shell 23 Grit/Fiber 1 Clay/Sand 7 Clay/Grit 1 Fiber/Clay 1 Composite 1 Total 370 Percentage 1.6% 0.8% 77.6% 0.3% 10.5% 6.2% 0.3% 1.9% 0.3% 0.3% 0.3% 100% far the most common temper, at 77.6%, followed by grit/ clay (10.5%) and grit/shell (6.2%). Clay/sand tempered sherds made up 1.9% of the assemblage. Sand tempered sherds made up 1.6%, all of which were red filmed. Other temper types made up very small percentages of the assemblage, including sand/grit at 0.8% and one example each of grit/ sherd, grit/fiber, clay/grit, fiber/clay, and composite temper (equal parts grit, shell, and clay). Altamaha sherds were also the most likely to contain inclusions. The most common of these were clay at 44.2% and shell at 37.7%. Fiber also was common, making up 22.1% of inclusions. One example each of hematite, grit, and sherd inclusions, as well as two unidentified inclusions also are present. Of the 145 Altamaha sherds that were analyzed using our newly developed system of varieties, the overwhelming majority of sherds were classified as Altamaha (97.2%). Two sherds were classified as Darien, meaning they had fine sandy pastes. The clay tempered Sapelo variety and the heavily shell tempered Belle Marsh variety were represented by one sherd each. These varieties may be useful in further refining the Altamaha chronology and understanding vessel function, but more work with larger sample sizes is needed. The sample collected for this study is too small to tackle these questions, but it was useful for thin section analysis, the results of which are reported below. 63 Twenty of the thirty-four sherds with an identifiable rim form were identified as rim fold/ strips. Six of these were flared, while the others were unidentified. Two rims were unmodified, one flared and one restricted. Two rims had narrow folds, and a single rim was rolled. Vessel form was identifiable for eleven sherds. Three of these were bowls and five were jars (Figures 61 and 62), while the others were identified as colonowares (Figure 63). While all of the red filmed sherds are likely colonoware, as SIMPAP has yet to identify a single red filmed vessel that was not colonoware, only five colonoware sherds were large enough to determine vessel form. Two of these were plates with red filmed interiors and complicated stamped exteriors. The others have unidentified vessel forms. Two of these have red filmed exteriors with plain interiors, one with an interesting vessel form that may have been a cup. The final piece has a plain interior and a complicated stamped exterior. All of these sherds came from Unit 42. Figure 57. Base Sherd with Two Painted Dots on the interior (UK6870.04). Figure 56. Altamaha Sherd with an Unidentified Incised Design Over Overstamping (UK6237.11). 64 B A F K D C L M J I H G E N O Figure 58. Altamaha Rims and Decorated Sherds. A-E: UID Altamaha rim fold/strips with cane punctations (UK6851.2, UK6921.1, UK6232.2, UK6210.1, UK6948.1), F-H: UID Altamaha rim fold/strips with fingernail punctations (UK6263.8, UK6851.1, UK7078), I: UID Altamaha triangular stylus punctations (UK6232.1), J: UID Altamaha rolled rim (UK7090.1), K-L: Altamaha Complicated Stamped rims (UK6948.2, UK6263.1), M-N: UID Altamaha rim and shoulder with paired hemiconical punctates and trailing (UK6263.7, UK6232.3), O: Altamaha Incised neck sherd with an unidentifiable incised design (UK6965). Figure 59. A Sample of Altamaha Complicated Stamped Sherds (UK6263.1, UK6950, UK7094, UK6976, UK7094, UK7094, UK6842.2, UK6265) 65 Figure 60. Sample of Altamaha Overstamped Sherds. UK6266, UK6893, UK6893, UK6266, UK6893, UK6893, UK6893, UK6893, UK6266) Figure 61. Complicated Stamped Altamaha Rim (UK6313). 66 A B D C E Figure 62. Sherds with Identifiable Vessel Forms. A-B: Altamaha Complicated Stamped jar rims with cane punctations (UK6975, UK6924), C: Altamaha Complicated Stamped bowl (UK6307), D: Savannah Burnished Plain bowl (UK6832), E: Altamaha Incised bowl (UK6844). B A C D E Figure 63. Colonoware with Identifiable Vessel Forms. A: Altamaha Red Filmed plate (UK6308), B: Altamaha Complicated Stamped rim (UK6211), C: Altamaha Red Filmed plate rim (UK6267), D: Altamaha Red Filmed unidentified vessel (UK6238), E: Altamaha Red Filmed possible mug (UK6312). 67 Other Mission period ceramics include one Lamar (Figure 52i) and one St. Johns sherd (Figure 52l). The Lamar sherd came from Unit 42. It has a series of triangular punctates below the lip, with two trailed lines running parallel to the rim and was likely part of a bowl. While the term Lamar is often used to describe ceramics from the mainland and interior, SIMPAP uses the term to refer to incised bowls that may have been manufactured during the Irene or Altamaha periods. We do not know whether these sherds were manufactured on the mainland or on Sapelo (Jefferies and Moore 2009). The St. Johns sherd came from Unit 43 Feature 50, one of the shell-filled pits. It has whole cane punctates extending below the rim and a notched lip. The rim is flared, and it is sherd tempered. We also recovered a single sherd that may be Creek from the first level of Unit 47. It is a sizable rim sherd with a sand/sherd temper and scrape marks on the interior. The surface treatment was classified as burnished plain, but it may be brushed. Four of the twelve possible transitional sherds had unidentified surface treatments. Four of these were plain and two were burnished. Four sherds were stamped, two with unidentified rectilinear patterns and two overstamped. Of these four, two had wide lands and grooves, and there was one example each of narrow and mixed lands and grooves. Two of the possible transitional sherds were decorated with incised designs on the body. Ten had grit/sand B temper, and two had grit/clay temper. A C Figure 64. A Sample of Irene/Altamaha Sherds. A-B: Irene/Altamaha Complicated Stamped (UK7012, UK7039), C: Irene/ Altamaha Incised (UK6966), D: Irene/ Altamaha Unidentified Stamped (UK6966). 68 D Unfortunately, possible transitional ceramics made up only 2.2% of the assemblage. This is likely a reflection of the short duration of the purported transitional time period, as well as the fact that we did not identify possibly transitional sherds until later in the project. While some sherds have been noted as Irene/Altamaha throughout the process, the possibly transitional category was not a focus of the analysis throughout the project. It is also likely that any period of transition would have been shorter than the other time periods represented by the ceramic chronology. While these sherds do have traits of both Irene and Altamaha sherds, a larger, dated, sample would be help- ful in determining their relationship to the rest of the ceramic chronology. Statistical Comparisons Statistical comparisons of Irene, Altamaha, and possible transitional sherds and contexts illustrate some useful differences between the three groups. Small sample sizes, especially for transitional sherds, along with abnormal distributions of quantitative data, made statistical analysis difficult. More work with larger samples may be useful in identifying more diagnostic traits. However, some comparisons were still helpful. The maximum thickness was recorded for every sherd larger than 3 cm 2, excluding base sherds, which tend to be much thicker than body sherds, and rims. Jefferies and Moore (2009) have suggested that there is a trend towards increasing thickness from the Savannah period through the Altamaha period, possibly due to increasing vessel size. However, this expected trend was not observed in this dataset. As expected, Savannah sherds had the thinnest average body thickness of 6.8 mm. Body thickness increased during the Irene period to an average thickness of 8.2 mm. Surprisingly, Altamaha sherds had a slightly lower average thickness at 8.1 mm. The difference is small and suggests that there may be no real difference between the two groups in terms of thickness and that this will not be a useful trait for identifying transitional sherds. It may also indicate that this is a late Irene assemblage. Transitional sherds had the greatest thickness at 8.8, which is surprisingly high. The sample size for this category is rather small, and it seems likely that a couple of unusually thick sherds are skewing the average. These measurements do include measurements from sherds near the base and from a variety of vessel forms, which may account for some of the unexpected 69 Figure 65. Boxplot of Maximum Body Thickness by Series. variability in the measurements. The possibility of an overlap between Irene and Altamaha ceramics was tested further using chi-square tests. Tests comparing specific traits across ceramic types were somewhat limited because the use of diagnostic traits would only reflect the groups that they were placed in based on the typology. For this reason, traits cannot be considered independent from type. In order to test the possibility that there may still be differences at the assemblage level, I used the chi-square statistic to compare assemblages from contexts dating to each time period. This provided larger sample sizes that were more suitable for statistical analysis, and it allows us to focus on assemblages, rather than individual sherds. These contexts were defined based on excavation notes and the inclusion of ceramics of predominately one type. Some mixing of ceramics from mul- 70 tiple time periods should be expected. Since the goal of this analysis was to identify differences at the assemblage level, all ceramics from each context were used in the comparison, regardless of ceramic series. The assemblage used to represent the Irene period came from Levels 2 and 3 of Unit 46, which was determined to be an Irene context, with minimal mixing from the preceding Savannah or subsequent Altamaha periods. It contained sixteen Irene sherds, six Altamaha, five possible transitional sherds, and two unidentified sherds. The Altamaha assemblage was pulled from Feature 50, a shell-filled pit that is associated with the Mission period. It consisted of 113 Altamaha sherds, two unidentified sherds, and a single Irene sherd. The possible transitional assemblage was pulled from the possibly transitional shell-midden and the zone just below it in Units 48 and 51. This assemblage consisted of 72 Altamaha sherds, twelve Irene sherds, five unidentified sherds, and three possible transitional sherds. Chi-square tests were used to compare the transitional context with both the Irene and Altamaha contexts. The largest difficulty in conducting statistical comparisons between these assemblages was the issue of sample size. For some traits, such as decoration, the overall sample sizes were too small to attempt statistical comparison. Sample sizes were sufficient for surface treatment, land and groove width, and temper. Even for the traits that were selected for this comparison, there were issues with the counts for individual chi-square cells. However, these issues were unavoidable and not so severe that chi-squares could not be calculated. They will be addressed further below. The first trait considered in this analysis is surface treatment. In order to reduce the number of cells with counts of fewer than five, some surface treatment categories were combined. All curvilinear complicated stamped designs were grouped together, as were all rectilinear stamped designs. As each of these is associated with a particular time period, Irene and Altamaha respectively, this should not be detrimental to this analysis. Plain and burnished plain surface treatments were also combined. As a result, there were only 7 cells with counts lower than five. Unfortunately, due to the nature of the data, these counts likely will not increase even with larger sample sizes. Traits such as rectilinear stamping, cross simple stamping, and check stamping, for example, will always be rare 71 to non-existent in Irene assemblages. For this reason, I still feel confident moving forward with a chi -square analysis. A chi-square test comparing the surface treatment of Irene, Altamaha, and possible transitional sherds returned a p-value of less than 0.001 (Table 20), suggesting that there is a statistically significant difference between the expected and observed surface treatments for these types. Based on the residuals, we can determine which surface treatments contribute most significantly to this result. Both rectilinear and cross simple stamping are much less common among Irene ceramics than expected, with residual values of -3.0 and -7.9 respectively (Table 21). Curvilinear stamping is much more common than expected, with a residual value of 7.7. These results are not surprising, as they are in keeping with the type descriptions. Among the possible transitional ceramics, the surface treatments that contribute most to the significance of the chi-square are plain and cross simple stamping, with residuals of 2.2 and -2.1 respectively. These factors contribute much less to the overall significance of the test than the notable traits from the Irene and Altamaha assemblages. Residuals indicate that most traits from the Altamaha assemblage contribute to the significance of the chi-square. The most significant of these is cross simple stamping, with a residual value of 10.0. This surface treatment is much more common among Altamaha than expected. Rectilinear stamping is more common than expected, with a residual of 3.0, and curvilinear stamping is less common than expected, with a residual of -9.5. Neither of these results are surprising, as this fits with the type description. Plain surface treatment also contributes to significance, with a residual of -2.5, indicating that there are fewer than the expected number of plain ceramics in the Altamaha assemblage. Based on the residual analysis, it seems that the transitional context has a mix of rectilinear and curvilinear stamping that is not found in the Irene and Altamaha assemblages, as well as a slightly higher than expected count of plain ceramics and a slightly lower than expected count of cross simple stamped sherds. Although the residuals for this trait do not indicate that it contributes greatly to the significance of the chi-square, it is also worth noting that there are no check stamped 72 Table 20. SPSS Output for Surface Treatment Chi-Square Test. Chi-Square Tests Asymptotic SigValue Pearson Chi-Square df 10 .000 45.248 10 .000 44.885 Likelihood Ratio nificance (2- a N of Valid Cases 222 Table 21. SPSS Output for Surface Treatment Crosstab with Residuals. Crosstab SurfaceTreatment Rectilinear Context Irene Count Residual Plain Curvilinear 'Cross Simple' Check Stamped 9 2 11 1 0 -3.0 .3 7.7 -7.9 -.3 Possi- Count 34 7 11 23 0 ble Residual .0 2.2 1.8 -2.1 -.7 Altama Count 49 4 3 44 2 ha Residual 3.0 -2.5 -9.5 10.0 1.0 Count 92 13 25 68 2 Transitonal Total Crosstab SurfaceTreatment Unidentified Context Irene Count 6 Residual Possible Tran- Count sitional Residual Altamaha Count 29 3.1 7 82 -1.1 9 Residual Total Total 111 -2.0 Count 22 73 222 sherds found in either the Irene or possible transitional contexts, though they are found in the Altamaha context. While it is not surprising that check stamped ceramics are absent from the Irene assemblage, as they are quite rare during this period, it is worth noting that they are missing from the possible transitional context. If this context is truly transitional, this may suggest that check stamping is a later Altamaha trait. Another chi-square test related to surface treatment compared land and groove widths across the contexts. For this test, only two cells had counts lower than five, with four each. No categories could be combined to fix this, but the counts are close enough that I feel confident moving forward with a chi-square analysis. This test also suggested a statistically significant difference between contexts, with a p-value of less than 0.001 (Table 22). As expected, the residuals indicate that the Irene assemblage had much lower than the expected number of sherds with wide lands and grooves and much higher than the expected number of sherds with narrow lands and grooves, with residuals of -9.1 and 5.1 respectively (Table 23). The transitional assemblage had fewer than the expected number of sherds with wide lands and grooves, indicated by a residual of -4.9. The Altamaha assemblage appears to have contributed the most to the significance of this test. This assemblage had much higher than the expected number of sherds with wide lands and grooves and a much lower than expected number of sherds with narrow lands and grooves, with residuals of 14.0 and -7.0 respectively. Again, this is not surprising, as it is in keeping with the type description. The Altamaha assemblage is the only one with fewer than the expected number of sherds with mixed narrow lands and grooves, as indicated by a residual of -3.0. The transitional assemblage seems to have a mix of both narrow and wide lands and grooves, as well as a number of sherds with mixed wide and narrow lands and grooves that most resembles the Irene assemblage. The final chi-square test was a comparison of temper across the assemblages. Of the three tests presented here, this one had the greatest number of cells with counts lower than five. Most temper categories cannot be grouped, but some could be combined. Sherds with mostly sand, but some grit temper, were combined with sand-tempered sherds in the sand category. Shell and grit/shell 74 Table 22. SPSS Output for Land and Groove Width Chi-Square Test. Chi-Square Tests Asymptotic SigValue Pearson Chi-Square Likelihood Ratio df nificance (2- a 6 .000 25.435 6 .000 26.504 N of Valid Cases 222 Table 23. SPSS Output for Land and Groove Width Crosstab with Residuals. Crosstab Land and Groove Width Wide Context 1 2 3 Count Both Total 7 10 8 4 Residual 2.0 -9.1 5.1 1.9 Count 16 49 10 7 Residual 2.0 -4.9 1.9 1.1 Count 15 87 4 5 -4.0 14.0 -7.0 -3.0 38 146 22 16 Residual Total Narrow Count 75 29 82 111 222 tempered sherds were combined into the shell category. The other category consists of temper types for which there was only one or two examples. These included, grit/sherd, grit/fiber, and composite tempers. The sherd with composite temper had roughly equal parts grit, clay, and shell. Even after these categories, temper was the least suitable trait for statistical analysis included in this study. It included nine cells with counts fewer than five and twelve cells with an expected count lower than five. As in the previously discussed examples, this is due to the nature of the data. Temper diversity increases with time, meaning that temper recipes that are represented in later samples will be absent from earlier samples. Thought the results of this test are less reliable statistically, they do reveal a culturally meaningful pattern, so it is worth proceeding with a chi-square. The chi-square returns a p-value of 0.024 (Table 24), indicating that there is a statistically significant difference between the expected and observed counts for this trait across the three assemblages. Based on the residuals, the Irene and possible transitional contexts contribute the most to the significance of the chi-square. The Irene assemblage has lower than the expected number of sherds with grit/sand temper and higher than expected numbers of sherds with grit/clay temper, with residuals of -2.2 and 4.7 respectively (Table 25). The possible transitional assemblage has greater than expected numbers of sand and grit/sand tempered sherds, with residuals of 2.8 and 2.2. It had fewer than the expected number of grit/clay tempered sherds, indicated by a residual of -3.2. Though these categories are not flagged as significant based on residuals, the Irene assemblage clearly has a low diversity in temper recipes, with counts of 0 for the sand, shell, and clay/ sand categories. Temper diversity appears to increase in the possible transitional assemblage, and increase further in the Altamaha assemblage. It is also worth noting that the rarer temper types, here combined into an other category, are most abundant in the Altamaha context. From this analysis, it appears that the transitional context has tempers that most resemble that of Altamaha assemblages, though they may be slightly less diverse, based on the lack of other categories in the transitional context. More testing would be needed to confirm this. Based on chi-squares comparing surface treatment, land and groove width, and temper, it 76 appears that the context that we have identified as possibly transitional is indeed unique. This assemblage has high frequencies of both rectilinear and curvilinear stamping, while Irene and Altamaha assemblages have mostly one or the other. It also lacks the check stamped surface treatment that is found in Altamaha assemblages. It has relatively high frequencies of both wide and narrow lands and grooves, while Altamaha and Irene assemblages have primarily one or the other, and it also has the highest frequency of mixed wide and narrow lands and grooves. Like the Altamaha assemblage, it has much greater temper diversity than the Irene assemblage, though possibly not as diverse as the Altamaha assemblage. It appears that these traits may be useful in the identification of transitional assemblages; however, more testing is needed. First, AMS dates are needed to confirm that the context that we have identified as possibly transitional does indeed represent a period of transition. While this analysis suggests that the assemblage is unique, this could be due to social factors such as family or ethnic group differences, especially when we take into consideration the diverse groups that would have occupied Sapelo at this time (eg, Worth 2007). Secondly, statistical analysis of additional traits would lend further support to this hypothesis. Decoration and vessel form could be incredibly useful in the identification of transitional contexts, but sample sizes for sherds with these traits were too small in this dataset to even consider for statistical analysis. Repeating some of these tests with larger sample sizes may also lend additional support to this argument. 77 Table 24. SPSS Output for Temper Chi-Square Test. Chi-Square Tests Asymptotic SigValue Pearson Chi-Square df 10 .024 22.856 10 .011 20.658 Likelihood Ratio nificance (2- a N of Valid Cases 222 Table 25. SPSS Output for Temper Crosstab with Residuals. Crosstab Temper Sand Context Irene Count Residual Possibly Count Transi- Residual Grit/ Grit/Clay Shell Clay/ Other Total 0 20 8 0 0 1 -.8 -2.2 4.7 -1.4 -.7 .3 5 65 6 5 1 0 2.8 2.2 -3.2 .9 -.8 -1.8 1 85 11 6 4 4 -2.0 .0 -1.5 .5 1.5 1.5 6 170 25 11 5 5 82 111 tional Altamaha Count Residual Total Count 78 222 Thin Section Analysis The thin section analysis was conducted using both sherds described in the above section on the Altamaha portion of this assemblage as well as sherds that were pulled from other previously excavated contexts, to allow us to test more varieties than were identified in our initial analysis. Overall, thin section analysis showed very little difference between the ceramic types in terms of temper. The two Savannah sherds (Figure 60) had the lowest average percentage of grit, at 11.2%, as compared to Irene at 15.8% (Figure 61), Altamaha at 18.3% (Figure 62), and possible transitional sherds at 19.4% (Figure 63). However, this may be influenced by the particularly small sample of Savannah sherds, with the entire series represented by only two thin sections. Clay was most common among the possible transitional sample at 4.5% and was represented by less than 1% in all other samples (actual percentages are displayed in the table below). It was commonly identified in thin section, even when it was not noted in macroscopic examination. Percentages for sand, paste, and voids were similar across all types, and all other temper types and inclusions (sherd, shell, fiber, mica, unidentified, and possible charcoal) were represented by an average of less than 0.1% to 2%. There was variation from sherd to sherd, largely in agreement with macroscopic temper identifications performed before thin section preparation, meaning that our macroscopic analysis is fairly accurate. However, there appears to be little variation across series. The thin section analysis of the Altamaha varieties that were selected (Altamaha, Belle Marsh, Darien, Mud River, and Sapelo) were more or less consistent with what was expected. The Altamaha variety showed the most diversity in unexpected inclusions, usually amounting to one or two per sherd. As expected, the temper was composed primarily of grit and sand. The Darien (Figure 62e) variety was also mostly grit and sand, but with a higher percentage of sand than the Altamaha variety (Figure 62 a-c), in keeping with our definition of this variety as having a finer paste. The Mud River sherd (Figure 62f) was also composed primarily of grit and sand, as expected. This variety is based on the presence of an iron oxide film around the quartz crystals. The presence of this film was noted on both grit and large sand particles. The Belle Marsh variety (Figure 62d) is defined by shell temper, but shell only made up 79 0.5% of the Belle Marsh thin section. Only one large piece of shell was observed in the thin section, even though several are visible along the sherds broken edges. Upon reexamination and comparison with the portion of the sherd from which the thin section was taken, it appears that one or two sizable pieces of shell may have fallen out of the sherd during the thin section preparation process. Similarly, the Sapelo variety (Figure 62g), which is defined by the presence of clay temper, was made of only 0.63% clay. However, it was noted during thin section analysis that there were fine pieces of clay that were missed by the point counting grid. It should be noted that this sherd had a lower percentage of clay than both the Altamaha and Mud River varieties. Of all of the varieties that were developed for this project, this one seems the most questionable, though this analysis is based on a single thin section, which may not be representative of the whole sherd. Overall, the varieties that were developed for this analysis seem to be supported by thin section analysis. Both the Belle Marsh and Sapelo varieties may require further investigation, but they were composed of the materials predicted by the newly created varieties, though at times in smaller quantities than expected. Surprisingly, there was no significant difference in temper frequencies across types, suggesting that tempering techniques did not change drastically over time. Based on this admittedly small analysis, it would seem that clay preparation techniques were relatively stable throughout the Late Prehistoric and Mission periods. A B Figure 66. Savannah Sherds in Thin Section. A: Savannah grit/sand tempered sherd (UK6829.1), B: Savannah sherd/grit tempered sherd (UKUK630.3). 80 A B C D E F Figure 67. Irene Sherds in Thin Section. A-C: Irene grit/sand (UK6863, UK6817.2, UK6866.1), D-F: Irene grit/clay (UK6825.2, UK6866.2, UK6867). 81 A B C D E F Figure 68. Altamaha Sherds in Thin Section. A-C: Variety Altamaha (UK6267.1, UK6309, UK6266.1), D: Variety Belle Marsh (UK6265.1), E: Variety Darien (UK6850.1), F: Variety Mud River (UK6310), and G: (next page) Variety Sapelo (UK6263.1). 82 Figure 68 Continued. Variety Sapelo in Thin Section (UK6263.3). G 83 A B C D E F Figure 69. Possible Transitional Irene/Altamaha Sherds in Thin Section. A-E: Irene/Altamaha grit/sand sherds (UK6100, UK6842.1, UK6842.6, UK6864, UK6861.2) and F: Irene/Altamaha sand tempered sherd (UK6857.9). 84 85 UK6857.9 6 UK6867 2015-15-1 7 48 46 - W 48 NA NA NA NA UK6861.2 48 50 50 NA 50 64 NA NA NA 3 UK6842.6 42 42 48 42 UK6864 24 48 UK6825.2 NA 47 UK6866.2 2015-18-1 46 - East UK6866.1 2015-18-1 46 - East 126 UK6310 126 6 126 UK6265.13 UK6850.1 UK6266.19 50 NA 48 45 42 3 114 126 Feature UK6842.1 UK6340.33 UK6263.3 Unit NA NA NA 50 NA 50 FS# UK6100 32 STP 346 UK6863 2015-26-1 46 UK6829.1 NA 47 UK6267.1 126 42 UK6817.2 NA 47 UK6309 126 42 Cat# 2 2 3 4 2 2 2 1 NA NA 2 NA 1 2 NA NA 3 3 NA 1 NA Level Grit/Sand Sherd/Grit Clay/Sand Altamaha Comp Stamped Savannah Check UID Altamaha Grit/Clay NA NA Altamaha Overstamped Irene Comp Stamped Grit/Clay Grit/Sand Grit/Sand NA NA NA NA Altamaha Overstamped Altamaha NA Irene Comp Stamped Irene Comp Stamped Irene Comp Stamped Body Body Body Body Body Body Body Body Altamaha Comp Stamped NA Rim Body Body Body Rim Body Body Partial Vessel Body Body Rim Rim Rim Fragment Type Altamaha Comp Stamped Mud River Quartz w/ Iron Oxide Altamaha Comp Stamped Belle Altamaha Red Film Darien Altamaha Comp Stamped Altamaha Grit/Shell Sand Grit/Clay Grit/Sand Grit/Sand Irene Comp Stamped NA Irene Comp Stamped NA Savannah Check NA Altamaha Red Film Altamaha UID Irene NA Altamaha Overstamped Altamaha NA Sapelo Temper Variety Series Table 26. Sherds Used for Thin Section Analysis. The table lists both the sherds context and the percentage of each temper component. Possibly Transitional Possibly Transitional Possibly Transitional Possibly Transitional Possibly Transitional Notes Possibly Transitional Piece Plot Scanning Electron Microscope Analysis The SEM analysis of thick sections was successful, though the slips were more difficult to identify using the epoxy embedding method than in preliminary studies using the edges of freshly broken sherds. However, this method stabilizes the sherd surface before cutting, which ensures that the slip (if one is present) will be present at the cut edge of the sherd. When breaking the sherds, the slips often flaked or chipped off, so that they could not be seen along the broken edge. The epoxy embedding method allows for a more standardized cross section of the slip and the slip is still identi- fiable. Of the six sherds examined with the SEM, only the Altamaha Red Filmed sherd had a readily identifiable slip. On this sherd, there was a visible change in texture between the body of the sherd and the slip at its surface. This texture change was not present on the sherds that were recorded as having blackened surfaces or possible orange or white slips, even after examination under higher magnifications than those needed to identify the red film. This indicates that most of the colorations that were previously recorded as possible slips are the result of oxidation during the firing process, rather than the application of a second, finer material to the vessels surface before firing. Based on this study, red filming represents a new technology, rather than a simple change in color preference. 86 Epoxy Slip Body Figure 70. SEM Photo of an Altamaha Red Filmed Sherd Epoxy Body Figure 71. SEM Photo of an Altamaha Sherd Originally Coded as Having an Orange Slip. 87 Epoxy Body Figure 72. SEM Photo of an Altamaha Sherd Originally Coded as Having a White/Tan Slip. Epoxy Body Figure 73. SEM Photo of an Altamaha Sherd with a Blackened Interior. 88 Epoxy Body Figure 74. SEM Photo of an Irene Sherd Originally Coded as Having a White/Tan Slip. Epoxy Body Figure 75. SEM Photo of an Irene Sherd with a Blackened Interior. 89 Conclusions The goals of this study were to determine whether Late Prehistoric mixed trait sherds on Sapelo Island represent a period of transition, as well as to determine when this transition occurred at Site 9Mc23. Unfortunately, we did not recover enough datable material to give a more specific date for this transition. However, it appears that the context that we identified as possibly transitional is different from Irene and Altamaha contexts at the site in terms of surface treatment, land and groove width, and temper. This might indicate that we have identified a period of transition. Even if this transition occurred over the course of a single generation, as has been suggested (Saunders 2000), it was not immediate. However, the possible transitional context may be the result of social, rather than temporal differences. The identification of a possible transitional period is based largely on the results of the excavation of the shell midden in Units 48 and 51, which contained both Irene and Altamaha ceramics, as well as some sherds with traits belonging to both series. The inclusion of all of these ceramic types in a single midden suggests that they were produced simultaneously and by the same people. The excavation of more contexts like this one and radiocarbon dates from those contexts would be helpful in confirming the transitional nature of this context. Testing for differences in non-diagnostic traits among Irene, Altamaha, and possible transitional sherds showed no statistically significant differences. However, statistical tests for differences between Irene, Altamaha, and possible transitional contexts suggested that there may be statistically significant differences at the assemblage level, which may be useful in the identification of transitional contexts. It seems that transitional contexts may be identified by the presence of ceramics with greater temper diversity than Irene assemblages, as well as mixed wide and narrow land and groove widths. Surface treatment may be the most useful trait. Transitional contexts have mixed curvilinear and rectilinear surface treatments and contain fewer check stamped and cross simple stamped sherds than Altamaha contexts. They may also contain more plain ceramics. The context identified as possibly transitional differed from the Irene and Altamaha contexts in temper, surface treatment, and land and groove width. These differences are evidence that the midden is different in some way from the Irene and Altamaha contexts. However, this does not confirm that the midden is transitional. It may be the case that the midden was deposited by a different ethnic or family group from the 90 other contexts that were considered in this study, especially when the great diversity of groups that congregated on Sapelo Island to avoid raids by the English-backed Westo and pirates is taken into consideration (Worth 2007). Thin section analysis did not identify significant changes in temper technologies through time, with the exception of an increase in the diversity of temper during the Altamaha period. The small sample size may have influenced these results, making it more difficult to accurately identify these differences. However, this analysis did confirm that there are differences in the tempers of the Altamaha varieties that were created for this study. While the sample sizes used were not large enough to make any definite statements about the significance of these differences, the creation of these categories opens up a new avenue for future study. Red filming was the only slip that could be identified in the SEM study, suggesting that this trait represents a change in technology, not the continued use of the same techniques with a new color. This is the best example of highly visible change identified in this study. The fact that this technique is primarily associated with colonowares, which are clearly associated with the Spanish, suggests that it might be the result of Spanish influence. It may not be the case that the Spanish were directly involved in this process, by directing Guale pottery production. It seems more likely that this is a revival of an old technique, encouraged by market forces. A comparison of Altamaha Red Filmed ceramics with other, earlier, red filmed ceramics is necessary to fully understand this development. While it is clear that the context that we initially identified as possibly transitional is different from the other Irene and Altamaha contexts used in this study, more investigation is needed to understand the significance of that difference. The excavation of more possible transitional contexts, and AMS dating of those contexts, is needed to confirm that these ceramics were actually produced at the time of the transition from Irene to Altamaha. Statistical comparisons of ceramics from different areas of Site 9Mc23 to determine how much variation can be expected across the site, may also be helpful in determining whether this is a factor in this analysis. 91 References Cited Caldwell, Joseph R. and Catherine McCann 1941 Irene Mound Site, Chatham County, Georgia. University of Georgia Press, Athens. Crook, Morgan R. 1980 Archaeological Indications of Community Structures at the Kenan Field Site. In Sa pelo Papers: Researches in the History and Prehistory of Sapelo Island, Georgia, edited by D. P. Juengst, pp. 89-100. Studies in the Social Sciences. West Georgia College, Carrollton, Georgia. Crook, Morgan R., Jr. 1984 Evolving Community Organization on the Georgia Coast. Journal of Field Archaeology 11(3):247-263. Deagan, Kathleen 1983 Spanish St. Augustine: The Archaeology of a Colonial Creole Community. Academic Press, New York. Deagan, Kathleen and David Hurst Thomas 2009 Epilogue. In From Santa Elena to St. Augustine: Indigenous Ceramic Variability (A.D. 1400-1700), edited by K. Deagan and D. H. Thomas, pp. 209-212. Anthropological Papers of the American Museum of Natural History. American Museum of Natural History, New York. DePratter, Chester B. 1991 W.P.A. Archaeological Excavations in Chatham County, Georgia: 1937-1942. Laboratory of Archaeology Series Report. University of Georgia, Athens. 2009 Irene and Altamaha Pottery from the Charlesfort/Santa Elena Site, Parris Island, South Carolina. In From Santa Elena to St. Augustine: Indigenous Ceramic Variability (A.D. 1400 -1700), edited by K. Deagan and D. H. Thomas, pp. 19-47. Anthropological Papers of the American Museum of Natural History. American Museum of Natural History, New York. Hughes, Daniel 2012 A Case of Multiple Identities in La Florida: A Statistical Approach to Nascent Cosmopolitanism. Historical Archaeology 46(1):8-27. Jefferies, Richard W. and Christopher R. Moore 2009 In Search of Mission San Joseph de Sapala: Mission Period Archaeological Research on Sapelo Island, Georgia. University of Kentucky. Submitted to Georgia Department of Natural Resources. Lechtman, Heather 1977 Style in Technology Some Early Thoughts. In Material Culture: Styles, Organization, and Dynamics of Technology, edited by Heather Lechtman and Robert Merrill, pp. 3 -20. West Publishing Co., New York. McMichael, Alan E. 1980 A Model for Barrier Island Settlement Pattern. In Sapelo Papers: Researches in the History and Prehistory of Sapelo Island, Georgia, edited by D. P. Juengst, pp. 47-60. Studies in the Social Sciences. West Georgia College, Carrollton, Georgia. 92 Moore, Christopher R. 2009 Analysis of Native American Ceramics. In In Search of Mission San Joseph de Sapala: Mission Period Archaeological Research on Sapelo Island, Georgia, edited by Richard W. Jefferies and Christopher R. Moore, pp. 70-159. University of Kentucky. Sub mitted to Georgia Department of Natural Resources. Moore, Christopher R. and Richard W. Jefferies 2014 Who Were the Guale? Reevaluating Interactions in the Mission Town of San Joseph de Sapala. In New Perspectives on Spanish Missions in the Indigenous Landscape, edited by L. M. Panich and T. D. Schneider, pp. 79-92. University of Arizona Press, Tucson. 2016 Investigating a possible Spanish Military Structure at the Site of San Joseph de Sapala, Sapelo Island, Georgia. Paper preented at the 2016 Conference on Historical and Underwater Archaeology, Washington, D.C. Reitz, Elizabeth J. and Joel Dukes 2008 Change and Stability in Vertebrate Use between the Irene Period and the Mission Period: Nonhuman Vertebrate Remains from Meeting House Field and Fallen Tree. In Native American Landscapes of St. Catherines Island, Georgia: The Data, edited by D. H. Thomas, pp. 778-298. Anthropological Papers of the American Museum of Natural History. American Museum of Natural History, New York. Sassaman, Kenneth E. 1993 Early Pottery in the Southeast Tradition and Innovation in Cooking Technology. University of Alabama Press, Tuscaloosa. Saunders, Rebecca 2000 Stability and Change in Guale Indian Pottery, A.D. 1300-1720. University of Alabama Press, Tuscaloosa. 2009 Stability and Ubiquity: Irene, Altamaha, and San Marcos Pottery in Time and Space. In From Santa Elena to St. Augustine: Indigenous Ceramic Variability (A.D. 14001700), edited by K. Deagan and D. H. Thomas, pp. 83-111. Anthropological Papers of the American Museum of Natural History. American Museum of Natural History, New York. 2012 Deep Surfaces: Pottery Decoration and Identity in the Mission Period. Historical Archaeology 46(1):94-107. Semon, Anna 2015 Exploring Mississippian Check-stamped Ceramics from Mortuary and Village Con texts on St. Catherines Island, Georgia, paper presented at the 72nd Annual meeting of the Southeaster Archaeological Conference, Nashville, TN. Simpkins, Daniel L. 1980 A Preliminary Report on Test Excavations at the Sapelo Island Shell Ring, 1975. In Sapelo Papers: Researches in the History and Prehistory of Sapelo Island, Georgia, edited by D. P. Juengst, pp. 61-75. Studies in the Social Sciences. West Georgia College, Carroll ton, Georgia. Simpkins, Daniel L. and Dorothy J. Allard 1986 Isolation and Identification of Spanish Moss Fiber from a Sample of Stallings and Orange Series Ceramics. American Antiquity 51(1):102-117. Stoltman, James B. 93 2001 The Role of Petrography in the Study of Archaeological Ceramics. In Earth Sciences in Archaeology, edited by Paul Goldberg, Vance T. Holliday, and C. Reid Ferring, pp 297-326. Kluwer Academic/Plenum Publishers, New York. Thomas, David Hurst 2008a Foragers and Collectors, Fishers, and Farmers: Implications for St. Catherines Island Archaeology. In Native American Landscapes of St. Catherines Island, Georgia: The Theoretical Framework, edited by D. H. Thomas, pp. 234-292. Anthropological Papers of the American Museum of Natural History No. 88, part I. American Museum of Natural History, New York. 2008b Foragers and Collectors, Fishers, and Farmers: Implications for St. Catherines Island Archaeology. In Native American Landscapes of St. Catherines Island, Georgia: The Theoretical Framework, edited by D. H. Thomas, pp. 234-292. Anthropological Papers of the American Museum of Natural History No. 88, part I. American Museum of Natural History, New York. 2008c The Guale People. In Native American Landscapes of St. Catherines Island, Georgia: The Theoretical Framework, edited by D. H. Thomas, pp. 22-25. Anthropological Papers of the American Museum of Natural History No. 88, part I. American Museum of Natural History, New York. 2008d Melding the Ceramic and Radiocarbon Chronologies. In Native American Landscapes of St. Catherines Island, Georgia: The Data, edited by D. H. Thomas, pp. 404-434. Anthropological Papers No. 88, Pt. II. American Museum of Natural History, New York. 2008e When Should a Forager Farm? In Native American Landscapes of St. Catherines Island, Georgia: The Theoretical Framework, edited by D. H. Thomas, pp. 198-210. An thropological Papers of the American Museum of Natural History No. 88, part I. American Museum of Natural History, New York. Thompson, Victor 2006 Questioning Complexity: The Prehistoric Hunter Gatherers of Sapelo Island, Georgia. Ph.D. dissertation, Department of Anthropology, University of Kentucky, Lexington. Thompson, Victor D. 2007 Articulating Activity Areas and Formation Processes at the Sapelo Shell Ring Com plex. Southeastern Archaeology 26(1):91-107. 2009 The Mississippian Production of Space through Earthen Pyramids and Public Build ings on the Georgia Coast, USA. World Archaeology 41(3):445-470. Thompson, Victor D. and C. Fred T. Andrus 2011 Evaluating Mobility, Monumentality, and Feasting at the Sapelo Island Shell Ring Complex. American Antiquity 76(2):315-344. Thompson, Victor D., Matthew D. Reynolds, Bryan Haley, Richard Jefferies, Jay K. Johnson and Laura Humphries 2004 The Sapelo Shell Ring Complex: Shallow Geophysics on a Georgia Sea Island. Southeastern Archaeology 23(2):192-201. Turck, John A. and Clark R. Alexander 2013 Coastal Landscapes and Their Relationship to Human Settlement on the Georgia Coast. In Life among the Tides: Recent Archaeology on the Georgia Bight, edited by V. D. 94 Thompson and D. H. Thomas, pp. 169-189. Anthropological Papers No. 98. American Museum of Natural History, New York. US Climate Data 2017 Climate Sapelo Island - Georgia. Electronic document, http:// www.usclimatedata.com/climate/sapelo-island/georgia/united-states/usga0501, accessed May 5, 2017. Voss, Barbara L. 2008 Gender, Race, and Labor in the Archaeology of the Spanish Colonial Americas. Current Anthropology 49(5):861-893. Worth, John E. 2007 The Struggle for the Georgia Coast. University of Alabama Press, Tuscaloosa. 2009 Ethnicity and Ceramics on the Southeastern Atlantic Coast: An Ethnohistorical Analysis. In From Santa Elena to St. Augustine: Indigenous Ceramic Variability (A.D. 1400-1700), edited by K. Deagan and D. H. Thomas, pp. 179-207. Anthropological Papers of the American Museum of Natural History. American Museum of Natural History, New York. 95 ...