The Java Plantation, and 18AN1285: Camp Letts. Rhode River Region, Anne Arundel County, Maryland

FY 2007 MHT NON-CAPITAL HISTORIC PRESERVATION GRANT

Limited Phase III Investigations at 18AN339: the Java Plantation, and 18AN1285: Camp Letts. Rhode River Region, Anne Arundel County, Maryland

VOLUME III

Written by Stephanie Taleff Sperling, Anne Arundel County’s Lost Towns Project Principal Investigators: C. Jane Cox, Assistant Director Anne Arundel County’s Lost Towns Project and Dr. Al Luckenbach, Director, Anne Arundel County’s Lost Towns Project. Submitted to: Maryland Department of Planning Maryland Historic Trust 100 Community Place Crownsville, MD 21032

DRAFT REPORT Submitted June 2008

ABSTRACT

This report, the final in a three-volume series, is the culmination of the multi-year research program. The first year’s efforts of survey and limited assessment of resources in the Rhode River region resulted in the relocation, identification, and updated or new site forms on forty-six sites in the watershed. That first years’ research also produced a comprehensive cultural and historic framework for the Rhode River area and placed it in regional context. In year two, assessment and evaluation was undertaken on five sites to determine their National Register eligibility. Sampling these representative sites allowed for further refinement of the regions’ historic and cultural context, details of which can be found in Volume II of this series.

Finally, in year three, Phase III level investigations were undertaken at two of the most promising and representative sites in the watershed, 18AN1285 (a Middle Woodland period prehistoric site) and 18AN339 (an expansive multi-component historic plantation.) While the data recovery efforts produced a wealth of incredibly rich information, the results demonstrate how much more can be learned from these sites. The sampling strategy devised for each site covered only a fraction of the site boundaries yet was still exceptionally informative. This work will provide a valuable starting point for further scholarly investigations of these two complex archaeological sites in the Rhode River drainage and serves to place each site within a broader historic context. As will be expanded upon in the report below, while much has been learned, valuable research questions for both sites remain and their potential to yield valuable information is intact.

MANAGEMENT SUMMARY

Anne Arundel County’s Lost Towns Project, in cooperation with the Anne Arundel County Trust for Preservation, Inc., began a multi-year investigation of the Rhode River drainage in 2006. The goal for the multi-year project was to survey, assess, and investigate archaeological resources within a limited watershed. The first year of survey and assessment resulted in a comprehensive look at the region’s archaeological resources and developed a region-specific historic context for the study and appreciation of these forty-six sites. The first phase of this work (detailed in Volume I of this series) gathered important information, updated existing site forms, and filed new site forms with the Maryland Historical Trust. The report also detailed major threats to the region’s archaeological resources, and as important, identified several sites that offered the best opportunity for more detailed studies. In the second year of work, the Lost Towns Project identified five representative sites on which an assessment and evaluation strategy was applied. Phase II investigations were conducted at five sites (three prehistoric and two historic), the results of which can be found in Volume II of this series. From these Phase II investigations, the final year’s research program grew as the team selected one representative historic site and one prehistoric site to investigate in more detail.

Data recovery was conducted on these two sites to more fully explore at least two millennia of cultural trends within the Rhode River region. The Phase III testing strategy involved additional test unit excavations at each site. Five additional 5 ft. square units were excavated at 18AN1285 (Camp Letts Middle Woodland site) and while original plans were for the excavation of 10-15 additional 5 ft. square units at 18AN339 (Java Plantation, a multi- component historic plantation), ultimately, 25 additional units were excavated this season.

Investigations at the Camp Letts site (18AN1285) revealed exciting and unexpected information about the Mockley, or Selby Bay, phase of the Middle Woodland period of prehistory (ca. A.D. 200 - A.D. 900). After careful analysis of the artifact assemblage and consultation with the prehistoric pottery experts at Temple University, it was determined the site shows evidence of attempted pottery manufacture on the Rhode River. Hundreds of partially fired, crumbly, friable, shell-tempered, near-pottery sherds were recovered from one of the shell

middens on site, clearly demonstrating a failed attempt at making coiled clay pots in the vicinity. This, coupled with the relative dearth of rhyolite recovered from the site (a non-local lithic type usually found in abundance on sites of this time period), demonstrates that Camp Letts offers a new perspective on the presumably transient and temporary nature of a short-term resource procurement camp from the Middle Woodland time period.

Site 18AN339, the historic Java Plantation (alternatively called Sparrows Rest, Squirrel Neck, or Contee’s Farm throughout its hundreds of years of occupation), has proven to be a rich, yet complicated and multi-facetted site; one that this limited Phase III study has only begun to comprehend. Shovel-test pit data immediately around the 18th century mansion ruins suggested an earlier 17th century loci in the western yard, thus the Project team began excavation of full 5- ft. square units in that vicinity. The 28 excavation units, many removed in a block excavation, revealed a plethora of intact features, including a substantial brick chimney base, associated structural postholes and molds, possible root cellars or small pits, and a trash-laden oyster midden. None have yet been sampled, but observation of exposed but imbedded artifacts and the temporal range of the artifacts in overlying layers, suggests that many of these features are associated with the 17th century occupation of the site by the Sparrow family. The data also strongly indicates that while this area was originally a primary domestic dwelling for the owners of Sparrows Rest plantation in the 17th century, the area was subsequently used both domestically and industrially, potentially as servants quarters and/or as a outbuilding associated with the later 18th century mansion. The site was actively in use for more than 300 years, making interpretation and analysis challenging.

Throughout these investigations, the Lost Towns Project has emphasized public outreach efforts, including volunteers, interns, and the public in our discoveries, through hands-on, in the field archaeology, lab experiences, and site tours in cooperation with the neighboring Smithsonian Environmental Research Center (SERC). Our on-going research, and efforts toward public outreach, has resulted in an exciting new partnership with SERC. In late Spring 2008, SERC announced the acquisition of the Kirkpatrick-Howat Farm. Nearly 600 acres of the farm has been incorporated into the SERC property holdings, including the 18AN339 site, along with several others detailed in Volume I of this report series. The SERC management has

iii

expressed great interest and commitment toward incorporating the cultural and historic resources on this newly acquired parcel into their long-standing program of land stewardship. With SERC’s underlying mission of research and public education in efforts to show linkages between land and water ecosystems, the historic role of people on the landscapes, as discovered through scientific archaeological research at 18AN339, promises a unique and valuable new perspective to the SERC programming. While no one could have envisioned this fortuitous acquisition, fortunately, the multi-year Rhode River project reported upon in these three volumes will be of practical and real use, as SERC looks forward to developing a strategy for the long-term management of cultural resources on its lands.

From the very first efforts three years ago, to the new and exciting SERC partnership, this project can be seen as a model for comprehensive research identification, assessment, and limited, focused investigations, all conducted with an eye towards education and ultimately stewardship. Today, we have a much better understanding of the resources, and potential research potential within the region, an understanding that extends to adjacent watersheds and the surrounding environment. While the Rhode River watershed was ultimately selected as the core research area, as can be seen in the Phase III reports offered below, the scope of cultural influence from the occupants of these sites reverberates throughout Anne Arundel County, Maryland, and beyond. The comprehensive study of prehistoric and historic populations who called the Rhode River area home, emphasizes how interrelated human populations have been, and continue to be, whether socially, politically, or environmentally. Not only does this demonstrate a broad pattern of human prehistory and history, it can perhaps teach us something about our modern condition and how human populations interact today.

ACKNOWLEDGMENTS

This final year of work on the Rhode River assessment and evaluation project was successful only because of the dedication of countless people. First, we must thank the staff at YMCA Camp Letts, including Stacie Vollentine, Associate Executive Director, for welcoming us and being so accommodating. We are also grateful to Betsy Kirkpatrick-Howat for permitting us three amazing years of excavating site 18AN339, located on her land. The recent acquisition of her property by SERC will extend and further the partnership that the Lost Towns Project has developed over the years with that organization. SERC Director Anson “Tuck” Hines and Outreach Coordinator Karen McDonald have been steadfast supporters of preserving and interpreting the rich cultural resources on their property, and we look forward to a successful future collaboration with them.

The professional staff of the Lost Towns Project is owed a heartfelt thanks for their many long hours of digging in extreme conditions, all the while educating and assisting the countless volunteers and interns that make our work possible. The results here are the product of their individual personal qualities. Director Al Luckenbach, Assistant Director C. Jane Cox, Archaeologist Extraordinaire Shawn Sharpe, Archaeologist and Volunteer Coordinator Jessie Grow, Archaeologist and Intern Coordinator Lauren Schiszik, Lab Director Erin Cullen, Lab Specialist Carolyn Gryczkowski, and our newest Archaeological Assistants, Maria Valverde and Steve Tourville, made excavating and interpreting both sites a pleasure.

We are deeply indebted to our field and lab volunteers who dedicate their free time to joining us in the thrill of discovery, including Lois Nutwell, Dave Turner, Tracy Beer, Diana Keener, Dave McKenna, Sarah Sandifer, David Stewart, and Cindy Olsen. We are also grateful for the help of interns, whom we hope will take the archaeological methods and techniques they’ve learned at the Project and start on their own career paths, including Stuart Biggs, Matthew Foley, Sally Gordon, Elva (Liza) Krohn, Christine Kujath, Jessica Lester, Michael McCleary, Emily Mineweaser, Christie Richardson, Vincent Shirbach, Victor Furtado de Mendoça Torres.

Special thanks goes to Dr. Michael Stewart and his students, Joe Blondino and George Pevarnik, at Temple University in Philadelphia for enthusiastically helping us interpret the complicated ceramic assemblage from the Camp Letts site.

Finally, January Ruck, a graduate student intern from the Historic Preservation Master’s program at the University of Maryland, wrote a thought-provoking thesis entitled “Reintegrating Public History & Environmental Education: Preservation and Interpretation of the Ruin at Java Plantation, Edgewater, Maryland” (2008) where she provided an argument justifying the expenditure of financial and human resources regarding stabilizing the remaining portions of the old brick mansion at site 18AN339. We are grateful for her research and analysis as we partner with SERC, who will attempt to preserve the graceful old ruins.

TABLE OF CONTENTS

ABSTRACT ...... i MANAGEMENT SUMMARY...... ii TABLE OF CONTENTS...... vii LIST OF TABLES ...... xiii INTRODUCTION...... 1 FIELD AND LABORATORY METHODOLOGY ...... 4 CHAPTER ONE: 18AN339 ...... 5 Introduction ...... 5 Historic Summary ...... 6 18AN339 Previous Archaeology...... 21 2007 Excavations ...... 22 Features at 18AN339...... 31 Feature 1 ...... 33 Features 2, 4, & 5 ...... 33 Features 3 & 3a ...... 34 Features 6, 7, 8, & 9 ...... 35 Features 10 & 11 ...... 37 Features 12 & 13 ...... 38 Feature 14 ...... 38 Features 15 and 16 ...... 38 Features 17, 18, 19, & 20...... 39 Features 21 and 22 ...... 40 Feature Summary ...... 41 Artifact Analysis...... 43 Kitchen-Related Artifacts ...... 45 Architectural Materials ...... 57 Arms-Related Materials ...... 62 Horse Furniture ...... 64 Personal Materials ...... 65 Slag ...... 75 Prehistoric Artifacts ...... 77

vii

Summary of Artifacts...... 79 Site 18AN339 Summary...... 82 CHAPTER TWO: 18AN1285 ...... 89 Introduction ...... 89 18AN1285 Previous Archaeology...... 91 2007 Excavations ...... 92 Units 4, 7, and 8 - Excavation Block...... 95 Unit 5 ...... 112 Unit 6 ...... 120 Site 18AN1285 Discussion ...... 121 Site 18AN1285 Summary...... 130 CHAPTER THREE: The Future of Cultural Resources on the Rhode River ...... 134 Management Plan for 18AN339...... 134 Management Plan for 18AN1285...... 135 The Rhode River Region...... 136 REFERENCES CITED...... 138 Appendix One: Artifact Catalogs for 18AN339 and 18AN1285...... 141 Appendix Two: Staff Qualifications...... 142 Appendix Three: Revised Site Forms for 18AN339 and 18AN1285...... 143 Appendix Four: Revised Determination of Eligibility Forms for 18AN339 and 18AN1285 ...... 144

viii

LIST OF FIGURES

Figure 1: Locations of 18AN339 and 18AN1285, shown on portion of the 1957 7.5 minute USGS quadrangle, South River, Maryland...... 3

Figure 2: Site plan, with excavation block, numbered units outside of block, and mansion location, at 18AN339...... 23

Figure 3: Results from the magnetometer survey from the west lawn at 18AN339...... 24

Figure 4: Partial comparison of assemblages from Units 4 and 5 at 18AN339...... 27

Figure 5: Numbered excavation units within block at 18AN339...... 30

Figure 6: Site plan showing numbered features in the excavation block at 18AN339...... 32

Figure 7: Features 2 (the brick hearth), 4, and 5 (burned areas inside hearth) at 18AN339; Feature 3 (the shell midden) visible to west...... 34

Figure 8: Feature 3 (the shell feature) shown in relative position to Feature 2 (the brick hearth, partially exposed in background) and the Squirrel Neck/Java mansion ruins at 18AN339...35

Figure 9: Feature 7 (to right) and Feature 8 (to left) at 18AN339, facing south; Feature 16 (brick rubble pile) partially visible to top of frame...... 36

Figure 10: Feature 9 (the post hole and mold) at 18AN339, visible in lower right...... 37

Figure 11: Features 15 (the rock rubble pile, shown to top), 16 (the brick rubble pile, center), 7 and 8 (the two post holes and molds, shown to right) at 18AN339, facing southwest ...... 39

Figure 12: Features 21 and 22 (shown left and center) in proximity to Feature 3 (the shell feature) at 18AN339...... 41

Figure 13: Total artifacts recovered from 18AN339 ...... 44

Figure 14: Total artifacts without brick at 18AN339...... 45

Figure 15: Total ceramics (minus pipes) recovered from 18AN339...... 46

Figure 16: Total earthenwares recovered from 18AN339...... 47

Figure 17: Example of tin-glazed earthenware sherds recovered from main excavation block at 18AN339...... 48

Figure 18: Total stoneware recovered from 18AN339...... 50

Figure 19: Selection of Rhenish stoneware, English brown stoneware, refined white earthenware, Rockingham, and North Italian slipware recovered from excavation block at 18AN339...... 51

Figure 20: Porcelain sherds with overglaze decal sunburst decoration recovered from Strata 2 and 3 at 18AN339...... 52

Figure 21: Nineteenth century soap dish from 18AN339...... 53

Figure 22: Total faunal remains recovered from 18AN339 ...... 54

Figure 23: Two mammal bone fragments with butcher marks recovered from the excavation block area at 18AN339...... 55

Figure 24: Iron utensils recovered from excavation block area at 18AN339 (l-r: knife, three-tined fork, two-tined fork, spoon bowl)...... 56

Figure 25: Total nails recovered from 18AN339...... 58

Figure 26: Window leads recovered from Unit 23 at 18AN339 with makers marks “*WM*” (l) and “*1671*” (r)...... 60

Figure 27: Copper alloy wrist escutcheon with knight motif from 18AN339...... 63

Figure 28: Conserved 18th century snaffle bit from 18AN339...... 64

Figure 29: Sample of buttons, buckles, and beads recovered from in and around the main excavation block at 18AN339...... 67

Figure 30: Obverse (l) and reverse (r) of 1724 George I halfpenny from 18AN339...... 68

Figure 31: Two bone comb fragments recovered from site 18AN339...... 69

Figure 32: Lead alloy cast toy fragment from 18AN339...... 70

Figure 33: Scissors and whetstone fragment recovered from excavation block at 18AN339...... 71

Figure 34: Pipe stem bore diameters from 18AN339...... 72

Figure 35: Marked tobacco pipes recovered from excavation block at 18AN339 ...... 73

Figure 36: Lead cloth seals recovered from Unit 12 (l) and Unit 10 (r) at 18AN339 ...... 74

Figure 37: Distribution of slag (by weight in oz.) around excavation block with portion of conjectural earthfast building footprint at 18AN339...... 76

Figure 38: Distribution of wrought nails (by weight in oz.) around excavation block with portion of conjectural earthfast building footprint at 18AN339...... 76

Figure 39: Results of the magnetometer survey from 18AN1285...... 93

Figure 40: Site plan, with numbered excavation unit locations, from 18AN1285 ...... 94

Figure 42: Plan view of midden with nearby features in Units 4, 7, & 8 at 18AN1285 ...... 98

Figure 43: Portion of ceramic assemblage recovered from the shell midden in Unit 8 at 18AN1285; (top) demonstrates diversity of color and size of shell-tempered sherds, including the sherd adhered to an oyster shell, three distinct coils are visible up-down in this sherd; (lower left) degraded sandstone pieces shown to left, very low-fired or unfired clay ball shown lower right; (lower right) friable, crumbly, shell-tempered sherds...... 100

Figure 44: Portion of ceramic assemblage recovered from shell midden (Stratum 3) in Unit 7 at 18AN1285...... 101

Figure 45: Plan view of Feature 1 at 18AN1285 ...... 104

Figure 46: Profile of bisected Feature 2 at 18AN1285...... 104

Figure 47: Total ceramic assemblage from Feature 1 at 18AN1285, including (clockwise from upper left) very low fired clay fragments, cord-marked Mockley, and burned Mockley that exhibits fireclouding...... 105

Figure 48: North wall profile of Unit 4, including profile of excavated Feature 1 at 18AN1285 ...... 106

Figure 49: North wall profile of Unit 4 at 18AN1285...... 107

Figure 50: Fabric-impressed Mockley sherds from Feature 2 at 18AN1285 (top) and plan view of excavated feature (bottom)...... 108

Figure 51: South wall profile of Unit 4 (to left) and Unit 7 (to right) at 18AN1285...... 109

Figure 52: South wall profile of Units 4 and 7 at 18AN1285 ...... 110

Figure 53: East wall profile of Unit 8 at 18AN1285 ...... 110

Figure 54: East wall profile of Unit 8 at 18AN1285 ...... 111

Figure 55: Unit 5, after excavation of shell midden (Strata 3 and 4), visible in north and west walls; facing north ...... 116

Figure 56: Ceramic assemblage from Unit 5, Stratum 5 at 18AN1285; most were recovered from just to exterior (east) of midden ...... 116

Figure 57: North (top) and west (bottom) wall profiles of Unit 5 at 18AN1285...... 118

Figure 58: Profiles from Unit 5 at 18AN1285 ...... 119

Figure 59: Total historic artifacts recovered from 18AN1285...... 122

Figure 60: Total prehistoric ceramics recovered from 18AN1285 ...... 123

Figure 61: Total lithics recovered from 18AN1285 ...... 125

Figure 62: Distribution maps from 18AN1285...... 128

Figure 63: Comparison of prehistoric artifacts from three excavation areas at 18AN1285...... 129

xii

LIST OF TABLES

Table 1: Features Identified at 18AN339...... 31

Table 2: Total Vessel Glass Recovered from 18AN339...... 53

Table 3: Personal Items Recovered from 18AN339...... 66

Table 4: Total Pipes Recovered from 18AN339...... 71

Table 5: Total Prehistoric Ceramics Recovered from 18AN339 ...... 78

Table 6: Total Prehistoric Lithics Recovered from 18AN339 ...... 79

Table 7: Total Assemblage from Unit 4 at 18AN1285...... 95

Table 8: Total Assemblage from Unit 7 at 18AN1285...... 97

Table 9: Total Assemblage from Unit 8 at 18AN1285...... 102

Table 10: Total Assemblage from Feature 1 at 18AN1285...... 106

Table 11: Total Assemblage from Unit 5 at 18AN1285...... 114

Table 12: Total Assemblage from Unit 6 at 18AN1285...... 120

Table 13: Total Prehistoric Artifacts Recovered from 18AN1285 ...... 123

Table 14: Total Debitage and Tools from 18AN1285...... 126

Table 15: Comparison of Prehistoric Artifacts from Three Excavation Areas at 18AN1285.....129

xiii

INTRODUCTION

The following report details the results of the limited Phase III excavations conducted during Year Three, the final year of the Investigation of Significant Archaeological Properties in the Rhode River Drainage. After two years of broadly investigating the rich cultural resources of the watershed, Lost Towns Project archaeologists chose one historic and one prehistoric site with very high research potential to study more fully. Java Plantation (also known as Sparrows Rest, Squirrel Neck, or Contee Farm during its long history) (18AN339), occupied continuously since at least the third quarter of the 17th century, is an incredibly rich historic period site with a minor prehistoric component. The Camp Letts site (18AN1285), a seasonally occupied Middle to Late Woodland period camp site, offered the prospect to further the knowledge base about this little understood cultural period. Excavations at both sites provided a unique and exciting opportunity to investigate largely undisturbed and highly intact resources of the Rhode River drainage (Figure 1).

This third year of the Rhode River survey also offered the Lost Towns Project the opportunity to conduct significant public outreach, by means of engaging volunteers, interns, and the community in our discoveries through hands-on field and lab work at both sites. This also helped to foster the Project’s mission to strengthen the connection between past and present, increase public awareness of the rich cultural past of Anne Arundel County, and promote historic and environmental preservation and conservation. In doing so, a number of educated amateur archaeologists have been created that will serve as stewards for the protection of cultural resources in this region and beyond.

The first step in each investigation this year was a geophysical survey, which assisted in the placement of initial excavation units. At Java Plantation, this led to exciting finds in the west yard of the ruins of the circa 1750 brick Georgian mansion house. Located only about 100 ft. from the two chimneys that mark the location of that once grand home, the remains of a 17th century earthfast building was found, providing a rare glimpse into the first chapter of Anne Arundel County history. Extensive documentary research into the history of the property revealed that this was the location of one of the houses at Sparrows Rest, both mentioned by Thomas Sparrow II in his 1675 will. Thousands of artifacts from the Sparrow family were

recovered from the excavation units, along with materials from the 18th and early-19th century Maccubbin occupation and the later 19th century Contee family who lived in the brick mansion. This intensive temporal mixing of material culture has proven to make site interpretation an exciting challenge and has raised many questions for future research.

At Camp Letts, the results of the excavation raised a number of interesting questions about the activities undertaken by the people of the later Middle Woodland time period, locally referred to as the Selby Bay phase (ca A.D. 200 – A.D. 900). A number of units were excavated that straddled two separate oyster shell middens, providing a good comparison between what was thrown out with the trash into the midden and what was happening at midden’s edge. Two intact features were found and excavated on the edge of one midden, and hundreds of partially fired earthenware sherds were found within this midden, suggesting that native people who lived here were engaged in pottery manufacture. This, along with an unusual collection of lithics, provides a new perspective on the lifeways of the people that roamed the Rhode River hundreds of years before contact with Europeans.

Overall, the archaeological investigations undertaken in the Rhode River watershed in 2007 have provided a bounty of fascinating data, integral to our understanding of the unique patterns of cultural and historic development in Anne Arundel County. The following report will begin by discussing the findings at site 18AN339 (Java), followed a discussion of the findings at site 18AN1285 (Camp Letts). First, the findings of the extensive documentary research conducted on the site history of the Java Plantation will be presented, followed by the results of the archaeological excavations there. The archaeological finds from site 18AN1285, Camp Letts, will then be presented, along with a general discussion of the prehistoric past of Anne Arundel County. In both cases, a great effort has been made to discuss the findings in a technical, yet approachable, and understandable way. This report concludes with a discussion of the future of the cultural resources in the Rhode River drainage.

Figure 1: Locations of 18AN339 and 18AN1285, shown on portion of the 1957 7.5 minute USGS quadrangle, South River, Maryland

FIELD AND LABORATORY METHODOLOGY

In 2005 and 2006, limited shovel testing was conducted at sites 18AN339 and 18AN1285. No additional shovel tests were excavated in this third year of the survey. The first step in the 2007 season was to conduct a magnetometer survey on each of the sites. The data gathered from both of these surveys assisted in the initial placement of the 5 ft. by 5 ft. excavation units. Thereafter, the placement of the units was based on prior research goals or new questions inspired by the data collected during the investigations.

All units were hand excavated using shovels, spades, or trowels. All soils were screened through ¼ in. mesh hardwire screen. All artifacts were collected 100%, with the exception of oyster shells (only those with “female” hinges were collected), and bricks at site 18AN339, where only those with distinct edges or markings were taken to the lab for processing.

Plan views were hand drawn in the field of the base of excavations in all units, and plans were drawn on separate graph paper of all significant features. Profiles of significant unit walls were also drawn on graph paper. These images were later digitized using AutoCAD software. Digital photographs were taken of all plan views and profiles and are stored on the Anne Arundel County government servers.

All artifacts were washed, sorted, counted, weighed, rebagged, and labeled according to Maryland State standards. The collections were cataloged using FileMaker Pro software. This data was later entered into an Excel spreadsheet for ease of analysis and for producing the tables and charts presented in this report. The artifact assemblages from both sites 18AN339 and 18AN1285 are permanently stored at the Anne Arundel County laboratory at Londontown, Maryland.

CHAPTER ONE: 18AN339

Limited Phase III Excavations at Java Plantation, a.k.a. Sparrows Rest, Squirrel Neck, or Contee Farm

Introduction

The Georgian mansion originally called Squirrel Neck is a landmark for people traveling the Rhode River, as it has been since its circa 1750 construction. Today, the graceful ruins, consisting of little more than two chimneys, can still be seen for miles and is a highlight of the Java History Trail, located on the SERC campus. It was this ruin that first drew The Lost Towns Project to excavate in its shadow in 2005, and it has provided a lovely backdrop to our excavations for the subsequent two field seasons. This third year of excavating at site 18AN339, referred to in this report as “Java” after the well-known 19th century name of the plantation, has revealed that the archaeological resources below the ground are equally as stunning as the observable architectural resource.

Earlier excavations made it clear that there was an earlier phase of occupation on the hilltop that pre-dated the mid-18th century construction of the mansion. A test unit dug in 2006 along the eastern front of the mansion made it clear that an earlier house had not simply been absorbed into the later construction. The data acquired during the 2006 field season and a 2007 remote sensing survey helped determine the focus of this year’s excavations. Many units later, the earlier, 17th century Thomas Sparrow family occupation was revealed to be situated only about 100 ft. from the 18th century mansion. The remains of a brick hearth, a few post holes, and thousands of 17th century domestic artifacts signaled the location of one of the earthfast dwellings at Sparrows Rest.

It was also made clear this season that the excavated soil strata at Java are heavily mixed with a collection of artifacts that span the entire period of site occupation, including a minimal prehistoric component. None of the features exposed this year were excavated, and the assumption must be that these too span the long period of site occupation. The strong concentration of 17th century materials makes it clear, however, that an area of intense occupation during the Sparrow occupation of the land has been pinpointed. It stands to reason

that the proximity to the mansion and increasing availability of goods through the centuries would provide for a certain admixture of later artifacts.

The following chapter begins with a narration of the most comprehensive historic summary written thus far for site 18AN339, produced after conducting additional documentary research. Next, a brief summary of the 2006 excavations is presented, followed by a technically detailed description of the results of the 2007 field season. An analysis of the exposed features and artifact assemblage follows. The chapter ends with a summary of the knowledge accumulated to date on site 18AN339 and questions to pose for future research.

Historic Summary

The land previously owned by the Kirkpatrick-Howat family, and recently acquired by the Smithsonian Land Trust, a non-profit organization, has been known by many names during its 350 year occupation, but was held by only four families for most of its history. In the 17th century, the land was called “Sparrows Rest”, “Sparrows Addition”, and “Locust Neck”. The Thomas Sparrow family owned the property for much of the 17th and early 18th centuries. By the mid-18th century, it was called “Squirrel Neck” and it was possessed by Nicholas Maccubbin and then his son, James (Maccubbin) Carroll. The 19th century brought the ownership of the Contee family, who called it “Java.” The ancestors of the Kirkpatrick-Howat family purchased the property in the early 20th century.

The first Englishman to lay claim to the land was Thomas Sparrow (I). Sparrow (I), born circa 1620 in England, immigrated to Virginia by 1635. In 1640, he married Elizabeth Marsh in Lower Norfolk County, Virginia, who herself had arrived in Virginia before 1637 (Russell ND). In 1649, the young couple, along with their servant, John Dennis, and their two young children, Thomas (II) and Elizabeth, came to Anne Arundel County with the Puritan Providence settlers (Sparrow 1990). In Maryland, Elizabeth Sparrow would give birth to second son, Solomon, around 1652. Sparrows Rest, containing 590 acres and the present location of site 18AN339, was patented on September 22, 1652 on the “West side of Cheasapeak Bay and on the West side of Road River” for Thomas Sparrow (I) (MSA, Anne Arundel County Patent Certificate, AB & H, folio 282). Sparrow also patented land on the south side of the Patapsco River in 1652, which would later be called Sparrows Point (this parcel of land, located at the mouth of the river and

presently occupied by the Bethlehem Steel Company, is still called by this name). Sparrow (I) died by September 1659, when Sparrows Rest was granted to “Thomas Sparrow son and heir of the first named Thomas Sparrow” (MSA, Anne Arundel County Patent Certificate, Liber 4, folio 97).

The Sparrow family had converted to Quakerism in the 1650s, and Thomas Sparrow (II) was such a devout Friend, he was reportedly fined five hundred pounds of tobacco for refusing to bear arms in the colonial militia in 1661 (Sparrow 1990). This Thomas Sparrow married Elizabeth Kinsey in 1659 (who later married Richard Johns, noted Calvert County Quaker and ancestor of the 19th century philanthropist Johns Hopkins) (Sparrow 1990). Their children included Charity, Solomon, Thomas (III), and Elizabeth. Sparrow (II) sold a portion of Sparrows Rest, called Locust Neck, to Major Thomas Francis on October 15, 1667 (Chancery Court Proceedings, Archives of MD, Vol. 51, folio 227). Francis later patented 42 acres of land called Francis His Addition (MSA, Anne Arundel County Patent Record 15, folio 22: 1674). Archival research suggests that Locust Neck and Francis His Addition may occupy the same area as the northeastern portions of the 140 acre Sparrows Addition, which was patented by the Sparrow family in 1675 (MSA, Anne Arundel County Patent Records 19, folio 8). Sparrows Addition encompassed present day Contee’s Wharf, Sheepshead Cove, and extended up Sellman Creek to Nettlefolds Branch, today called the Cypress Bog.

Thomas Sparrow wrote his will on January 1, 1674/5, where he begins by stating “I Thomas Sparrow of the County of Ann Arundell in the Province of Maryland Planter being Sick and Weak in Body but of a Sound Disposing mind taking into Serious Consideration that all men are Mortall and that mans life is but Transitory as the flower of the field…” (MSA, Anne Arundel County Wills, Liber 2, folio 76:1675). He goes on to state “I do give and bequeath my Son Thomas Sparrow all my plantation which I now Live on…” along with his parcels on the Patapsco, and specifically states that the land should stay within the male line of Sparrows forever. He wills his wife, Elizabeth, to have use of half of the plantation until Thomas (III) reaches the age of 21 (he was born circa 1660, which would make him about 14 at the time), and requests that “the building now begun upon my now dwelling plantation to be finished with all Convenient Speed”. He goes on to will the “use of that parcel of Land which my sister Elizabeth lives…with the timber house and what Else is necessary for her Occupation upon the Said Land

from the time of her first going on the Said Land until Eleven years being fully Completed and Ended…”. This document suggests that there were at least two dwellings on Sparrows Rest by 1675, and perhaps more. It is possible that the remains of the impermanent structure newly discovered on the hilltop at 18AN339 could be one of these 17th century houses.

The inventory of Thomas Sparrow (II), dated October 25, 1676, reveals the assemblage of a wealthy 17th century planter. His estate, valued at 27,068 pounds of tobacco, included, among other items, a large collection of farm animals (including “1 old horse”), three feather beds, 12 red leather chairs, three small pewter dishes, one bible, a small silver spoon, and a warrant for 250 acres of land (this valued at 500 pounds), to name a few items (MSA, Anne Arundel County Inventories and Assessments, Liber 2, folio 359). It was neighbors Major Thomas Francis and Richard Tydings (married to Chastity Sparrow) who conducted this inventory.

Following Sparrow’s death, his sons, Thomas (III) and Solomon, and his wife, Elizabeth (now married to Richard Johns), contest the land transaction of the 42 acre Locust Neck parcel between Francis and their father (MSA, Chancery Court Proceedings, 1669-1679 Volume 51, folio 227, 1678.) This valuable water access land is likely the eastern most portion of the Sparrows Rest and Sparrows Addition land located along the shoreline of today’s Sellman Creek. The Sparrow heirs claimed that Thomas (II) died before the deed could be finalized. The court ruled in favor of Francis, and Locust Neck would remain in his family until 1699.

Major Thomas Francis (the elder) drowned in a boating accident on March 19, 1685. He and his wife were reportedly returning from a trip across the river to Tulip Hill, the adjacent plantation owned by the Galloway family. According to local legend, his wife survived because her hoop skirt acted as a life preserver and buoyed her (Samuel Asher, personal communication, 1987). Mr. Asher also mentioned that Francis’ headstone was located in a “small wooded area east of the big house [the Java mansion], about halfway between the house and Sheepshead Cove.” This gravestone was transcribed in the 1950s, but even at that time was reported to be vandalized and lying on the ground. The transcription was written as follows:

Here Lies the body of major Thomas Francies Who deceased March 19 anno 1685 adged 42 years Tho now in silence I am lowly laid here tie that place few mortal made Others fore doe no thou thyself morngre [mourn] morngre ye no more but doe yourself redeme And the fortuner I hope youll plainly see Such future comforts as are blessing thee For the grim death thought tell to pass us Rejoice and thing that we shall summonld be None to be exempted in the eternity Cause then its so grieve ye no more I fear that God should thee at that (anew anger wat sore) Even to death and all to let you see Such grieves to him offensive be

Possession of Locust Neck apparently passed from Major Thomas Francis to his son, also called Thomas Francis. This Francis subsequently left Locust Neck (a part of Sparrows Rest) and Francis His Addition to his wife, Mary Francis, in his will dated August 29, 1698 (MSA, AACo Wills, Liber 6, folio 173). The following year, on September 12, 1699, Mary Francis (now living in Cecil County) sells all of Locus Neck and Francis His Addition (totaling 188 acres) to Thomas Sparrow (III), and the original 590 acres of Sparrows Rest and 100 acres of Sparrows Addition were back in the possession of the Sparrow family (MSA Chancery Court Proceedings, Archives of Maryland, Liber 51, folio 20-26).

Thomas Sparrow (III) had three wives, a total of eight children, was a devout Quaker, an Annapolis merchant, and owned property in both Maryland and in North Carolina (Sparrow 1990). He married Anne Burgess on June 8, 1697 and buried her just a month and a half later, on July 25, 1697 (Russell ND). Later that same year, he married Sophia Richardson on November 11, 1697. Sophia was the mother of Thomas (IV) and Solomon, and after she died in 1705, he contracted William Coale (possibly his uncle, married to his Aunt Elizabeth) and Daniel

Richardson to care for his children for 15 years in exchange for use of Sparrows Rest and Sparrows Addition for 21 years (MSA Liber WT#2, folio 310:1705). The land totaled 690 acres “together with all houses Edifices gardens orchards pastures Emolumnets appendances & appurtenances” and Coale and Richardson would owe Sparrow 100 pounds of “good merchantable wool” yearly for the rent. They were also charged with providing Thomas (IV) and Solomon “good and sufficient meate Drinke Cloathes lodging and washing and also shall and will cause ye said Children to be Educated in such schoole learning as can conveniently be procured in ye Province of Maryland.” This document suggests that this Thomas Sparrow (III) did not reside at 18AN339 during this time.

Sparrow (III) remarried a woman named Ann West by the end of 1705, and they eventually had five children together. During this time, it appears that he acquired at least two more parcels of land in the Rhode River area, called Squirrell Neck and Thomas’s Quarter Plantation, probably located to the south of site 18AN339. In 1712, he leases these two parcels to William Coale for 21 years for the “summ of fifty pounds Current money” and states that rent of “one Barrell of Indian Corn” should be paid to Sparrow yearly “at his Mansion house upon the plantation whereon he now lives” (MSA Liber 1B2, folio 76:1712). What is not clear in this document is where Sparrow and his family are living at the time. If the agreement with Coale from 1705 (now seven years prior) still stands at this point, the Sparrow family are living somewhere other than 18AN339 and Sparrows Rest. His wife, Ann West, was a wealthy woman in her own right (she was the daughter of Col. John West), and it is possible they were living on her family estate or on another landholding. Sparrow is, however, referred to as “Thomas Sparrow of Road River”, so this suggests he might have been living on Sparrows Rest at the time.

Also during this time, Thomas Sparrow (III) was apparently spending time on his plantations near the Pamlico River in North Carolina. He held over 3,000 acres in that state and in 1706 became one of the original landholders in the town of Bath, the oldest incorporated town in North Carolina (Sparrow 1990). In 1711, he participated in Cary’s Rebellion, a religious conflict lead by Thomas Cary (Paschal 1955). Cary, a South Carolinian and a supporter of the Quaker party, served as North Carolina’s governor for a short time in the early 18th century, during which time Bath was the impromptu seat of the colony’s government. His rebellion

centered on Quaker rights in North Carolina after Cary was removed from office by the English crown and replaced by Edward Hyde, a member of the Church of England. During the summer of 1711, Cary assembled a group of men, including Thomas Sparrow (III), and fought numerous armed skirmishes for Quaker rights. Eventually, the royal marines from Virginia sided with Hyde’s forces and overtook Cary and his followers. Cary was arrested and sent back to England and his men were deemed traitors. By 1713, however, the English crown granted them full pardon (Paschal 1955).

Sparrow (III) wrote his will on June 15, 1713. He would live for many years after he wrote this document, and it was not endorsed until May 12, 1719. There is also evidence that he wrote a second will in North Carolina in 1717 (Sparrow 1990:400). In the 1713 will, he states “I Thomas Sparrow of Road River in Annar[undall] County being intended to take a Voyage & not knowing how it may Please god to dispose of me & calling to mind ye uncertainties of this Life doe therefore make Constitue ordaine & appoint this to be my Last will & testament” (MSA Wills Liber 15, folio 101:1713). He begins the will by leaving his island called Crany Island [today called Harkers Island, located near the southern tip of the Outer Banks] in North Carolina to his sons Solomon and John, and follows by giving his son Kinsey “all that parcel of Land I Purchased of Mary Frances being one hundred & forty six acres Part of Sparrows rest formerly Sold by my father to Tho Francis lying on Road River.” This is the only mention of Sparrows Rest in the document, and interestingly, his eldest son, Thomas (IV) is only specifically given “one Cow & Calfe with Seven Years Priviledge of them & theire Increase on Crany Island.” His father does, however, specify that his four sons should divide all of his lands, rights, titles, and interests equally.

Sparrow’s failure to mention the remaining portions of Sparrows Rest and Sparrows Addition in his will could stem from the fact that only seven days prior to the drafting of his will (on June 8, 1713), he leased “by Estimation one hunds acres more or less” of Sparrows Rest to John Gresham for 30 pounds current Maryland money (MSA, Liber 1B#2, folio 45:1713). This document indicates that Gresham was the then-owner of the neighboring plantation called Shaw’s Folly (located north of Sparrow’s Rest), and this lease document was enacted for 21 years. The document was revisited on June 10, 1723 when it appears the lease agreement was further endorsed.

At some point after the death of Thomas Sparrow (III), his eldest son, Thomas Sparrow (IV) acquired Sparrows Rest, Sparrows Addition, and Squirrel Neck, among other parcels of land. This Thomas Sparrow had a “crippling disability” and was unable to farm (Sparrow 1990:400). He instead made his living as an inn holder in Annapolis, and became the crier of the provincial and Anne Arundel county courts, and later acted as the doorkeeper of the General Assembly. Apparently, he also invented a machine to catch and cure fish for export (Sparrow 1990:400). There is also court documentation from November 1747 that he fathered an illegitimate child with a woman named Martha Freeman, and was fined 30 shillings for the act (Russell ND).

Thomas Sparrow (IV) was convinced by his guardian, Thomas Gassaway, to convey Sparrows Rest and Sparrows Addition to him in 1720 (MSA, AACo Deeds, Liber CW#1, folio 373). However, Sparrow later decided to take Gassaway to court to reacquire the parcels. On February 4, 1724, the Chancery Court granted Sparrow ownership of the two parcels, totaling 490 acres (MSA, Chancery Court Proceedings, Liber SY#1, folio 88:1724).

It appears, however, that by 1746, Sparrow (IV) is finally ready to permanently part with his ancestral lands. On September 9, 1746, he sells a portion of Sparrows Rest to Richard Tydings (probably related by marriage) for 10,000 pounds of tobacco (MSA, AACo Deeds, Liber RB 2, folio 252). This portion of Sparrows Rest seems to be further west and inland, west of Muddy Branch. In May of the following year (on May 21, 1747, to be specific), he sells the remainder of Sparrows Rest and Sparrows Addition to Nicholas Maccubbin for 1,000 pounds current Maryland money (MSA, AACo Deeds, Liber RB 2, folio 485). A total of 690 acres were sold to Maccubbin, including the easternmost, waterfront portions of the land along the Rhode River; site 18AN339 is located in this eastern portion of Sparrows Rest. With the selling of this parcel to two different men, the land use history of the area becomes divided, and here ends the 94 year Sparrow possession of the land.

As an interesting side note, after Sparrow (IV) died in 1753, his young son, Thomas Sparrow (V), was left under the guardianship of Jonas and Catherine Green, publishers of the Maryland Gazette (Sparrow 1990). The Greens were neighbors of the Sparrows in Annapolis, and young Thomas became an apprentice at the Gazette. This Thomas Sparrow later became a

silversmith, and during the Revolution, forged the original seal of the State of Maryland, still on display at the State House in Annapolis (Russell ND).

Nicolas Maccubbin was a wealthy merchant-planter from Annapolis, who married Mary Clare Carroll, also of Annapolis, on July 21, 1747, just two months after his acquisition of the Sparrow lands (Lee 2004). This was a marriage of great wealth. Mary’s dowry included 1500 acres of Prince George’s County lands and waterfront land in Annapolis. Upon her father’s death, her brother, Charles Carroll (known as The Barrister), became one of six richest men in Maryland due to his inheritance (Lee 2004).

While there is little documentary evidence to pinpoint exactly when the mansion at Java was built, it is safe to assume that it was erected during the Maccubbin ownership of the land. The mansion at Squirrel Neck (as it was called at the time) is very similar in design to Mount Clare, the home built by The Barrister. Historians generally believe that they were constructed by the same builder, most likely Patrick Creagh (Trostel 1981). Creagh was a well-established Annapolis builder, and also had connections to both the Maccubbin and Carroll families. Judging by the evolution of designs, Squirrel Neck was constructed first, being the simpler of the two buildings. As an aside, the name “Mount Clare” was given to that house as a way to honor the Barrister’s grandmother, Clare Dunn Carroll, and sister, Mary Clare Carroll Maccubbin.

Whoever designed both mansions had an eye toward classic Georgian architecture. The Squirrel Neck house was constructed of brick and had five parts, the center section being two and a half stories tall with a gambrel roof, the two hyphens being one story tall with a gable roof. Two story brick pilasters flanked the central portion of the house, and glazed header bricks running up the center of them were a design element used at both Mount Clare and Squirrel Neck. The Squirrel Neck mansion lacked a water table or belt course, typical elements of a Georgian-period house. The gambrel roof also set the Squirrel Neck mansion apart from many others built in the mid-eighteenth century; houses of the period were more typically built with a hipped or gabled roof, as was the case at Mount Clare. However, Padsworth Farm, a mansion built between 1735 and 1745 on the Patuxent River just seven miles from Squirrel Neck, also possessed a gambrel roof, as did a number of Georgian period houses in the northern colonies (Trostle 1981).

While Maccubbin acquired Sparrows Rest and Sparrows Addition in his 1747 agreement with Thomas Sparrow (IV), he did not file a patent certificate for the parcel of land called Squirrel Neck until 1765 (MSA, Pat. Cert. Folder 1481). During this time, the Maccubbin family was probably spending most of their time in Annapolis, at the ancestral home of Mary Clare Carroll Maccubbin, purchased by her husband in 1746. The family grew in the mid-18th century with the birth of Nicholas in 1751, Charles (in 1755?), James in 1761, and Samuel in 1763. Additional children included John Henry, Mary, and Susanna. The 1776 census listed the slaves possessed by Maccubbin at the time, including eleven black males, eleven black females, and 25 black children (all 22 adults were taxable).

Charles Carroll, the Barrister, died in 1783 and left his estate to his two nephews, Nicholas and James Maccubbin, under the assumption that they would change their names to Carroll (Lee 2004). James would eventually inherit Mount Claire upon his aunt’s death in 1817, and also inherited the brick mansion at Squirrel Neck upon his father death in 1787. He therefore became owner of the two grand houses created by the same builder in the mid-18th century. The will of Nicholas Maccubbin, written on May 28, 1784 (and endorsed on March 15, 1787), states “I give and bequeath to my son James Carroll formerly called Maccubbin the following Lands, Squirrel Neck containing fifty Acres more or less, Sparrows Addition containing one hundred Acres more or less Sparrows Rest containing five hundred and ninety acres more or less” (MSA, AACo Wills, Liber THH2, folio 9: 1787). James (Maccubbin) Carroll was now the owner of what once was the entire Sparrow family landholdings, including “all my Horses, Cattle, Hogs, Sheep, Plantation Utensils, Household furniture, and all other things…” on the land when the elder Maccubbin died. Interestingly, his father specified that James was not to receive all of the slaves on the 740 acres, as he had when distributing his other plantations to his other sons. Instead, he specifies at least 12 slaves by name (along with their children) for James to own, and gives the “Nigro woman named Jenny at Squirel Neck and all her Children” to his daughter Mary (Maccubbin) Brice.

James (Maccubbin) Carroll was an absentee landowner and left Squirrel Neck under the care of a superintendent named William Johnson. The 1798 Federal Tax Assessment lists a five part mansion (two story brick house; one story Passage brick; one story brick Kitchen; one story brick Passage; one story brick Washhouse) valued at $12,000, eight small dwellings, and 40

slaves on the property. This suggests a very large farming operation was being undertaken on the plantation, but it is not clear if anyone was actually living in the Squirrel Neck mansion during this time.

On February 1, 1819, James Carroll sold his entire 740 acres of Sparrow lands to Dennis Boyd, who immediately endorsed the land to John Contee. This left Contee as the owner of the three parcels, although he still owed $30,000 on the sale to James Carroll. The lands were resurveyed on June 15, 1825, and a patent was issued to John Contee three years later on August 12, 1828. It was this patent document that first called the property “Java.” It would not be until November 10, 1832 that Contee finished paying his debt in full (to the heirs of James Carroll, who died January 27, 1832), at which point he finally received the deed (PGCo Wills, PC1-123, Dec. 2, 1839) (Lee 2004).

John Contee was a wealthy landowner from Prince George’s County and served in the U.S. Navy during the War of 1812. He was an officer on the U.S.S. Constitution (also known as “Old Ironsides”) when it captured and destroyed the H.M.S. Guerriere on August 19, 1812 off the coast of Nova Scotia and the H.M.S. Java on December 29, 1812 off the coast of Brazil. Local legend asserts that he purchased the Squirrel Neck lands with prize money he was later given for these captures. However, many historians believe that this prize money would have been nominal to John Contee’s fortune, totaling only about $1000 for an officer of his rank. He had amassed enough wealth by the time he purchased Java that this prize money would not have been a significant factor (Lee 2004). It does seem clear, however, that he called the property “Java” because the battle and his service aboard the Constitution left a lasting impression, and he desired to name his new property after the conquered British ship.

John Contee probably did not reside at Java full time, as he maintained a life-long residence in Prince George’s County. He apparently left the daily operations of the plantation in the hands of an overseer, continuing the tradition started by James (Maccubbin) Carroll. He died in 1839, and willed Java to his second wife, Ann Snowden Contee. The inventory of his landholdings, made in February, 1840, shed some light on the immense size of the farming operations that were taking place on Java at the time. There were 84 slaves listed as living on the property, which would provide a labor force of approximately 40 people. These slaves produced

70,000 pounds of tobacco in 1839, a tremendous amount relative to the other farms in the region. In the stores, there were listed 1,500 pounds of bacon, 1,000 pounds of pork, and four barrels of herring, which would have probably been used to feed all of the slaves. There were few essential items listed as being in the main house, suggesting it was not furnished for day to day living, and the overseer was probably living in a different dwelling on the plantation (Greenburg and Hyatt 1990).

However, Ann Contee and her two young sons were listed on the 1840 census as residents of the First District of Anne Arundel County (the Rhode River area), so she must have moved her family to Java soon after her husband’s death. Evidently, she and her boys moved back to Prince George’s County by 1850, as she is not listed in the Rhode River area census for that year. She is, however, listed in the 1850 Slave Schedule as the owner of 76 slaves residing in the First District, suggesting she was the absentee landlord of a thriving plantation. Contee’s Wharf, located on her property on the Rhode River, must have been an integral part of this plantation system, and it is shown for the first time on the detailed 1846 U.S. Coastal and Geodetic Survey map.

Ann Contee conveyed her interests in Java to her two sons, Charles and Richard, on July 28, 1859. The Contee boys divided the land into a south section belonging to Charles (totaling 360 acres) and a north section belonging to Richard (totaling 380 acres) along the course of an old road (Lee 2004). It is in the northern section where site 18AN339 and the Java mansion are located. On the same day he obtained the land from his mother, Richard Contee mortgaged his portion to his father in law for $23,500 to be repaid in 10 years (MSA, AACo Land Records, NHG8, page 303). He and his brother also borrowed money from the Farmers Bank of Maryland at about the same time (Lee 2004).

The Contee brothers married two sisters of the John Bowling family (Richard married Ann in 1858 and Charles married Betty in 1860) and both couples settled on their respective parcels of Java land. The 1860 Slave Schedule for the First District lists Richard as owner of 34 slaves and Charles as owner of 40 slaves, making them two of the largest slave owners in the Rhode River watershed. The 1860 census lists Richard (age 24) as having $23,000 in real estate and $19,770 in personal estate, while Charles (age 29) has $21,600 in real estate and $21,220 in

personal estate. There are also five white farmhands with the last name “Contee” listed under Charles in this census. All of this suggests that both Contee brothers were still running productive, successful farming operations largely based on slave labor at the start of the Civil War.

However, by the close of the War, both Contee brothers encountered serious financial troubles. The labor-intensive approach they both utilized had become a thing of the past with the abolition of slavery. They defaulted on the loans they took out from the Farmers Bank of Maryland, and the county sheriff was instructed to sell both parcels to the highest bidder. Their father in law, John Bowling, purchased the notes on both farms, so while the Contees were no longer owners of the parcels, they were still permitted to live there.

The 1870 census lists Richard Contee (age 34), his wife Ann (age 29), and two additional Contee women (cousins?), Eloise (age 19) and Sylvia (age 17) on the First District census, but there is no sign of Charles and Betty Contee. Richard’s real estate value is now listed as $20,000 (only $3,000 less than 1860), but his personal estate is listed as only $2,300 ($17,470 less than 1860). According to the 1870 Agricultural Census, he paid $1,000 in wages to farm labors that year, giving him just about $1,300 in profits. Greenburg and Hyatt (1990) estimate that by this time, only about 125 acres of Contee’s landholdings were in production (down from a high of around 270 acres in 1860), meaning that marginal areas of Java, including Francis Field near the river, have not been plowed since the 1860s.

Richard Contee was still living at Java in 1875 when his father-in-law, John Bowling, died, but he had started to lease out portions of the property to tenants. Bowling forgave Contee the debt he was owed for purchased the mortgage, and willed him back his original northern portion of Java. By 1877, Contee and his wife Ann were living in Baltimore and the entire farm had been turned over to tenants. Ann was apparently in poor health by this time, and they filed a petition with the Anne Arundel County courts to have the land sold, rather than retain it under the direction of Bowling’s will, asserting “under the system of tenancy the place [was] rapidly declining in value…from want of proper cultivation and attention [and the buildings were] falling to pieces and there were no means with which to repair them, the proceeds of the place not being sufficient for that purpose even” (Greenburg and Hyatt 1990:I-10). Richard and Ann

reportedly later moved to the Snowden family mansion, Oakland, in Prince George’s County by 1887. This was the ancestral home of Richard’s mother’s family.

John Bowling also willed Charles Contee his original southern portion of Java in 1875. In 1878, Charles conveyed the land to his brother-in-law, also called John Bowling, to be held in estate for Charles’ wife Betty. Charles died in 1887, and after Betty died in 1894 their children sold the land to Thomas Myers in 1896. This portion of Java would later become the successful Java Dairy Farm in the mid-20th century under the ownership of Robert Lee Forrest, who obtained the land in 1915. When Forrest died in 1962, he donated his 368 acres to the Smithsonian Institute, who established the Smithsonian Environmental Research Center (SERC) there in 1965. Since then, their landholdings have expanded to over 2,600 acres of pristine land (Lee 2004:8).

The late 19th century land transactions for the northern portion of Java are not so clear. A Chicago resident called John V. Lemoyne obtained the land sometime after 1882, and he apparently left the farm in the hands of a manager or tenants (Greenburg and Hyatt 1990). Lemoyne apparently used the Java mansion (about 150 years old by this time) as a summer house, but was apparently not there around 1890 when the place was struck by lightning was partially destroyed. The roof and interior framing of the house fell, and portions of the front and rear walls collapsed (Lee 2004). The house was apparently rebuilt using the fire-damaged brick, a fact that would heavily contribute its slow demise. A number of architectural changes were also made to the house. Several bow windows were added to the sides of the house and the new roof dormers were built as “eyebrow” dormers, rather than the more typical shed dormers. The northern hyphen was rebuilt to attach to the old kitchen, but the southern hyphen was not rebuilt. A series of brick-walled terraces were also built south of the house around this time.

Lemoyne sold the land to Jasperson Smith of New York City in 1897 and eventually Mrs. Elizabeth Gordon bought the land in 1917. Mrs. Gordon later married a Mr. Kirkpatrick- Howat, and their son and his wife are today still the owner of this portion of Java. Apparently, soon after buying the land, Mrs. Gordon lived in the reconstructed Java mansion during summers. But even at this time, in 1916, the house is described as “Dwelling Ruins” on a 1916 plat of the parcel. She was determined to rebuild the old mansion, even going to far as to have

bricks delivered. However, it was about this time that the adjoining property (the old Sellman farm) containing a sturdy 19th century brick mansion came on the market. She purchased this farm in 1923 (bringing her total landholdings in the area to about 700 acres) and abandoned the old Contee house (Samuel Asher, personal communication, 1987).

An undated newspaper article (circa 1970?), reminiscing about a trip made to the Java mansion in the mid-20th century, reveals the state of the house at that time. A man named James C. Willfong wrote about a trip he took to the area in 1954 with his daughter, Sue, and Dr. Henry Chandlee Forman, noted early historical archaeologist and architectural historian. They stopped first at Mr. Kirkpatrick-Howat’s house (the old Sellman house), and then accompanied Kirkpatrick-Howat to see the Java mansion. Willfong writes,

“Java was one of the strangest houses I had ever seen. Here was a two-story gambrel roof with a center entrance that would qualify it as Georgian. It was clearly old –really old – and many years earlier someone with very peculiar ideas had built bay windows on the second floor on the end which faced the river. To me they were awful, and I thought it was too bad… [they] were a later alteration…and they were a fair match of the original brick, but there was just enough contrast to make you wonder why somebody had bothered to do what had been done…[it seemed] what had probably been a 17th century mansion house had been reduced to a bit of early Victorian hodgepodge – an eyesore, too.” Willfong went on in the article to say that all of the window frames and glass were gone by this time, and the damage from the circa 1890 lightning strike was still visible in the one gable end. He states, “There was no settlement on either side of the slice, but there was a gaping hole, at least an inch wide, from the ground right up to the eaves.” Dr. Forman and Mr. Kirkpatrick-Howat agreed that the house was beyond repair, but “if the entire building were stuccoed if might be livable, but it would still be an eyesore.” Mr. Kirkpatrick-Howat stated that he and his father felt the same way and they decided to leave it as it was “as a kind of monument to some visionary 17th century styling and some 19th century misguided architectural taste – and lightning.”

In this article, Willfong also commented on the gigantic black walnut tree that was located just to the north of the ruins. This black walnut was, according to Mr. Kirkpatrick- Howat and the Maryland Department of Forests, the largest walnut tree in America. Even in 1954, the tree was in poor shape, and it finally died sometime in the early 21st century. This tree was such a local landmark, that then 86-year old Captain Harry Benning stated in a 1975 interview that this walnut tree could be seen “from way over on the Eastern bay, you could see that round hump on the horizon…I’m told the old schooner captains used it for a mark. It sure stood out on the horizon.”

Today, all that remains of the old Squirrel Neck mansion that Nicholas Maccubbin built around 1750 are two chimneys and a portion of the northern hyphen. As Lost Towns Project archaeologists have been working in their shadow for the past two years, bricks have fallen out of the chimneys on a number of occasions. For the time being, nature continues to take its toll on the old house. However, in early 2008, SERC, through the Smithsonian Land Trust (a non- profit organization), purchased all of the land owned by the Kirkpatrick-Howat family, and part of their preliminary cultural management plan for the property involves an effort to stabilize the ruins. They will be assisted in this endeavor by the Master’s thesis written by Lost Towns Project intern, January Ruck, who wrote “Reintegrating Public History & Environmental Education: Preservation and Interpretation of the Ruin at Java Plantation, Edgewater, Maryland” (2008). In this compelling document, Ms. Ruck provides an argument justifying the expenditure of financial and human resources regarding stabilizing the old mansion ruins. Her analysis, along with input from Project staff, will provide an initial framework for SERC as they plan for the future of the incredibly rich cultural heritage of their newly acquired lands.

18AN339 Previous Archaeology

In the Fall of 2006, Lost Towns Project archaeologists began Phase I/II testing at the Java site, and testing continued through January of 2007. A total of 31 shovel test pits (STPs) were excavated at 50 foot intervals in the portion of the site studied during this phase of investigation; an additional six STPs were excavated on the tenant house knoll to the east of the mansion ruins. No strong evidence for a domestic occupation was identified in the eastern portion of the site, but the light scatter of architectural materials may warrant further testing in the future in this area. The STP grid across all of site 18AN339 was based on the orientation of the house, with grid north lying 11 degrees west of true magnetic north (Cox et al. 2007b).

Artifacts spanning the entire historic occupation of the property were recovered from the 31 STPs excavated in the area surrounding the mansion ruins; however, distinct concentrations of kitchen and architectural artifacts were identified in the west, rear yard of the mansion. These concentrations not only revealed a backyard activity area related to the 17th, 18th and 19th century occupation of the site, but also suggested the potential for an activity area that predated the construction of the circa-1747 Maccubbin house (Cox et al. 2007b).

The locations of three excavations units were then selected (Figure 2). Unit 1 was placed at the base of the eastern façade of the mansion ruins, between the main block and the hyphen or passage leading to the northern wing. The dimensions of the unit measured approximately three and a half feet by five feet, as the interior one and a half feet of the unit was located under the standing brick foundation. Unit 1 revealed four episodes of construction and destruction spanning the occupation of site 18AN339. “Episode 1” represented the 20th century deposits that followed the 1890s rebuilding of the structure after it burned. “Episode 2” represented the 1890s reconstruction. “Episode 3” represented the original mid-18th century construction of the house. Finally, “Episode 4” represented the original ground surface that existed before the construction of the mansion. The strata in this episode may also be associated with the Sparrow occupation of the site in the late 17th and early 18th centuries (Cox et al. 2007b).

Unit 2 was placed to the south of the mansion ruins, on approximately the same line as the eastern façade to test for subsurface features related to a formerly extant southern passage (see Figure 2). Only the northern half of the five by five foot unit was excavated due to the

presence of dense brick rubble located immediately below the sod. A robbed foundation trench, called Feature 1, was identified subsurface in the western portion of the unit related to the southern hyphen that was most likely destroyed during the fire that burned most of the structure in the 1890s. The hyphen was most likely constructed of brick and did not have a basement, as does the corresponding hyphen on the northern end of the building (Cox et al. 2007b).

Unit 3 was placed to the rear and northwest of the Maccubbin house ruins, based on a concentration of artifacts recovered in the vicinity during the STP survey (see Figure 2). The unit measured five by five feet and produced several 17th and 18th century diagnostic artifacts. This included a significant assemblage of faunal remains and large quantity of kitchen and architectural-related objects. This suggested a highly trafficked area, perhaps a kitchen, initially utilized before the mid-18th century construction of the mansion. Excavation of Unit 3 was halted at a dense lens of fire-cracked ironstone, possibly indicating an intact structural feature associated with an earlier structure (Cox et al. 2007b).

2007 Excavations

A total of 25 units would eventually be excavated during the 2007 field season. The majority of these (23) were ultimately clustered in an excavation block located about 100 ft. west, or inland, of the mansion ruins (see Figure 2). While the testing planned for the 2007 season was limited to excavation of 10 to 15 units, the edge of an articulated brick feature was found in the corner of Unit 16 on one of the last field days of the season. Excavation continued beyond the planned field session so that the team could better understanding the nature of this feature, which ended up being a 12 ft. long brick hearth that once belonged to a 17th century impermanent structure. Only a small portion of this building has been uncovered as of June 2008, and none of the 22 identified features have been tested, thus all of the artifacts discussed below have been recovered from the plow zone. The following discussion of methods and findings will conclude with our preliminary conclusions and research questions for further study.

Figure 2: Site plan, with excavation block, numbered units outside of block, and mansion location, at 18AN339

The first step in this phase of excavations at the Java site was to conduct a magnetometer survey in the west, inland yard of the mansion ruins (Figure 3). A number of anomalies were initially detected about 100 ft. away from the mansion ruins. A second mag survey was later conducted in the east lawn of the site, and while a number of hits were detected, time constraints prohibited conducted additional excavations in this area in the 2007 season.

1140

1120

1100

1080

hyphen

1060 52590 52580 passage 52570

52560 1040 52550

52540 52530 52520 main block 1020 52510 52500 52490 52480

52470 1000 52460 52450 N1000,W1000 -1080 -1060 -1040 -1020 -1000 -980 -960

Scale in Feet

0 20 40 60 80 Figure 3: Results from the magnetometer survey from the west lawn at 18AN339

While the archaeologists waited for final processing of the west lawn survey data, two excavation units were laid out. Unit 4 was placed between the northern hyphen of the 18th century ruin, to the east and Unit 3 to the west, while Unit 5 was placed south of this, roughly in line with the location of the southern hyphen foundation (see Figure 2). It was hoped that placing these two units in similar orientation to the mansion ruins would provide insight to use variations between hyphens through a comparison of the artifact assemblages.

The two units had very different stratigraphy. Unit 4 consisted of three strata that overlay dense rock, while Unit 5 consisted of two strata that overlay a small artifact-bearing soil lens and subsoil. In Unit 4, Stratum 1 (10YR 4/3 brown sandy loam) represented the upper sod level and artifacts recovered (n=211) included a few wrought (n=3) and cut nails (n=3), two whiteware rim sherds, some utilitarian wares (undiagnostic salt glazed stoneware and coarse red earthenware), and a few burned glass and burned whiteware sherds. Stratum 2 (10YR 5/3 brown sandy loam with brick flecking) produced over 1,000 artifacts. Notable artifacts included one tin-glazed earthenware cup base, two creamware sherds, 17 whiteware sherds (including six that were burned), three ironstone sherds, a number of coarse earthenware and stoneware sherds, 45 wrought nails, 52 cut nails, and a number of glass vessel fragments. Stratum 3 (90% 10YR 5/3 brown silt with 10% 10YR 6/6 brownish yellow silty clay with frequent brick specks, moderate oyster, and moderate rocks) produced only 124 artifacts, including one lead-backed tin glazed earthenware sherd, one Rhenish stoneware sherd, one white salt-glazed stoneware sherd, and one 90 pearlware sherd. No whiteware or ironstone was recovered from Stratum 3. In addition, four wrought and four cut nails were recovered. In general, the artifacts recovered from Strat 3 had an earlier manufacture date than those recovered from the upper strata; however, the presence of cut nails suggests 19th century admixture. At the base of excavation, a dense collection of rocks were noted, comprising about 70% of the unit after the removal of Strat 3. No discernable pattern was identified, but the rock was primarily located in the northwestern and central portions of the unit. These rocks were similar to those noted in the base of Unit 3 to the northwest.

Unit 5 was the only unit excavated to the southwest of the mansion ruins (see Figure 2). Its stratigraphy consisted of two strata that overlay a third soil lens that was located in the southeast corner of the unit. Stratum 1 (10YR 5/3 brown sandy loam with occasional brick and

coal) produced a total of 152 artifacts, including five partial square nails, two pearlware sherds, two burned tin glazed earthenware sherds, and four coarse earthenware sherds. Stratum 2 (10YR 5/6 yellowish brown sandy clay with coal, brick, and occasional oyster shell) produced 222 artifacts, including one creamware and one pearlware sherd, one English brown salt-glazed stoneware sherd, ten coarse earthenware sherds, one pipe stem fragment, three hand wrought nails and four square nails, and a handful of vessel glass fragments. Stratum 3 (10YR 5/3 brown clay loam) was noted during excavation of Strata 1 and 2. It was concluded that Stratum 3 was an area of slightly higher artifact concentration at the base of the plow zone and was not a cultural feature. It was, however, excavated separately and only 69 artifacts were recovered, including a number of coal fragments, two cut nails, six flat glass fragments, one coarse earthenware sherd and one salt-glazed stoneware sherd. This stratum overlay sterile subsoil and excavation was halted.

Unit 4 produced 1,394 artifacts from its three strata, while Unit 5 only produced 443. This is consistent with the general trend noted during shovel testing in the western yard of the mansion; fewer artifacts were recovered from the southern portion than the northern portion. Figure 4 provides a general graphic overview of the assemblages. Note that brick has been excluded from this figure; it comprised a near identical percentage (approximately 50%) of artifacts recovered from each unit (Unit 4, n=718; Unit 5, n=214).

Figure 4: Partial comparison of assemblages from Units 4 and 5 at 18AN339

A higher percentage of architectural materials, including brick, mortar, and nails were recovered from Unit 4 (65.9% of the Unit 4 assemblage vs. 44.5% in Unit 5), but a near equal number of wrought (n=52) and cut nails (n=59) were collected from this unit. The ceramic assemblage of Unit 4 further suggests 19th century admixture of artifacts. Several sherds of a 19th century green-tinted hard-paste porcelain vessel with drain holes in the base (probably a colander), along with 19 whiteware and three ironstone sherds are most likely were used during the Contee occupation of Java Farm. However, tin-glazed earthenware sherds (n=11), including one lead-backed sherd, one Rhenish stoneware sherd, two English Brown stoneware sherds, and two white saltglaze stoneware sherd are possibly representative of the earlier Sparrow occupation. The archaeologists noted that the stones at the base of the unit were laying in an artifact-producing Stratum 4, but excavation was halted at the base of Stratum 3. Additional work should be undertaken in the vicinity of Unit 4 in the future to explore Stratum 4 and possibly clarify the usage of this portion of the site.

In general, the number, percentage, and functional variety of artifacts recovered from the southern portion of the site were less than that recovered to the north. Only coal was recovered in greater quantities in the southern yard. This suggests an absence of a domestic area that pre-

dates the 1747 construction of the mansion in this portion of the site. One anomaly was detected during the magnetometer survey to the west of Unit 5 that could warrant further analysis, but in general it appears the research potential of this portion of the site is limited.

Unit 8 was the final unit excavated that was not eventually part of a larger block of units (see Figure 2). This unit was placed in the western portion of site 18AN339 based on an anomaly noted in the vicinity during the magnetometry survey. Two strata were identified in this unit. Stratum 1 (10YR 5/3 brown sandy loam) produced few artifacts (n=140), and brick constituted about 87% (n=122) of the stratum assemblage. Three hand wrought nails, one small tin-glazed earthenware sherd, and one whiteware sherd were the only diagnostic artifacts. One thin iron wire, measuring nearly a foot in length, and a threaded iron pipe were also recovered from Stratum 1. These were possibly the cause of the magnetometer hit in this area. Stratum 2 (10YR 5/3 brown compact sandy clay loam with brick flecking) produced 285 artifacts, and 55% of this was brick (n=157). Notable artifacts recovered from this stratum included 24 hand wrought nails, two Rhenish stoneware sherds, one English stoneware sherd, three pearlware sherds, three whiteware sherds, a few red-bodied coarse earthenware sherds with black lead glaze, one honey-colored gun flint, and 12 olive bottle glass fragments. In addition to these, one possible obsidian flake was recovered from this strat. Stratum 3 (10YR 5/4 yellowish brown silt loam) was exposed at the base of excavations. A light scattering of brick and stone rubble was exposed, but no artifacts were noted during trowling. Due to the temporal mix of artifacts and the apparently sterile soil encountered at the base of the unit, excavation was halted and the unit was backfilled.

The remaining test units were ultimately clustered in two large blocks to the south of Unit 3. A total of 22 units were fully excavated in the block in 2007, and a number of interesting features were uncovered (Figure 5). All of the features were clustered in the northern block. Four units (Units 6, 10, 12, and 15) were clustered to the southeast and while no features were noted, ironstone rubble, similar to that seen at the base of Unit 4, was noted in varying degrees of density at the base of all of them. This ironstone was sporadically noted in limited densities throughout the remainder of the northern block. However, a relatively higher percentage of artifacts were noted in the southern block than the northern block.

In general, the northern portion of the block was excavated in two layers, while the southern block was excavated in three layers. This was done because nearly every unit in the northern block hit a feature at the base of stratum 2. Stratum 3 in the southern units produced few artifacts and included the ironstone rubble. These units were also slightly deeper than those to the north, terminating about 1.25 ft. below ground surface.

Soil colors and textures were generally quite similar throughout the blocks, containing varying densities of brick, shell, and rock rubble. Strat 1 was typically a 10YR 4/4 dark yellowish brown silty loam, while Strat 2 was typically a 10YR 5/4 yellowish brown silty clay loam. In the four units in the southern block, Strat 3 was mottled with more sand and contained 60% 10YR 7/4 very pale brown and 40% 10YR 5/4 yellowish brown sandy loam.

27 13 25

7 22 9 11 19 20 24

1418 17 26

16 23 28

21 10

Key: Grid Rock True Brick Shell North 12 15 0 10 Scale (in ft.) 6

Figure 5: Numbered excavation units within block at 18AN339 30

Features at 18AN339

A total of 22 features were uncovered in the excavation block located to the west of the mansion ruins; only Feature 1 was located outside of this block and was identified during the 2006 investigation (Figure 6). As stated above, none of the features have been excavated and few have been fully uncovered, so it is difficult to determine what is contemporaneous with what. A description and initial impressions of each feature follows (Table 1):

Table 1: Features Identified at 18AN339

Feature Associated Feature Description Number features

1 Robbed foundation trench in Unit 2; only feature identified in 2006 2 4, 5 The brick hearth 3 3a, 6 Large shell feature to west of hearth 3a Mortar-rich area at southern end of shell feature 4 2 Burned area inside of hearth; varying degrees of charring 5 2 Heavily burnt area inside of hearth, located at exterior end of firebox 6 3 Possible post hole located within large shell feature, to west of hearth 7 Post hole and mold, probably original post, located to northeast of hearth 8 Post hole and mold, probably repair post, located to northeast of hearth 9 Post hole and mold located to southeast of hearth 10 Small circular stain located just west of hearth, possible post mold, related to chimney construction? 11 Small circular stain located just west of hearth, possible post mold, related to chimney construction? 12 13 Circular stain located near southwest corner of hearth; oyster, brick fleck, and charcoal inclusions; surrounds feature 13 13 12 Circular stain located near southwest corner of hearth; infrequent brick and oyster shell, very infrequent charcoal inclusions; surrounded by feature 12 14 Dense brick layer, circular in shape, located off of northwest corner of hearth 15 Destruction rubble pile of predominately ironstone and sandstone; generally in line with hearth 16 Destruction rubble pile of predominately brick and daub rubble; partially overlays two post holes (Features 7 & 8) 17 18, 19, 20 Part of feature complex partially exposed in northern portion of excavation block; semi-circular, located furthest east in complex 18 17, 19, 20 Part of feature complex partially exposed in northern portion of excavation block; semi-circular, surrounds Feature 19 19 17, 18, 20 Part of feature complex partially exposed in northern portion of excavation block; semi-circular, surrounded by Feature 18 20 17, 18, 19 Part of feature complex partially exposed in northern portion of excavation block; located furthest west in complex 21 22 Moderately mortar-rich feature located at northern end of large shell feature; truncated by Feature 22 22 21 Heavily mortar-rich feature located at northern end of large shell feature; appears to have partially truncated Feature 21

21 22 20

19 18 17 8 7 16 3 3 14 15

6 4 3 5

Key: Grid Rock True 10 Brick 3a 11 Shell North 0 10

Scale (in ft.) 12 2 9 13 Figure 6: Site plan showing numbered features in the excavation block at 18AN339 32

Feature 1

Feature 1 was identified at the base of Stratum 2 in Unit 2, excavated in 2006 (Cox et al. 2007b). It represents the western portion of the unit and was defined as a robbed foundation trench associated with the no longer standing southern hyphen of the mansion. This was a rubble filled trench, measuring 2.3 ft. wide and filled with large architectural rubble of mortar, stone, and brick, along with copious window glass, nails, and occasional ceramics and vessel glass. This rubble is likely associated with the destruction of this wing of the building during the 1890s fire.

Features 2, 4, & 5

Features 2, 4, and 5 represent a brick hearth and the burned area inside the firebox (see Figure 6). Feature 2 is the brick hearth itself, which measures 12 ft. long on the exterior and 8 ft. long on the interior. The width of the hearth is by 5 ft. long on the exterior by 3 ft. long on the interior. The hearth is better preserved and the bricks are more fully articulated in the southern portion; the northern bricks are loose and crumbly, and it is difficult to determine the precise outline of the feature. Features 4 and 5 represent portions of the burned area inside the hearth (Figure 7). Feature 4 is a 10YR 3/4 dark yellowish brown silty clay loam with brick and charcoal flecking and occasional mortar, rocks and oyster shell. Feature 5 is a heavily burnt area further to the exterior of the firebox. It consists of a 10YR 4/3 brown silty clay loam with larger bricks and brick flecking, heavy charcoal, and occasional mortar and shells.

Figure 7: Features 2 (the brick hearth), 4, and 5 (burned areas inside hearth) at 18AN339; Feature 3 (the shell midden) visible to west. This single-sided hearth and burned area was at one time most likely located on the gable end of an earthfast building. No distinctive foundation was identified in the surrounding area, and a number of post holes with molds were noted to the east of the hearth (Features 7-9). This hearth is most likely related to these posts.

Features 3 & 3a

Feature 3 represents a large shell-rich feature located to the west of the brick hearth (Figure 8). The majority of the soil within the feature is comprised of a 10YR 5/3 brown silty loam with frequent whole oysters and oyster fragments, and frequent daub and brick fragments. The exposed portion of this feature measures approximately 12 ft. long by 8 ft. wide, but much of it has yet to be uncovered. A number of artifacts, including faunal remains and an iron spoon handle, were noted in the top layer of the feature; the spoon handle was taken back to the lab for curation. Feature 3a seems to be a portion of the greater Feature 3 with a similar concentration of oyster shell, but more mortar was noted in this section of the feature, along with higher concentrations of brick and charcoal.

Figure 8: Feature 3 (the shell feature) shown in relative position to Feature 2 (the brick hearth, partially exposed in background) and the Squirrel Neck/Java mansion ruins at 18AN339 Features 6, 7, 8, & 9

Features 6, 7, 8, and 9 represent post holes and their associated molds situated to the east and west of the brick hearth (see Figure 6). Feature 6 is actually located within Feature 3, the shell midden, and represents a slightly darker stain filled with a higher percentage of brick

rubble. Because the stain is located within the larger shell midden, its shape is difficult to discern. It is the only post hole noted to the west of, or behind, the brick hearth.

Features 7, 8, and 9 are all located to the east of the hearth. Features 7 and 8 represent a post and repair post and their associated molds situated to the northeast of the hearth. It appears Feature 7 was the original post and Feature 8 was the later repair post. This was surmised because the post mold in Feature 7 contains few artifacts, while the mold of Feature 8 is chock full of rocks and brick rubble, suggesting more artifacts were in the area when the post was removed. The hole of Feature 7 contains a 10YR 4/3 brown loam mottled with 25% 10YR 5/6 yellowish brown clay, while the mold of Feature 7 contains a 10YR 3/3 dark brown clay loam. The post hole contains a 10YR 4/3 brown silty loam mottled with 50% 10YR 6/6 brownish yellow clay, while the post mold of Feature 8 contains a 10YR 3/3 dark brown silty loam with large rocks and brick bats (Figure 9).

Figure 9: Feature 7 (to right) and Feature 8 (to left) at 18AN339, facing south; Feature 16 (brick rubble pile) partially visible to top of frame

Feature 9 is located to the southeast of the hearth and represents another post hole and associated post mold (Figure 10). It appears that the post hole is not fully uncovered and extends south and west into unexcavated areas. No artifacts were noted in the hole or mold, suggesting it is probably first generation. The post mold contains a 10YR 3/3 dark brown clay, while the hole contains a 10YR 3/6 dark yellowish brown clay mottled with a 10YR 4/6 dark yellowish brown clay. The post mold here in Feature 9 seems very small compared to the large size of the hole.

Figure 10: Feature 9 (the post hole and mold) at 18AN339, visible in lower right It should be noted here that Features 7 and 9 are 16 ft. apart, while Features 6 and 7 are also 16 ft. apart. This measurement was a typical size for the width of an earthfast building in the 17th century.

Features 10 & 11

Features 10 and 11 appear to be small post molds without associated holes situated just behind (to the west) of the southern edge of the brick hearth (see Figure 6). These small stains consist primarily of a 10YR 3/3 dark brown sandy loam with infrequent brick flecking, mottled

with 10% 10YR 4/6 dark yellowish brown sandy clay. Their proximity to the hearth suggests some sort of relation with the chimney. It is possible that these could have been small support posts used during construction of the hearth or posts to support a leaning wattle and daub chimney.

Features 12 & 13

Features 12 and 13 represent partially-exposed stains located behind (west of) the southwest exterior corner of the hearth (see Figure 6). These stains appear to be circular in shape and Feature 13 is located within Feature 12. Judging by the exposed curvature of the features, they would probably be quite large, possibly measuring 5 ft. or greater in diameter. Feature 12, the exterior feature, contains a 10YR 3/3 dark brown sandy clay with inclusions of oyster shell, brick flecks, and charcoal mottled with 35% 10YR 4/6 dark yellowish brown sandy clay. Feature 13, the interior circular feature, consists of a 10YR 3/3 dark brown sandy loam with 20% 10YR 4/6 dark yellowish brown sandy loam with inclusions of infrequent brick and oyster shell and very infrequent charcoal. There is a possibility that these features could be structural and related to the earthfast building, but too little of it is exposed to be certain.

Feature 14

Feature 14 represents a dense brick layer containing several brick bats and crumbled brick (see Figure 6). The soil between the bricks contains a 10YR 5/3 brown sandy loam with a few larger ironstone rocks. This feature seems to extend in a circular shape off of the northwest corner of the hearth. This portion of the hearth is very loosely articulated and poorly defined, and the feature seems to be a continuation of the crumbed bricks. There is a possibility that Feature 14 could be structural in some way considering its proximity to the hearth, or could be a layer related to the destruction of the building.

Features 15 and 16

Features 15 and 16 represent destruction rubble piles located to the east and northeast of the hearth (see Figure 6). Feature 15 is located just east of the northern wing of the hearth and

contains over 90% ironstone and sandstone rubble. Some brick flecks and bits of oyster shell were noted in a matrix of a 10YR 3/3 dark brown silty loam (Figure 11).

Figure 11: Features 15 (the rock rubble pile, shown to top), 16 (the brick rubble pile, center), 7 and 8 (the two post holes and molds, shown to right) at 18AN339, facing southwest Feature 16 is located to the north of Feature 15 and consists of more brick bats, brick rubble, and daub with a minimal amount of stone rubble. This feature overlays the two side-by- side post holes and molds, called Features 7 and 8. The soil in between the bricks is a 10YR 5/3 brown silty loam. Most of the brick here is fragmented and soft, similar to what was noted in the northern edge of the hearth itself.

Features 17, 18, 19, & 20

Features 17, 18, 19, and 20 are a series of what appear to be related stains in the very northern portion of the exposed excavation block (see Figure 6). The features are generally circular in shape, and none of them are fully exposed. Moving from west to east, Feature 17 consists of a 10YR 4/4 dark yellowish brown silty clay mottled with 50% 10YR 6/6 brownish

yellow silt and a number of ironstone rocks. This feature appears to have been cut at a later date by Feature 18, located just to the west. This semi-circular feature measures approximately 6 ft. in diameter and consists of 10YR 5/3 brown silty loam with 50% 10YR 4/4 dark yellowish brown with inclusions of infrequent brick bits, mortar, and bone.

Lying in the center in of Feature 18 are Features 19 and 20. Feature 19 appears to have roughly the same shape as Feature 18, and its exposed diameter measures approximately 3 ft. This feature consists of a 10YR 4/4 dark yellowish brown silt mottled with 40% 10YR 5/3 brown silt and frequent shell, mortar, and brick flecks. Feature 19 may have been excavated at a later time than Feature 20, as it appears to have truncated it. Feature 20 is located just to the west of Feature 19 and consists of a 10YR 5/6 yellowish brown silt with inclusions of occasional brick, mortar, and shell.

The nature, date, or meaning of this feature complex is unclear. Some have speculated that this could represent a large structural post hole for the earthfast building, as it is situated just to the north of the brick hearth. A large building would require a large post hole and mold, and the complex is in a location expected of an end post. It could also represent a portion of a second, later structure built in the area. Further excavations may reveal the nature of this complex.

Features 21 and 22

Features 21 and 22 represent stains filled with mortar that are located just to the north of Feature 3, the large oyster-rich feature (see Figure 6). Feature 21 is located to the west of Feature 22 and consists of a 10YR 4/4 dark yellowish brown silty loam with inclusions of mortar chunks and brick bits. It appears Feature 22 was partially excavated into Feature 21, and it consists of a 10YR 4/4 dark yellowish brown silt loam with heavy mortar and brick inclusions. The use or function of these features is unclear, but their proximity to the edge of the shell feature, Feature 3, suggests some sort of relationship (Figure 12).

Figure 12: Features 21 and 22 (shown left and center) in proximity to Feature 3 (the shell feature) at 18AN339

Feature Summary

To summarize, a total of 21 features were fully or partially exposed at 18AN339 during the 2007 field season; only Feature 1 was identified in 2006 to the south of the mansion ruins. The two largest features identified this season were a single-sided brick hearth (Feature 2) that once provided heat and light for an earthfast building, and an oyster shell-rich feature (Feature 3) situated to the west of the hearth. Assuming these features are contemporaneous, the shell feature would probably be situated just outside the gable end of the building.

The hearth is of a decent size for a 17th century earthfast building, measuring 12 ft. long on the outside by 5 ft. wide. Varying stages of burning were noted inside the hearth, as would be expected in a large hearth of this size; it is highly doubtful that the full firebox would be filled with burning timber throughout the year. The hearth is in varying stages of degradation. The bricks are more articulated and intact to the south and more crumbled and friable to the north.

Oddly enough, the northern end was more deeply buried, while the southern end was closer to the existing ground surface. It is possible that at some point since the destruction of the building, the plow clipped the northern end of the hearth causing it to become partially destroyed. Alternatively, later use of the building or its very destruction could have negatively impacted the hearth.

At least three post holes with molds were also identified this field season (Features 7, 8, and 9). These posts almost certainly mark two walls of the same earthfast building that possessed the chimney base. The posts are situated 16 ft. apart, which is a typical measurement for post-in-ground buildings of the 17th century. A fourth feature, located behind the hearth and contained within the large shell-rich feature, may represent another post hole (Feature 6). This feature is marked by slightly darker soil than the surrounding shell-rich matrix and contains more brick bats. Coincidentally, this feature is located 16 ft. from one of the other northern post holes. Having said that, it is difficult to imagine what the building might have looked like if it continued 5 ft. behind a single-sided chimney.

Two small stains located just behind the chimney base on its southern side may represent the locations of small posts utilized during construction of the chimney, or possibly to support a leaning wattle and daub chimney (Features 10 and 11). It seems most likely that the chimney itself was constructed of wattle and daub rather than brick, judging by the relative paucity of brick in the vicinity. There is, however, a brick and daub-rich feature situated east of the hearth, probably representing destruction rubble (Feature 16). A second rubble layer, consisting primarily of ironstone and sandstone with limited brick and daub, is located just to the east of the hearth (Feature 15).

A number of other features and feature complexes may also represent structural posts in the vicinity. A mid-sized semi-circular feature is located adjacent to the northwestern corner of the hearth, consisting mainly of friable brick rubble (Feature 14). A feature complex is located directly to the north of the hearth, and this might represent an incredibly large post hole and mold (Features 17-20). Finally, a semi-circular feature complex located off the southwestern corner of the hearth might also be a partially exposed large post hole and mold (Features 12 and 13).

It should also be mentioned that just because areas between these described features appear to be blank on the site map (see Figures 5 and 6) does not mean these are areas of culturally sterile subsoil. It simply means that there are no obvious features in that area, and as is suggested on the site map, bits of brick, mortar, oyster shell, and charcoal occur in varying amounts throughout the base of excavation. Only further digging will reveal what is contemporaneous with what and where some of these features actually begin and end.

Artifact Analysis

For ease of discussion, the following artifact analysis places the collection into generalized artifact categories, including kitchen-related artifacts, architectural-related artifacts, arms-related materials, horse furniture, and personal materials. This technique loosely follows the functional groups outlined by South (1977). Each group was distributed spatially using Surfer software, and a few of these are presented in the following discussion.

The artifact assemblage from Java was recovered solely from the plow zone, and no intact features have yet been excavated. Therefore, statistical analyses like mean ceramic dates, Binford pipe stem examination, minimum ceramic vessel count, or a highly detailed faunal analysis were not conducted because the artifacts came from mixed contexts. The entire assemblage will be curated at the Anne Arundel County archaeology laboratory in Londontown, Maryland, and conservation is ongoing.

A total of 34,141 artifacts have been recovered from 18AN339 over the course of two field seasons; the 2007 field season alone produced 30,153. This material culture demonstrates a wide cross-section of the rich cultural heritage of the site. Prehistoric materials were recovered alongside 20th century artifacts, telling of the long occupation of the area. All of the artifacts recovered during this field season came from the plow zone, which due to erosion, is of varying depths throughout the site. It is difficult to make assumptions about the lifeways of one family over another with such a mixed assemblage. However, with careful examination, we can draw some conclusions.

A very basic breakdown of the assemblage into generalized categories reveals that the highest percentage (32%) of artifacts were in the masonry group, which includes brick, mortar, 43

plaster, daub, and tiles (Figure 13). However, brick will be excluded from the remainder of this analysis due to inconsistencies in field collection methods. Over 10,000 bricks were recovered from Java, but it is not clear how many more were discarded in the field, particularly when we screened the brick rubble-heavy soils that overlay the brick hearth. Therefore, when the brick is excluded from this basic breakdown, the masonry category now only represents 4% of the assemblage (n=944). This difference is displayed visually in Figures 13 and 14.

Figure 13: Total artifacts recovered from 18AN339

Figure 14: Total artifacts without brick at 18AN339

As Figure 14 shows, a nearly even percentage of faunal remains (including bones, teeth, shells, and scales; n=6,929) and metal materials (n=6,702) were recovered from across the entire site. Ceramics, including tobacco pipes, make up 17% of the assemblage (n=4,043). Lithics (including debitage, fire cracked rock, gun flints, and slag), glass (including window glass, table glass, and beads), and floral materials (charcoal, seeds, and nuts) round out the remaining 22% of the assemblage.

The highest concentration of artifacts were recovered from the excavated areas located to the south and east of the hearth, probably once situated inside the earthfast building.

Kitchen-Related Artifacts

The kitchen-related artifacts recovered from Java include ceramics, table and bottle glass, faunal remains, and cutlery. A great deal was found inside or just to the south of the footprint of the 17th century earthfast building; this is also where the highest concentration of all artifacts were found. Many of these artifacts date to the Sparrow period of occupation (circa 1650-1748), but all were recovered from mixed contexts. However, with careful analysis we can make a few conclusions about the lifeways of the various occupants of the site.

Ceramics

Removing the pipes from the ceramic assemblage leaves a total of 2,086 fragments of pottery recovered from Java. These ceramics span hundreds of years of occupation of the site, from the 28 pieces of low-fired prehistoric pottery to the 247 pieces of 19th and 20th century whiteware and much more in between. A total of 79% of the ceramics are earthenware (n=1657), 15% are stonewares (n=311), and 6% are porcelains (n=118). Figure 15 presents this generalized breakdown of the entire ceramic assemblage.

Figure 15: Total ceramics (minus pipes) recovered from 18AN339

The earthenware assemblage is highly varied, but tin-glazed earthenwares represent over 50% of these ceramics. Refined white earthenwares (creamware, pearlware, whiteware, ironstone, and unidentified white earthenware) comprise about 33% of the assemblage. Figure 16 presents the total earthenwares from the site below.

Figure 16: Total earthenwares recovered from 18AN339

Consistent with the total artifact distribution, the majority of the tin-gazed earthenwares were recovered from south and east of the hearth. Considering the proximity and location of these 17th and 18th century artifacts in relation to the contemporaneous building, it seems highly likely that these cups, saucers, and dishes were utilized by the Sparrow family when they occupied this building. This also suggests that there has been little horizontal shifting of the artifacts over the centuries.

Many of these ceramics were hand painted with blue decoration, but some were more colorful, containing orange, black, or purple geometric designs (Figure 17). Most of the sherds, having been recovered from the plow zone, are too small and broken to decipher what image the artist was attempting to portray.

Figure 17: Example of tin-glazed earthenware sherds recovered from main excavation block at 18AN339

A total of six polychrome lead-backed tin-glazed earthenware sherds were also recovered from the vicinity. The lead-backing on these ceramics suggests they may have been used for wall hangings or decorative purposes, rather than for dining.

Other ceramics recovered that date to the Sparrow occupation of the earthfast building include one sherd of North Italian Slipware that was manufactured from roughly 1610-1675, seven sherds of North Devon coarse earthenware (circa 1680-1720), one Buckley-like earthenware sherd (circa 1720-1775), and 13 manganese mottled earthenware sherds (circa 1680- 1750). All of these ceramics were recovered from the immediate vicinity of the earthfast building and none were recovered from any of the shovel test pits or the three units excavated closer to the mansion. This strongly suggests that we are excavating in the locus of the thus- identified 17th century Sparrow-period occupation.

The refined white earthenwares recovered (creamware, pearlware, whiteware, ironstone) largely post-date the Sparrow period of occupation at the site. English creamware is not seen on American sites until the 1760s, almost 15 years after the Maccubbin acquisition of the property. However, these artifacts do speak to the long Maccubbin and subsequent Contee occupation of the property. A number of creamware dishes, blue shell-edged pearlware tablewares, and transfer-printed whiteware sherds were recovered during the excavations. The presence of these later 18th and 19th century artifacts is telling of the ways in which this portion of the site was utilized after the construction of the brick Georgian mansion in the 1750s. It is possible that after the old earthfast building was torn down, this part of the site was used only for refuse disposal. Another possibility is that the old building was reused for storage, industry, or housing of tenants or slaves who had little material wealth. Future excavations are necessary to determine if any of the exposed features date to the 18th or 19th century and to determine the spatial layout of the site determined by these later owners of the land.

The stoneware assemblage, presented in Figure 18, shows Rhenish as the highest single percentage recovered (30%). Many of the Rhenish stonewares from Java are beautifully decorated, some with cobalt and manganese ornamentation that dates to the last quarter of the 17th century. A number of the sherds are also incised and sprig molded. Only eight of the sherds were identified as Rhenish Brown, which often served more utilitarian purposes than its highly decorated cousins.

Figure 18: Total stoneware recovered from 18AN339

Many of the earlier English brown stoneware fragments appear to be mugs or tankards either slip dipped or with a band of iron oxide at the rim. One of these fragments contains the mark “AR”, indicating it was manufactured during the reign of Queen Anne (1702-1714). A few mid-18th century white-bodied stoneware dish fragments were also recovered from the excavation block, although these post-date the Sparrow occupation of the site (Figure 19).

Additionally, 18 sherds of domestically-produced gray-bodied stonewares were recovered, accounting for only about 6% of the total stoneware assemblage. These blue- decorated wares were not produced in this country until about 1725, but were very popular and widely available during the later 18th century and through the 19th century. It is not clear if these vessels were used during the Sparrow occupation of the land.

Figure 19: Selection of Rhenish stoneware, English brown stoneware, refined white earthenware, Rockingham, and North Italian slipware recovered from excavation block at 18AN339

Only 118 sherds of porcelain were recovered from both seasons of fieldwork at 18AN339. This accounts for 6% of the entire ceramic assemblage. All of the sherds are hard pasted, and a few have hand painted blue decoration. A number of porcelain sherds are decorated with an orange sunburst decal pattern (1890-present). These sherds were recovered from a number of units in the northwest corner of the excavation block, and speak to the late 19th and early 20th century occupation of the site (Figure 20). It should be noted that these sherds were recovered from strata 2 and 3. This makes it clear that these soil layers are not intact, but have the same level of temporal admixture of artifacts seen in stratum 1.

Figure 20: Porcelain sherds with overglaze decal sunburst decoration recovered from Strata 2 and 3 at 18AN339

A number of fragments of a greenish colored porcelain vessel were recovered from Unit 4, located to the west of the main excavation block (Figure 21). A red, hand-painted overglazed floral decoration is located on the interior rim of the vessel. Most of these fragments mended to form what appeared to be a small, oval-shaped bowl with drain holes in the bottom, much like colander. A number of ceramics experts looked at this vessel and concluded that it is most likely a late-19th century English porcelain soap dish, but they were unable to assign a specific maker or date. This unit did produce proportionately more 18th and 19th century materials than did the excavation block, which is not surprising, considering its proximity to the mansion ruins.

Figure 21: Nineteenth century soap dish from 18AN339

Vessel Glass

A total of 1,433 vessel glass fragments were recovered from Java. Table 2 below presents a breakdown of this portion of the assemblage.

Table 2: Total Vessel Glass Recovered from 18AN339

Type Number Percentage Vial 45 3% Table 49 4% Unidentified 477 33% Bottle 862 60% TOTAL 1433 100%

Of those fragments that could be identified, over 60% were bottles. And of these 862 sherds, over 81% were olive green wine bottle sherds (n=701). The table glass assemblage consists of over 78% clear drinking glasses (n=38), along with a few cobalt and yellow unidentified forms.

Faunal Remains

Nearly 7,000 bones, teeth, scales, and shells were recovered from 18AN339. Figure 22 presents the breakdown of this assemblage graphically.

Figure 22: Total faunal remains recovered from 18AN339

This assemblage reflects the long occupation of the hilltop, although the overwhelming majority was recovered from the excavation block. Relatively few faunal remains were noted in the units and shovel tests excavated closer to the mansion, possibly suggesting we have not found the kitchen trash area utilized during the main period of occupation of the mansion house. Within the main excavation block itself, the southern portion produced the most faunal remains, consistent with the distribution of total artifacts. There was probably a lot of butchering taking place next to the firebox, not to mention the casual discarding of bones once the meat had been consumed. Appropriately, many of the bones from this area had butcher or saw marks on them (Figure 23).

Figure 23: Two mammal bone fragments with butcher marks recovered from the excavation block area at 18AN339

Of the bones that could be identified, the overwhelming majority were mammals (93%). About 4% of the bones were avian and less than 1% were piscine. All of the 665 teeth and tooth fragments recovered were from mammals, many from pigs or cows.

Oyster shells make up the overwhelming majority (99%, or n=1417) of the shells recovered at the site, but it should be noted that in general, only those with “female” hinges, or those that were relatively whole, were taken back to the lab for processing. Fragments of shells were often discarded in the field. The remaining shells recovered from the site included four clam shells, one snail shell, and two egg shells.

Cutlery

A total of seven knife fragments, two table fork fragments, one iron spoon bowl fragment, one possible pewter spoon handle, and one iron spoon bowl represent the entire cutlery assemblage recovered from Java. All of these were recovered from the main excavation block or the immediately surrounding areas (Figure 24).

Figure 24: Iron utensils recovered from excavation block area at 18AN339 (l-r: knife, three- tined fork, two-tined fork, spoon bowl)

Nearly all the cutlery was made of iron or an iron alloy. This is most likely more reflective of what has been preserved in the ground, rather than the actual preferences of the Sparrow family. The presence of the two forks around the remains of a circa-1670 house is striking, as forks are rarely seen on archaeological sites in the Chesapeake until the very end of the 17th century. However, it should be stated that the forks were not recovered from intact contexts and could have been disposed of in the 18th century, when they were more widely available and accepted. One of these forks was two-tined, while the other was three-tined;

generally speaking, two-tined forks were manufactured as early as the early 17th century, while three-tined forks are more of an 18th century invention.

Three knife fragments recovered mend to form a whole knife from Unit 15, located south of the building footprint. These fragments, along with a second whole knife blade recovered from Unit 10, further the suggestion that this portion of the site was used as a refuse disposal area.

The two spoon fragments were also recovered from outside of the building footprint. One simple, undecorated iron spoon handle was recovered from within the large shell feature, Feature 3, located behind the brick hearth. And one narrow, drawn out spoon bowl was recovered from Unit 28, located along the southern edge of the building. The shape and material of this bowl suggest it dates to the early 18th century, still within the Sparrow period of occupation.

Architectural Materials

Thousands of architectural artifacts were recovered from Java, including nails, brick, mortar, daub, window leads, and flat glass. Like all of the objects to which a date can be assigned, these artifacts span the entire historic period occupation. Again, no features have been excavated at the site, and all of the artifacts were recovered from mixed contexts.

Nails

A total of 5,318 nails were recovered from the site. Figure 25 shows the breakdown of the nail assemblage graphically.

Figure 25: Total nails recovered from 18AN339

The wide date ranges of these artifacts suggest they were left here during the Sparrow, Maccubbin, Contee, and later 20th century occupations of the site. While this past season of excavations proved there was a 17th century presence on the hilltop, we have yet to determine how long this earthfast building stood, or how this part of the plantation was utilized after the construction of the circa-1750 mansion. There are no drawings of the mansion until the 1880s, when an outbuilding is depicted to the northeast of the main house (see Cox et al. 2007b). The presence of cut and wire nails recovered from the excavation block this past season could be a result of this, or another outbuilding in the vicinity.

Over 61% of the nails recovered at the site were hand wrought (n=3248). Relatively few wrought nails were recovered from outside of the excavation block. Within the block, the highest percentage of these nails was recovered from the units to the east of the hearth, or within the footprint of the earthfast, wooden building (see Figure 38). The presence of thousands of hand wrought nails from the vicinity strongly suggests that the building was repaired multiple times during its lifetime. The repair post (Feature 8) identified to the northeast of the hearth further endorses this theory (see Figure 6).

A total of 423 machine-cut nails were recovered from Java, representing about 8% of the nail assemblage. The majority of the cut nails were recovered from the units situated to the east 58

of the brick hearth, on the mansion-side of the hill. A simple distribution shows that this would put many of the nails around the footprint of the building, but this begs the question, how long was this building utilized? Assuming this is one of the same buildings that Thomas Sparrow referred to in his 1675 will, the earthfast building would be nearly 150 years old by the time cut nails were becoming widely available (circa 1820). It is not out of the question that the structure could have been adaptively reused for over a century, but it does seem somewhat unlikely. With luck, future excavations into some of the features in the area will answer this question.

One test unit, Unit 27, did produce a strikingly high number of cut nails (n=41) relative to the units around it. Unit 27 (see Figure 5) is situated to the northwest of the hearth. Many of these nails are identical to one another, suggesting that this may represent a single episode of deposition. Could this be the spot where a box or container of nails was accidentally dropped sometime during the 19th century Contee occupation? It seems very possible.

About 27% of the nails from the site were cataloged as “square.” In other words, the laboratory professionals were unable to determine if the nails were hand wrought or machine cut. This descriptor does state that they were able to tell with certainty that these nails were not wire. The 197 nails called “unidentified” were degraded to the point that they were not able to be put into any definitive category.

Only 1% of the nail assemblage (n=39) was wire, or modern, nails that became widely available after about 1890. It should be noted that about half of these nails came from Strat 1, while the other half came from Strat 2. As an example, Unit 17, located just inside the 17th century brick hearth, produced two wire nails from strat 2. This strat also produced 117 hand wrought nails, Rhenish stoneware, blue decorated tin-glazed earthenware, and a window lead. This unit demonstrates that no intact stratigraphy has been encountered in the large block excavated at 18AN339.

Window Leads

A total of 34 window leads were recovered from Java. All of these were recovered from the excavation block area; none came from the shovel tests or the units near the mansion ruins. But as with all of the dateable materials, the leads were recovered from mixed contexts in Strats

1, 2, and 3. However, most of the leads were recovered from units to the east of the brick hearth, many along the edges of the footprint of the building, demonstrating that much of this early material has not shifted significantly along the horizontal plane.

The presence of these leads demonstrate that the 17th century building that stood here contained glazed casement windows, which is not surprising considering the wealth of the Sparrow family. Even if this building is not one of the two houses Thomas Sparrow II referred to in his 1675 will, it would be surprising if a house of the time period did not contain at least a few casement windows.

Two of the leads recovered from Unit 23 were marked with initials and a date. Unit 23 was also the location of a post hole and mold (Feature 9), situated just off the southeast edge of the brick hearth (see Figures 5 and 6). When unfolded, one lead was marked with the initials “WM” surrounded by stars, while the second was marked with the date, “1671”, also surrounded by stars (Figure 26). The similar star design on the leads suggests that these leads may have come from the same casement window and that they were certainly made by the same maker.

Figure 26: Window leads recovered from Unit 23 at 18AN339 with makers marks “*WM*” (l) and “*1671*” (r)

The “*WM*1671*” mark is fairly ubiquitous in terms of marked window leads recovered from temporally similar sites. Two identical leads were recovered from the St. John’s site and the van Sweringen site in St. Mary’s City, Maryland. In fact, the initials “WM” have been seen on leads recovered from sites ranging from Jamestown, Gloucester County, Virginia, New Jersey, Ontario, Canada, and London, England with corresponding dates ranging from 1671 to 1687 (Hanna et al. 1992). This strongly suggests these leads were manufactured in England and speak to the vast trading patterns entrenched by the late-17th century.

Flat Glass

A total of 582 pieces of flat glass were recovered from the excavations at Java, but of these, only about 25% (n=145) was definitively called window glass. Most of this window glass was aqua or clear in color, which could have been found in an earthfast building of the 17th century.

Much of the flat glass was recovered from inside the footprint of the post-in-ground building, but the highest counts were recovered from Unit 1 (n=84) and Unit 2 (n=47), both excavated in the 2006 field season. These units were in the immediate vicinity on the brick mansion house which was occupied from circa 1750 through the 1920s. Considering the increasing availably of window glass through the 18th and 19th centuries and the numerous recorded alterations that took place on the mansion, it is of little surprise that proportionately more glass would be located in this area.

Brick, Mortar, Daub

Considering the long period of occupation at the site and the construction of at least two houses on top of the hill, it is little surprise that we recovered thousands of fragments of brick, mortar, plaster, and daub from our excavations. As mentioned above, brick was collected inconsistently, particularly when the soils that overlay the hearth were screened. These soils were rich with brick rubble, and the decision was made to only collect large fragments. Regardless, 10,014 red brick fragments were recovered from the site. Very few of these were complete bricks, but a number of them exhibited partial glazing. Glazed bricks were used as decorative elements in the pilasters of the circa-1750 brick mansion.

A full quarter of the mortar recovered from the site came from Unit 1, excavated along the west wall of the mansion ruins (n=175; 20.023 oz.). Most of the remainder of the mortar came from the units excavated behind, or to the east of, the brick hearth. Not surprisingly, the highest percentages in this area were noted in the units that overlay the northern and southern mortar-rich areas of the shell feature (Feature 3; Units 7, 14, and 27).

Daub, due to its resemblance to brick in the field, was probably also collected inconsistently. However, 182 daub fragments were recovered from the site, the majority coming from the units that surround the brick hearth. The presence of this daub in the area strongly suggests that the chimney was constructed of wattle and daub at some point during the life of the earthfast building.

Arms-Related Materials

Typical of 17th and 18th century sites in the Chesapeake, a number of firearm parts and firearm-related items were recovered from Java. This included gun flints used in flintlock rifles, lead shot, and various gun parts including an escutcheon, a frizzen, and a trigger guard. It is difficult to say if more iron gun parts are part of the Java collection, considering much of the unidentified metal is quite rusted. Conservation is ongoing, and many of these objects will be re- analyzed to attempt to determine what purpose they once served.

A total of 133 pieces of flint or chert were recovered from the site, but of these only 25 were definitively called gun flints. It is certainly possible that some of these unidentified lithics were brought to the New World as ship ballast and knapped by members of the Sparrow family. It is also possible that they were purchased in their pre-manufactured state and were later broken and discarded. Some could also certainly be domestic chert, utilized by Native Americans. The recovered flints varied in color, but 40 were honey-colored, traditionally thought of as French flint, and 75 were called black, gray, or smoky, suggesting they could have originated in England.

Only seven whole, unfired lead shots were recovered from the site. These ranged in size from one small .14 caliber birdshot, to three .38 caliber balls, to one .63 caliber ball. A number of possibly fired shots (spherical lead objects that resembled shot) were also recovered from

excavation block and the immediately surrounding areas. While these objects would have been available in stores in the 17th and 18th centuries, it was quite typical that a family would have made their own shot at home.

Three gun parts were definitively identified by Lost Towns Project laboratory staff. A beautifully decorated coppery alloy wrist escutcheon was recovered from Unit 12, situated just to the south of the footprint of the earthfast building (Figure 27). This decorative musket element would have been prominently located on the top of the weapon, and features the profile of a knight or conquistador. Three holes pierce the object for attachment to the gun. The motif suggests this escutcheon most likely dates to the late 17th century.

Figure 27: Copper alloy wrist escutcheon with knight motif from 18AN339

In addition to this escutcheon, an iron frizzen was recovered from Unit 15, just next to Unit 12. This frizzen was the part of the weapon utilized to strike the gun flint and cause a spark. Additionally, an iron trigger guard was recovered from Unit 25, located along the northern wall of the earthfast building .

Horse Furniture

A few pieces of horse furniture were recovered from the site. Two snaffle bits were recovered from the exterior of the earthfast house footprint, one in Unit 7 and the other in Unit 15. Snaffle bits are generally the most common type of horse bit, and the more intact one recovered from Java consists of a portion of a bit mouthpiece with an attached rind. The bit was designed so the reigns could have direct contact with the horse’s mouth, thereby giving the rider more control in steering the animal. This particular snaffle bit resembles one depicted in A Guide to Artifacts of Colonial America, discarded circa 1730 (Noel Hume 1969:241) (Figure 28).

Figure 28: Conserved 18th century snaffle bit from 18AN339

Additionally, one horseshoe fragment was recovered from the artifact-rich Unit 28, located just outside the footprint of the earthfast building.

Personal Materials

The category of personal materials is very general and somewhat misleading, as it could probably be argued that many of the artifacts discussed in this narrative are personal in nature. However, for the purposes of this report, items such as tobacco pipes and pipe-related materials, clothing-related artifacts, such as buttons and beads, and items like scissors and combs will be discussed in the following section. Lead cloth seals, also referred to as “bale seals”, have also been placed in this loosely defined artifact category, and will be discussed further at the end of this section.

Table 3 provides a list of all the artifacts that have been placed in this category, not including pipes or lead seals; many of these artifacts will be discussed at greater length below.

In terms of the clothing-related artifacts recovered from Java, three glass beads, 18 buttons, 13 buckles, and one hook and aigh were found (Figure 29). The vast majority of these were found in the excavation block immediately surrounding the footprint of the earthfast house.

Table 3: Personal Items Recovered from 18AN339

Object Material Count Weight Notes Unit Strat Bead Glass 1 0.023 Colorless 11 2 Bead Glass 1 0.02 Milky white 18 2 Bead Glass 1 0.014 Red 22 1 Buckle Copper/Alloy 1 0.123 Piece of buckle 1 4 Buckle Copper/Alloy 1 0.182 “S. COOK”; shoe buckle; 1660s-1720s 9 2 Buckle Copper/Alloy 1 0.155 12 2 Buckle Copper/Alloy 1 0.175 Rounded on top 12 2 Buckle Copper/Alloy 1 0.043 Trapezion shape double loop; 1600-1700 13 2 Buckle Copper/Alloy 1 0.185 Figure-of-eight double loop; 1600-1700 16 1 Buckle Copper/Alloy 1 0.11 16 2 Buckle Copper/Alloy 1 0.053 Partial; small 19 2 Buckle Copper/Alloy 1 0.159 “L”; plain shoe buckle; 1720s-1790s 21 2 Buckle Copper/Alloy 1 0.069 26 1 Buckle Iron 1 0.396 Possible shoe buckle; 1720s-1790s 23 1 Buckle Iron 1 0.208 Trapezion shape double loop; 1600-1700 23 2 Buckle Iron 1 0.213 Figure-of-eight double loop; 1600-1700 28 1 Button Bone 1 0.026 1 4 Button Bone 1 0.043 5 holes 2 2 Button Copper/Alloy 1 0.241 1 1 Button Copper/Alloy 1 0.154 9 2 Button Copper/Alloy 1 0.097 23 2 Button Glass 2 0.028 White, 4 holes 6 3 Button Glass 1 0.011 White, 2 holes 6 3 Button Glass 1 0.013 White, 4 holes, fragmented 27 2 Button Iron 1 0.05 14 2 Button Pewter 1 0.168 Iron shank; 3 circular cuts 23 1 Button Pewter 1 0.16 23 2 Button Porcelain 1 0.017 Half of a button STP 26 Button White Metal Alloy 1 0.119 Nipple button 10 2 Button White Metal Alloy 1 0.116 Nipple button 12 2 Button White Metal Alloy 1 0.035 12 2 Button White Metal Alloy 1 0.183 Nipple button 15 2 Button White Metal Alloy 1 0.12 26 3 Coin Copper Alloy 1 0.32 George I halfpenny, 1724 26 2 Chape Copper/Alloy 1 0.049 “TW” stamped into front; 1660-1720 12 1 Comb Bone 1 0.016 Fragment 15 2 Comb Bone 1 0.101 Fragment 17 1 Hook and Aigh Copper/Alloy 1 0.001 No hook, just aigh 22 2 Scissors Iron 1 1.474 Fragment 2 1 Scissors Iron 1 0.379 Fragment 7 2 Toy White Metal Alloy 1 0.089 Boy figurine with hands and feet missing 27 2 Whetstone Sandstone 1 0.814 Partial 19 2 66

Figure 29: Sample of buttons, buckles, and beads recovered from in and around the main excavation block at 18AN339 Of the buckles that could be identified, most date from the mid-17th through the mid-18th century. A single copper alloy chape, or buckle attachment, was recovered from strat 1 of Unit 12, located just outside the footprint of the earthfast building. The initials “TW” are stamped into this the front of this artifact and it resembles a stud chape, although the very tip has been broken off.

A George I halfpenny dating to 1724 was recovered from Unit 26, in the area of highest artifact concentration. This was the only coin recovered during the two seasons of excavation at Java (Figure 30). This coin was made the final year of the George I halfpences. On the obverse is a right-facing bust of the king with the inscription “GEORGIVS REX”, while the reverse has a depiction of the regal female, Britannia, with the fitting inscription “BRITANNIA” and the date of 1724.

Figure 30: Obverse (l) and reverse (r) of 1724 George I halfpenny from 18AN339

Two bone comb fragments were recovered from the excavation block. These combs are rectangular in shape and appear to be single-sided.

Figure 31: Two bone comb fragments recovered from site 18AN339

Only one toy fragment was recovered from the site. A small, cast lead alloy figurine shaped like a boy was recovered from Unit 27, located just to the west of Unit 13. This boy is missing his hands and feet, suggesting he was originally attached to something (Figure 32). There is incredible detail in this cast figurine, and his little eyes, nose, mouth, and hair is still quite visible.

Figure 32: Lead alloy cast toy fragment from 18AN339

A whetstone made of sandstone and two iron scissor fragments were recovered from the site. One of the scissor fragments was recovered from Unit 2, located to the south of the mansion ruins; the other was recovered from Unit 7, located in the northwestern corner of the excavation block. The whetstone was found in Unit 19, close to Unit 7. This object, while broken, is flat and rectangular with squared sides and a worn groove in the middle. This whetstone would have been highly useful for re-sharpening iron tools like scissors and knives, as these tools were prone to dent easily and the stone could make them useful again (Figure 33). It is, however, not out of the question that this whetstone dates to the prehistoric time period; a number of prehistoric artifacts were recovered from the area. Sandstone whetstones have been recorded on a number of prehistoric sites in Western states, where the native peoples would have used them for sharpening other stone tools or metal they obtained from trade.

Figure 33: Scissors and whetstone fragment recovered from excavation block at 18AN339

Tobacco Pipes

The total number of tobacco pipe fragments recovered at Java totals 1,957. It should be noted that all of the pipes recovered were made of white clay and none were made of red clay or terra cotta. Table 4 presents a breakdown of the recovered fragments:

Table 4: Total Pipes Recovered from 18AN339

Pipe Portion Number Percentage Bowls 1169 60% Stems 743 38% Joints (bowls and stems) 40 2% Heels 5 0.3% TOTAL 1957 100.3%

A diverse assemblage of bore diameters was noted from the pipe stems recovered at the site. These pipe stems reflect the long period of occupation at Java, although a few conclusions can be drawn (Figure 34):

Figure 34: Pipe stem bore diameters from 18AN339

As Figure 34 depicts, the highest percentage of recorded pipe stem bore diameters is between 5/64” and 6/64”. In general, pipes with bore diameters in this range were manufactured between about 1680 through about 1750, which spans the Sparrow period of occupation. This date range is also suggests that although the hilltop was occupied after the circa-1750 Maccubbin construction of the mansion house, this family was not living here year-round, as were their predecessors. At the minimum, the Maccubbins were not utilizing this portion of the site in the same manner as it was used by the Sparrows. Further excavations on the water side of the house may reveal similar Maccubbin period use-areas.

A number of pipes bowls and stems were decorated with rouletting, molded geometric designs, or makers marks. The marks on some of these pipes further suggest they were used during the Sparrow period of occupation (Figure 35).

Figure 35: Marked tobacco pipes recovered from excavation block at 18AN339

Smoker’s Companions and Strike-a-lights

A few tobacco pipe-related artifacts were found in the vicinity of the brick hearth. Two smoker’s companions and three strike-a-lights were recovered from the main excavation block.

Smoker’s companions have been referred to as the “Swiss Army Knife” of tobacco smoking in the Colonial Chesapeake. These handy little iron tools were used to pack, fill, clean, and light tobacco pipes by means of rounded ends, flat ends, and tongs. The two recovered from Java were found on what would have been the exterior of the 17th century earthfast building, one in Unit 10 to the south and one in Unit 14 to the west.

Three pieces of flint were identified as strike-a-lights by Lost Towns Project laboratory specialists. All three were recovered from units that would have been inside the earthfast building (Units 23 and 25). A strike-a-light was another pocket tool that was used to strike either another piece of flint or a piece of iron to cause a spark to set fire to dry tinder or tobacco. The flat end of a smoker’s companion was designed to make a spark with a flint strike-a-light. These

flints are often curvilinear with a concave surface on one side, and the technology used for making fire with these objects was used by native peoples all over the world. The three objects recovered from Java could have been used as gun flints first before being adaptively reused by the Sparrow family as strike-a-lights.

Lead Cloth Seals

Bale seals, better called lead cloth inspection seals, were utilized in the 17th and early 18th centuries to indicate that bolts of cloth had been subjected to various types of inspection during its trip across the Atlantic. Two marked seals were recovered from the southern excavation block at Java (one from Unit 10, the other from Unit 12) (Figure 36). The markings and designs on both of them suggest they were English royal alnage seals, dating to pre-1724 (Luckenbach and Cox 2003).

Figure 36: Lead cloth seals recovered from Unit 12 (l) and Unit 10 (r) at 18AN339

The first seal, recovered from Unit 10, was an alnage, or searchers seal. The marks on this seal (a crown with the initials “TX” on the reverse; “31” on the obverse) suggest that the cloth met certain quality standards and taxes had been paid on the bolt. This single, circular seal measures 24mm across, making it one of the largest seals recorded in Anne Arundel County (see Luckenbach and Cox 2003).

The second seal, recovered from Unit 12, was also an alnage seal that measures about 15mm across; it is heavily twisted so an exact measurement is hard to ascertain. This double, circular seal does have all of its parts intact (four disks connected by a lead strip), however mangled they may have become over the last 300 years. One side of the seal has a rouletted edge, while the other side has a crown over a rose next to the letter “I” over the number “2”. There might indeed be more markings on the seal that are not discernable due to the twisted nature of the artifact, but because it is so delicate, laboratory professionals were unable to see more.

Slag

While certainly not the most desired or exciting artifacts, the hundreds of pieces of slag recovered from the site merit some discussion. Slag is a waste material that is left as residue from metal smelting. No evidence was found at Java this season to suggest that there was a furnace capable of iron smelting in or near the excavation block (i.e., vitreous clay or heavy amounts of charcoal), but the 931 fragments recovered that weigh over 18 pounds strongly suggest that there was some sort of smithing taking place in the immediate vicinity of the earthfast building at some point during the occupation of 18AN339.

This slag clusters heavily in the southeast portion of the excavation block (Units 6, 10, 12, and 15), further suggesting that this was some sort of refuse disposal area (Figure 37). If at some point during the life of the earthfast building smelting was taking place within it, it would stand to reason that much of the waste material was simply thrown out of a door or window. As countless archaeological excavations have proven, it was very common to dispose of waste near the source. And as is the case with nearly every material type thus far discussed, this southeastern corner of the block produced the highest concentrations of artifacts.

70 1100 65 60 55 50 1090 45 40 35 30 1080 25 20 15 1070 10 5 0 -1090 -1080 -1070 -1060 -1050 -1040 -1030 -1020 -5 0 10 20 30 Ounces of Slag FT.

Figure 37: Distribution of slag (by weight in oz.) around excavation block with portion of conjectural earthfast building footprint at 18AN339

34 32 1100 30 28 26 24 1090 22 20 18 1080 16 14 12 10 1070 8 6 4 -1090 -1080 -1070 -1060 -1050 -1040 -1030 -1020 Ounces of Wrought Nails 0 10 20 30 FT.

Figure 38: Distribution of wrought nails (by weight in oz.) around excavation block with portion of conjectural earthfast building footprint at 18AN339

Comparing the distribution of slag to the distributions of nails raises the question, could the earthfast building have been used at some point during its life as a place of nail manufacture? The highest concentration of wrought nails and unidentified iron fragments was recovered from the same portion of the site as the slag, suggesting that the building could have been used for metal manufacture (Figure 38).

Future excavations into features in the area may reveal just how long this building was utilized. This portion of the yard could have been used as a work area during the Maccubbin and Contee occupations, regardless whether or not the old earthfast building still stood at the time. This postulation gains credence when it is considered that Java was not actually occupied full- time by the owners between about 1700 and 1830; therefore, industry taking place in the immediate vicinity of the earthfast house, and later the brick mansion, would not be an eyesore or so noisy as to disturb the landowners. The daily operations of the plantation were run by overseers and tenants for much of these decades, and as a successful and lucrative farming operation, countless nails and sheet metal would be called into service to construct outbuildings and repair equipment; hundreds of pieces of sheet metal and unidentified iron were also recovered from the excavation block. It stands to reason that some sort of blacksmithing operation would be required at the site. This earthfast building, once no longer used for a domestic capacity, could very well have used for this more industrial use.

Prehistoric Artifacts

Relatively few prehistoric artifacts were recovered from site 18AN339, but enough were recovered to be able to clearly state that Euro-Americans were not the first to utilize this hilltop. Prehistoric ceramics, a projectile point, and stone debitage were found spread across the site. Few diagnostic artifacts were recovered, and the assemblage seems to suggest light, ephemeral use of the hilltop by the native peoples of Anne Arundel County. Larger, more substantial prehistoric sites have been recorded all along the Rhode River (see Cox et al. 2007a; Cox et al. 2007b), but the hilltop was probably desirable to the Indians for the same reason it was to the Europeans who came here: it is the highest point in the area, providing a fantastic view of the Rhode River watershed, the soil is well-drained, and there is a small spring in the vicinity.

A total of 28 prehistoric ceramics were recovered from the site. These are listed by temper below in Table 5. Ten gravel tempered sherds, three sand tempered sherds, and 14 shell tempered sherds were found in the units in and around the main excavation block; one additional sherd was too small to determine temper. None of the fragments appeared to be decorated in any way, and most were too small to definitively determine what type they were and thereby when in the Woodland time period (circa 1,600 B.C.-circa 1,600 A.D.) they were manufactured. These artifacts are probably so small due to plowing and trampling by the Euro-Americans who subsequently occupied this hill for the last 350 years.

Table 5: Total Prehistoric Ceramics Recovered from 18AN339

Temper Count Weight Unit Strat Gravel Tempered 9 0.298 21 2 Gravel Tempered 1 0.075 26 3 No Temper 1 0.001 15 2 Sand Tempered 1 0.044 4 2 Sand Tempered 1 0.099 11 2 Sand Tempered 1 0.009 19 2 Shell Tempered 1 0.026 24 2 Shell Tempered 7 0.141 25 2 Shell Tempered 1 0.004 26 2 Shell Tempered 4 0.104 26 3 Shell Tempered 1 0.419 28 1 TOTAL 28 1.22oz

Lithic types recovered from the site include quartz, quartzite, chert, jasper, ironstone, and rhyolite. While it is not certain that all of these lithics are prehistoric, particularly the chert as it is easily confused for European gun flint fragments, Lost Towns Project laboratory staff determined these listed in Table 6 to be most likely utilized during prehistory. Quartz, quartzite, and ironstone are locally available materials, but the chert, jasper, and rhyolite would have made their way to site 18AN339 by long-distance trade.

A total of 130 pieces of debitage were recovered from the site, including cores, flakes, and shatter, here defined as angular chunks of debitage that can be produced at any stage of tool manufacture (see Table 6). Fourteen fire-cracked rocks were also recovered from the site, but these could have been utilized during the historic or prehistoric time periods. The single diagnostic artifact in the prehistoric assemblage is a quartzite Bare Island stemmed point. This

point has an average date of 1,500 B.C., and a range of 2,500 B.C. to 500 A.D., during the Late Archaic-Early Woodland transition (Hranicky and Painter 1989).

Table 6: Total Prehistoric Lithics Recovered from 18AN339

Material Type Count Weight (in ounces) Quartz Core 1 5.118 Flake 23 0.028 Shatter 16 3.007 TOTAL 40 5.118 Quartzite Projectile Point – Bare Island 1 0.457 FCR 2 39.0 Flake 5 0.652 Shatter 21 2.609 TOTAL 29 42.718 Chert Core 6 3.541 Flake 50 2.624 Shatter 4 0.386 TOTAL 60 6.551 Jasper Core 1 0.874 Flake 1 0.021 TOTAL 2 0.895 Ironstone FCR – TOTAL 12 26.538 Rhyolite Flake – TOTAL 2 0.023 TOTAL Projectile Point 1 0.457 FCR 14 65.538 Core 8 12.2 Flake 81 3.348 Shatter 41 6.002

Summary of Artifacts

The 34,141 artifacts recovered from two years of excavation at 18AN339 provide a broad snapshot of life in Anne Arundel County over thousands of years. Obviously, the assemblage strongly skews to the historic period occupation of the site; only 173 prehistoric artifacts were recovered. But the only diagnostic prehistoric stone tool recovered dates to as early as 4,000

years ago, demonstrating that people were utilizing the hilltop for thousands of years before Thomas Sparrow patented Sparrows Rest on September 22, 1652. The prehistoric assemblage suggests that local native peoples were using this site for minor campsites and some tool manufacture, but most evidence of their occupation has probably been obliterated by the heavy footprint left by people since 1652.

Although no intact strata have been excavated at the site and all of the artifacts have been recovered from mixed contexts, the majority of diagnostic historic-period artifacts date to the Sparrow period of ownership (1652-1747). The majority of the ceramic assemblage is tin-glazed earthenwares, the majority of the nail assemblage is hand wrought, and most of the diagnostic personal items, including the tobacco pipes, date to the late 17th and early 18th centuries. It seems likely that a large part of the reason why this assemblage skews to the Sparrows because of where we have concentrated our excavations (apparently within the 17th century occupation locus). It seems likely that we have not yet identified the areas of Maccubbin or Contee refuse disposal, where were would expect to see evidence of the higher quantity and variety of goods that were becoming more readily available.

The Sparrow-period artifacts speak to the life of a wealthy landowning family during the early colonial period in the Chesapeake. Lovely clothing adornments, such as decorative buckles, beads, and buttons were recovered, in addition to lead cloth seals that certified the raw textiles purchased by the family had been taxed by the English crown. Other objects such as musket adornments, decorated tobacco pipes, forks, knives, and spoons, highly decorated ceramic plates and glass drinking vessels, and casement window leads begin to paint a picture of the life lived by the family when they occupied the post-in-ground building.

The presence of later 18th and 19th century artifacts in the vicinity raises the question of how long this earthfast building stood and what purpose it served after it was no longer used domestically. Over 4,600 hand-wrought and square nails were recovered from the site seems like an excessive amount, even considering the amount needed to construct and repair an earthfast building. When comparing the similar distribution of slag and hand wrought nails, we have considered the possibility that some type of metal production took place in the house after it was abandoned by the family that lived here. Squirrel Neck plantation, and later Java, was an 80

incredibly successful farming operation and would have required a great deal of nails and sheet metal to sustain itself.

Hundreds of cut nails were also recovered from the vicinity of the footprint of the earthfast building. It is possible that these represent some later repair to the old, well-used building during the Contee occupation of the land. The presence of 19th century ceramics suggests that if this building was indeed standing when the Contee family owned the land and lived in the brick mansion house, this smaller building could have been used for storage of household goods or housing of tenants or plantation workers.

Site 18AN339 Summary

A major research goal at the outset of excavations in 2006 was to identify the earlier, 17th century occupation of the site that pre-dated the construction of the circa 1750 brick Georgian mansion house. A test unit excavated in 2006 along one of the few remaining standing walls of the mansion revealed that this building had not simply “absorbed” an earlier structure. So where was the earlier dwelling?

A shovel test pit survey was undertaken in 2006, followed by a 2007 magnetometer survey that covered much of the hilltop. When data from these surveys was compared, the potential 17th century portions of the site were narrowed to an area located about 100 ft. west, or on the inland side, of the mansion. As the end of the 2007 field season neared, the edge of an articulated brick feature was discovered in the corner of Unit 16. Once fully exposed, the 12 ft. long brick hearth surrounded by post holes clearly marked where an earthfast structure once stood. Analysis of the extensive artifact assemblage recovered from the plow zone overlying the features revealed the remains of a 17th century domestic building that was probably standing well into the 18th century, if not longer.

Extensive historic research and the recent archaeological investigations have clarified much about the inhabitants of site 18AN339. Thomas Sparrow, Englishman and Quaker, patented this land as Sparrows Rest in 1652, though the number of prehistoric artifacts recovered makes it clear the he was not the first person to make use of this hilltop. The prehistoric ceramics recovered from the site suggest Woodland period occupation, but a Bare Island projectile point suggests that native peoples could have utilized the area as early as 4,000 years ago.

While Thomas Sparrow (I) was the first European to lay claim to the land, it is not clear if he ever actually lived on Sparrows Rest. The documents only tell us that he died by 1659, when he willed the land to his son, Thomas Sparrow (II). This Thomas Sparrow wrote his will 1675, when he mentions two dwellings on Sparrows Rest. The first house he references is when he requests that “the building now begun upon my now dwelling plantation to be finished with all Convenient Speed” (MSA, Anne Arundel County Wills, Liber 2, folio 76:1675). The second

reference comes when he mentions that his sister, Elizabeth, should be allowed to continue to live on the parcel of land where she now resides that contains “the timber house.”

It seems highly likely that the brick hearth uncovered this field season was part of one of these two buildings. The hilltop is one of the most prominent spots in the area, providing a spectacular view of the Rhode River drainage. More than that, the artifact assemblage recovered from the area around the hearth skews heavily toward a domestic occupation that began in the 17th century. Delft ceramics fragments, Rhenish stoneware mugs, hundreds of pipe stems with bore diameters that date to the time period, and temporally significant buckles and lead cloth seals all speak to the life of a wealthy landowner in the early colonial time period. Two marked window leads were recovered from the edge of the building footprint (overlying a post hole), one with the date “1671” and the other with the initials “WM.” Similar markings on the two artifacts suggest they came from a single window lead, and an identical, whole lead was recovered from the contemporaneous van Sweringen site in St. Mary’s City (Hanna et al. 1992).

There are two possible scenarios for the 17th century occupation. This site could represent the incomplete house Sparrow intended his wife and children to live in. Alternatively, it may be the “the timber house” where his sister, Elizabeth, lived. Assuming the window lead was installed in the casement window of the house soon after its 1671 production, this is most likely “the timber house” that already stood by the time of Sparrow’s 1675 death. It is not out of the question, however, that the lead could have more than a four-year lag time between manufacture in England and use in a house in the colonial frontier of 17th century Maryland. Major construction was undertaken in 1676 at the van Sweringen site in St. Mary’s City, so it is possible the 1671 window lead could have been installed around that time.

While many questions remain about the 17th century building uncovered during the 2007 field season, several revelations are clear and well illustrated. First, the building did not have a stone or brick foundation, but was post-in-ground. The presence of a repair post suggests later alterations, restoration, or re-purposing of the building. Second, the building had a single-sided brick hearth, suggesting it was located at the gable end of the house. The presence of daub on site reveals that the chimney was constructed of wattle and daub at some point in the life of the house. The hearth measures 12 ft. on the outside and 8 ft. on the inside, which is a typical, if 83

slightly large size for a 17th century building. The postholes are situated almost exactly 16 ft. apart from one another. Earthfast buildings in the early colonial Chesapeake were often 16 ft. wide, represented the common colonial measurement of a “perch.” The postmolds, postholes, and hearth orientation suggest the building was built on an east-west axis. The highest concentration of temporally-related artifacts was recovered from just south of the building footprint, suggesting there was a window or a door in this area where refuse was disposed into the yard. The building had several casement windows, as a number of window leads were recovered from both sides of the house.

Thomas Sparrow (II) passed Sparrows Rest and all of his subsequent Rhode River acquisitions (Sparrows Addition and Locust Neck) on to his son, Thomas Sparrow (III). It is questionable how long Sparrow (III) actually lived on Sparrows Rest. In 1705, after the death of his second wife, he contracted relative William Coale and Daniel Richardson to care for his sons, Thomas (IV) and Solomon, in exchange for Coale and Richardson’s sole use of Sparrows Rest. Sparrow (III) had purchased about 3,000 acres near Bath, North Carolina at about the same time, and it appears he spent a great deal of time there and in the Outer Banks. Records indicate he participated in Cary’s Rebellion in 1711, an armed skirmish between the North Carolina Quakers and the English crown (Paschal 1955).

However, a lease agreement between Sparrow and William Coale in 1712 makes reference to a “mansion house” where Sparrow is living. The agreement states that Coale will lease two Rhode River parcels (Squirrel Neck and Thomas’s Quarter) in exchange for rent of one barrel of Indian corn paid yearly to Sparrow “at his Mansion house upon the plantation whereon he now lives” (MSA Liber 1B2, folio 76:1712). It is not clear where Sparrow is living, but he is referred to as “Thomas Sparrow of Road River,” strongly suggesting it is somewhere on Sparrows Rest.

Research questions raised by this recent investigation include whether it is possible that the “mansion house” of Sparrow (III) is one of the same buildings referenced by his father 37 years earlier. The artifact assemblage certainly suggests that this newly discovered earthfast building was occupied in some capacity through the early 18th century. English brown stoneware

manufactured during the reign of Queen Anne (1702-1714) and a 1724 George I halfpenny are well-dated examples that show some sort of occupation during the early 18th century.

Sparrow (III) passed all of his Rhode River landholdings to his son, Thomas Sparrow (IV). This Sparrow most likely did not live on the Rhode River, as documents suggest he had a “crippling disability” and was unable to farm (Sparrow 1990:400). Rather, he made his living as an inn holder in Annapolis and was involved in the provincial government (Sparrow 1990). In 1747, he sold all of his ancestral lands to Nicholas Maccubbin, a wealthy Annapolis merchant, who became the owner of site 18AN339.

It was Maccubbin who built the grand brick Georgian mansion house, and he and his heirs owned the land until 1819. He called the house and his plantation Squirrel Neck, and it seems likely that he did not live here year-round. Rather, he and his family probably split their time between his house in Annapolis and this country getaway. While Maccubbin and this site represent the upper echelon of colonial society, the ‘Merchant-Planter,’ much can be learned about those who actually worked on and ran the plantation, including Maccubbin’s slaves, from further archaeological excavations.

In the meantime, the artifact assemblage suggests that the small earthfast building may have still been standing along side the mansion in the mid to late 18th century, perhaps having been re-purposed as an auxiliary plantation outbuilding. It is somewhat difficult to imagine a hundred year old timber-framed building standing for long next to a grand mansion, but the recovery of hundreds of creamware and pearlware fragments from the area suggests that the building could have been used to house tenants, plantation workers, or enslaved laborers. If this is the case, further investigations can possibly reveal the adaptive reuse of an earthfast building long after it fell out of fashion as a home for landowners and came to be viewed in a utilitarian way.

A large amount of slag, or waste from metal smelting, was found in the southern side of the house and just to the exterior of the south side of the building. This is also the location of the highest percentage of wrought nails recovered. Hundreds of pieces of sheet metal and iron waste were also found in the area. This possibly suggests that some type of blacksmithing operation

was taking place in the old building. The 1798 Federal Tax Assessment lists the brick mansion, eight small dwellings, and 40 slaves on the property at the time. A plantation as large as this, with a slave labor force of nearly 80, would not only require countless nails and metal to repair and maintain the machinery needed to run the farm, but would have the labor to make these items on-site. The presence of a blacksmith on property would not be unexpected.

John Contee, a wealthy Prince George’s County landowner, acquired the property in 1819 and by 1828 it was known officially as the Java Plantation (Lee 2004). He named it after a British ship that he helped capture and destroy during his naval service during the War of 1812. But it seems clear that Contee did not live at Java full-time, and instead followed in his predecessor’s footsteps as an absentee landowner. The distribution of the 423 cut nails recovered scattered around the footprint of the earthfast building suggests an outside chance that the old building may have still been standing and was being repaired with newly available machine cut nails as late as the 1820s or 1830s. A number of whiteware fragments in the vicinity further suggests that some type of domestic occupation could have been taking place. Regardless of whether or not the old post-in-ground building stood during the Contee occupation, the proximity of the current excavations to the mansion suggests a certain degree of 19th century artifact admixture is not unexpected.

It was not until John Contee died and passed the land on to his wife, Ann, that the Squirrel Neck mansion was occupied year round. Census records show Ann and her two sons, Richard and Charles, were living there in 1840. Documentation from this time period indicates that Java plantation was an incredibly successful operation. In 1839, 84 slaves produced over 70,000 lbs of tobacco, far exceeding every other farm in the Rhode River watershed (Greenburg and Hyatt 1990). Ann passed the land containing the mansion on to her younger son, Richard, and he and his wife also appear to have been full-time residents of the brick mansion.

Considering nearly 100 years passed between the construction of the brick mansion and when a family actually lived there year-round, it does not seem out of the question that the post- in-ground building could stand in the back yard serving some sort of utilitarian purpose. But the relatively few number of cut nails and later 19th century ceramics and bottle glass suggest the house was probably finally torn down by this time and the Contees were clearly not using this 86

area of the plantation as the dumping ground for their household trash. There is no evidence that the wooden building burned down; little charcoal was found in the area and few of the nails were annealed. If it did not simply collapse from old age, it was most likely torn down sometime in the 18th or early 19th century.

Some later 19th and 20th century materials were found in the excavation block, including 39 wire nails and a few porcelain dish fragments decorated with orange starburst decals. This is not entirely surprising, as the brick mansion house was sporadically occupied after the end of the Contee occupation in the 1870s. The mansion was struck by lightning around 1890, but the owners attempted to rebuild it and continued to live there sporadically until the 1920s. The later artifacts recovered probably represent this periodic occupation of the main house. It does appear as if some plowing may have been taking place around the remains of the old earthfast house, as a number of later artifacts were recovered from all levels of the soil strata. Wire nails were recovered from the same layers as 17th century delftware, for example.

Nearly a century has passed since anyone could live in the brick mansion that was occupied off and on for about 175 years. The archaeological data suggests that there’s a possibility that the earthfast building might have stood for almost as long. This question will shape future research questions as research and public programming efforts continue in the shadow of the mansion ruins at 18AN339 under the new ownership of SERC. Future research efforts on the hilltop include sampling some of the features to determine their respective dates, which will help clarify how the building was used, repaired, and re-purposed over the centuries. Future work should record the full dimensions of the earthfast building, and determine if it was built in stages, thereby providing a better sense of the spatial layout of the hillside, both before and after construction of the brick mansion. Finally, more testing is warranted on the opposite side of the brick mansion to determine if parts of the water side of the hilltop were utilized by the Sparrow family and how the Maccubbins and Contees viewed this portion of their space.

The research potential of site 18AN339 is almost limitless. With over 350 years of continuous historic occupation and perhaps thousands of years of prehistoric occupation before that, there is a great deal that can be learned here about the rich history of Anne Arundel County. As the three year MHT grant comes to an end, we intend to continue to share with County 87

residents the excitement of discovering this small piece of our shared heritage. Recently, the Lost Towns Project was informed that the property upon which 18AN339 (and several other historic and prehistoric sites) has been acquired by the Smithsonian Land Trust, and the land, its management, and its opportunity for public programming, will be consolidated within the Smithsonian Environmental Research Center. While details of partnership and the ongoing research and educational programming are under discussion, the past three years of historic context development, archaeological testing and public programming supported by the MHT has laid a strong foundation for the continued research opportunities on 18AN339.

CHAPTER TWO: 18AN1285

Limited Phase III Investigation at 18AN1285: Camp Letts, a Middle to Late Woodland Campsite Introduction

The Middle Woodland time period, by all accounts, was a time of great change in the prehistory of the Eastern United States. Most scholars believe that Eastern Algonquian-speakers from the Great Lakes region migrated into the Middle Atlantic region sometime during this phase, broadly dating from about 300 B.C. to A.D. 900 (Dent 1995:9). This long-distance movement of populations probably occurred sporadically during the later Middle Woodland, dating from about A.D. 1 to A.D. 900 (Potter 1993:3), and many scholars (see Schindler 2006) use this narrow time period as the defining marker of the Middle Woodland. This portion of the Middle Woodland in Anne Arundel County is referred to as the Selby Bay phase.

It was these migrating native peoples who were the direct ancestors of the Algonquian Indians met by Europeans when they arrived in the Chesapeake Bay in the early seventeenth century. Therefore, the cultural trends initially developed in the Middle Woodland period had weighty consequences on the eventual Indian-European interactions that marked the beginning of the end of Chesapeake Indian society. A greater understanding of the culture of the Middle Woodland can provide context and perspective for understanding the complex Late Woodland societies that the Europeans encountered and recorded.

The Middle Woodland is marked in the archaeological record by a “technological homogenization” (Dent 1995:235) of ceramic and projectile point types and lithic preference. Native peoples in the region had been manufacturing pottery since the beginning of the Early Woodland time period (ca 1,000 B.C. to 300 B.C.), but by about A.D. 200, a type known as Mockley ware came to dominate the Middle Atlantic region. Mockley ceramics are oyster shell- tempered, thick-bodied, and often cord-marked or net-impressed (MAC Lab 2008). Projectile point types became more standardized, with a preference seen for Jacks Reef and the side- notched Selby Bay type. Preference for lithic type changed from earlier time periods when Indians utilized locally available materials to a desire for non-native materials. Rhyolite, a type

of igneous rock, became the exotic material of choice. This lithic type would have to be procured from above the fall line, signaling the development of long-distance trading networks. All of this “homogenization” strongly suggests increasing interaction between groups in the Middle Atlantic region.

In Anne Arundel County and along most of the Chesapeake Bay, many Middle Woodland sites are marked by shell middens, or large piles of shell (usually oyster in this region) representing an area of waste disposal. Archaeologists have discovered discarded ceramics, lithics, bone, and other plant residue in these middens, enabling them to state that the Middle Woodland was a time of large-scale exploitation of the available resources in the Bay and its tributaries (Dent 1995). These middens were formed during seasonal occupation of the low- lying areas along the water where the Indians would hunt and gather for subsistence. Semi- permanent base camps were established further inland, where the predicable anadromous fish were harvested and where some evidence of the first plant cultivation is seen by the end of the Middle Woodland (Potter 1993:109). This has led some scholars (see Gilsen 1978) to call this period one of a transition between hunting and foraging to the sedentary lifestyle that typified the Late Woodland.

Nearly 200 Middle Woodland period sites have been recorded by laypeople and archaeologists working in Anne Arundel County over the years, and their locations are on file with the Maryland Historical Trust (MHT). Sites identified and excavated in the Rhode River watershed over the years have contributed to the knowledge base about the Selby Bay phase of the Middle Woodland, including the Smithsonian Pier site (18AN284). This site, excavated in advance of the construction of the Education Center on the SERC campus, demonstrated that native populations tended to visit this coastal area during the autumn and spring to access the abundant local oyster population (Gibb and Hines 1997). Additionally, the majority of the prehistoric sites identified during Year One of this Rhode River survey dated to the Middle to Late Woodland time periods, strongly suggesting that this watershed was intensively used during the Selby Bay phase.

This season at Camp Letts (18AN1285) was focused on gaining further knowledge of the transition phase between the Middle and Late Woodland time periods. Earlier investigations at

the site had produced cultural materials from both periods, possibly suggesting the site could have been occupied near the end of the Selby Bay phase. It was also hoped that the two distinct middens previously identified on two separate portions of the peninsula could lend insights into spatial use of a procurement site.

18AN1285 Previous Archaeology

Site 18AN1285, a prehistoric campsite with oyster middens, was identified on the Camp Letts property in 2005, during the first year of the Rhode River survey (Cox et al. 2007a). The site is located on a small peninsula on the southwest shore of Bear Neck Creek, opposite Holly Hill Harbor and the mouth of Whitemarsh Creek (see Figure 1). The vegetative setting consists of a mixed deciduous and coniferous forest with relatively little undergrowth. Twelve STPs yielded prehistoric ceramics, debitage, and burnt animal bone in 2005.

Further archaeological work was conducted in May and June of 2006 during the second year of the Rhode River survey. During this Phase II testing, the Lost Towns Project identified this prehistoric site as eligible for the National Register and results suggested that the site offered a high level of research potential when compared to other prehistoric sites identified in the area. Two 5-ft. by 5-ft. units and a 5-ft. by 2.5-ft. unit were excavated (see Figure 40). Units 1 and 2 were placed in the northern portion of the site and were the larger of the three. Three strata were identified in both units, and artifacts recovered from Unit 1 included an undiagnostic rhyolite projectile point, thermally-altered quartzite and chert, and shell-tempered and Potomac Creek ceramics dating to the Late Woodland (A.D. 900 – contact) time period. Unit 2 was located to the south and west of Unit 1 and produced one quartz Selby Bay knife, shell-tempered and Potomac Creek ceramics, and one quartz flake (Cox et al. 2007b).

Unit 3 was located near the southern tip of the peninsula and measured 5-ft. by 2.5-ft. It was placed here to further investigate the shell midden noted during the earlier phases of work. Three sherds of Mockley ware, a thick shell-tempered ceramic dating to the later Middle Woodland time period (ca. A.D. 200 – A.D. 900), were recovered from Stratum 1. No artifacts were recovered from the shell midden itself, which was represented by Stratum 2 (Cox et al. 2007b).

The artifact assemblage suggests that this site dates to the late Middle Woodland through Late Woodland time periods of prehistory. This peninsula was most likely used as a temporary campsite that was returned to seasonally, primarily for oyster procurement. Based on the wealth of data recovered, and the potential for better understanding these sites that are so typical along the shorelines of the Rhode River and throughout Anne Arundel County’s riverine-rich environment, further testing was recommended.

2007 Excavations

The first step undertaken during the 2007 field season was to reestablish the 2006 grid and to perform a magnetometer survey (Figure 39). Considering the relatively pristine and unpolluted condition of the site, and the fact that previous investigations revealed little historic activity on the point, the magnetometer data held the potential for pinpointing possible prehistoric features, such as a hearth or fire-cracked rock concentration. While the overgrowth and density of tree cover did present a surveying challenge, the magnetometry data suggested a few areas of higher potential for such features; however, only the northern half of the site was tested in this manner. The promising anomalies were ground-truthed as discussed below.

260 52600 52550 52500 240 52450 52400 52350 52300 220 52250 52200 52150 52100 200 52050 52000 51950 180 51900 51850 51800 51750 160 Magnetic Signature nano-Tesla's 140

120

100

220 240 Figure 39: Results of the magnetometer survey from 18AN1285 A total of five units were excavated at the site in 2007, all measuring 5-ft. square. Three of these, Units 4, 7, and 8, were clustered in a block roughly in the center of the site, which had produced a high number of ceramics and lithics in 2006. One unit, Unit 5, was placed near the southern point to further investigate the oyster midden noted in that area in 2006. The final unit, Unit 6, was placed on the northern end of the site to correspond to the location of magnetic anomaly noted during the magnetometer survey (Figure 40).

Figure 40: Site plan, with numbered excavation unit locations, from 18AN1285 94

Units 4, 7, and 8 - Excavation Block

Unit 4, the first unit opened in 2007, was placed generally between Units 1 and 2 in the center of the peninsula to further investigate the high concentration of ceramics and debitage found in 2006. Two cultural levels were initially excavated that contained quartz, quartzite, and chert debitage, charcoal, oyster shells, and a great deal of shell-tempered, thick-bodied Mockley ware prehistoric ceramics (n=30). This was a strikingly high number of ceramic sherds relative to the units excavated in 2006, and many of these sherds were quite large.

At the base of Strat 2, four more large Mockley ceramics were recovered from a dark stain in the southwest corner. Additionally, eight heavily burned prehistoric ceramics were coming from over another stain in the northwest corner, which also seemed to be full of oyster shell. In sum, 42 Mockley sherds (nine cord-marked) were recovered from Strats 1 and 2 (Table 7; Figure 41). Rather than excavate these partially exposed features, another unit adjacent and to the west was opened.

Table 7: Total Assemblage from Unit 4 at 18AN1285

Stratum Material Type Count Weight (in ounces) 1 Earthenware Unidentified, high sand content 1 0.021 Lithic Chert flakes (secondary) 2 0.093 Shell Oyster 16 10.00 Brick 2 0.039 TOTAL STRATUM 1 21 10.153 2 Earthenware Mockley, cord-marked 9 1.905 Earthenware Mockley, unidentified surface treatment 20 0.538 Earthenware Mockley, interior burning 4 0.23 Earthenware Unidentified, shell-tempered, heavily burned 4 0.145 Earthenware Unidentified exterior spall 1 0.016 Earthenware Unidentified, high sand content 4 0.25 Lithic Quartz flake (secondary) 1 0.044 Lithic Quartzite, fire-cracked rock 1 1.696 Lithic Unidentified ironstones 3 0.62 Charcoal 2 0.004 Shell Oyster 37 32.0 TOTAL STRATUM 2 86 37.448 3 Earthenware Unidentified shell-tempered 4 0.029 Earthenware Unidentified, shell-tempered, heavily burned 1 0.03 Earthenware Unidentified sand-tempered 1 0.1 Shell Oyster 1 0.532 TOTAL STRATUM 3 7 0.691

Figure 41: Sample of ceramics recovered from Unit 4, Stratum 2 at 18AN1285; cord-marked Mockley (l) and shell-tempered sherds with varying stages of burning found in the northwest unit corner (r) Unit 7, adjacent to Unit 4, contained two cultural levels that came down on a dense oyster shell midden. The two strata above the midden produced 20 Mockley sherds (nine of which were cord-marked), four thermally-alerted sandstone rocks, and one possible brick or low-fired clay fragment. Also, a number of oyster shells were recovered, some that were covered with an orange, crystalline encrustation. Under a microscope, this orange crust was revealed to be soil that had turned into hard sediment after years exposed to the elements in the midden.

The dense oyster midden, called Stratum 3 in Unit 7, was excavated separately. Thirty Mockley ware ceramic sherds (nine large enough to be noted as cord-marked or fabric- impressed), fire-cracked rocks, charcoal, small pieces of bone, and hinged oyster shells came from this midden alone. Coincidentally, it seemed the Unit 4/Unit 7 wall marked the edge of the midden, which also meant the two dark, Mockley-producing stains in Unit 4 were located at the exterior edge of the midden (Figure 42; Table 8).

Table 8: Total Assemblage from Unit 7 at 18AN1285

Stratum Material Type Count Weight (in ounces) 1 Earthenware Mockley, cord-marked 5 0.163 Shell Oyster, orange encrusted 1 0.4 Shell Oyster 5 3.053 TOTAL STRATUM 1 11 3.616 2 Earthenware Mockley, cord-marked 4 0.125 Earthenware Mockley, unidentified surface treatment 9 0.214 Earthenware Mockley, with ochre inclusions 2 0.018 Masonry Brick fragment 1 0.033 Lithic Sandstone, thermally altered 1 1.695 Lithic Sandstone, unidentified use 3 0.069 Shell Snail 1 0.018 Shell Oyster, orange encrusted 2 0.073 Shell Oyster 20 12.0 TOTAL STRATUM 2 43 14.425 3 (Midden) Earthenware Mockley, cord-marked 7 0.779 Earthenware Mockley, fabric-impressed 1 0.082 Earthenware Mockley, with ochre inclusions 1 0.044 Earthenware Mockley, unidentified surface treatment 21 0.444 Lithic Quartzite, fire-cracked rock 1 0.102 Lithic Chert, fire-cracked rock 1 1.637 Bone Unidentified 1 0.006 Charcoal 19 0.322 Shell Oyster 85 61.0 TOTAL STRATUM 3 (MIDDEN) 137 64.416 4 Earthenware Mockley, cord-marked 7 0.819 Earthenware Mockley, unidentified surface treatment 13 0.308 Lithic Quartz shatter 1 0.062 Charcoal 1 0.005 Shell Oyster 6 3.199 TOTAL STRATUM 4 28 4.393

EU8

Feature 1

Midden EU4

Grid True North EU7 0 5 Feature 2 Scale (in ft.)

Figure 42: Plan view of midden with nearby features in Units 4, 7, & 8 at 18AN1285 The midden appeared to extend on a north-south axis, so in an attempt to find another edge, Unit 8 was opened, adjacent and to the north of Unit 7. Excavation revealed two cultural layers above the midden, which tapered out to the western side of the unit, consistent with what was seen in Unit 7 (see Figure 42). Also consistent with Unit 7, a high number of Mockley ceramics came from Strata 1 and 2 (n=39; 12 were cord-marked) (Table 9). In addition to these sherds, three pieces of fire-cracked rock, more orange sediment-encrusted oyster shells, and a few historic period artifacts (one cut nail, one possible brick fragment) were recovered.

An unusual collection of artifacts came from the shell midden in Unit 8 (see Table 9). A total of 264 fragments of shell-tempered ceramics were recovered, and many of them were very friable (Figure 43). No surface treatment was able to be determined from any of the sherds, but there was a great variety in color, texture, and width; however, the overwhelming majority was shell-tempered. Only five sherds were recovered that appeared to have been tempered with sand, although the paste of many of the shell-tempered sherds was very sandy, suggesting these sherds

may have come from the same clay source. Many of the sherds showed evidence of burning and a few of them had red ochre inclusions, possibly from the clay source.

As mentioned, many of the sherds were very friable, or very crumbly and brittle. One of these ceramic sherds was actually still attached to an oyster shell. It was brought back to the lab and cleaned with great care (see Figure 43). This vessel was coil-constructed, and three distinct coils and two brakes are still clearly visible in this particular sherd. Additionally, one piece of unfired or very low-fired clay was recovered from the midden that was the same color and consistency as the majority of the sherds from the midden. This was found along with two small, possibly degraded sandstones pieces (see Figure 43). Other artifacts found in the midden included a charred nut or seed, a quartz shatter fragment, and two rocks that may or may not have been cultural. One rounded quartz stone had only a single deep pock mark in the side that broke through the cortex, possibly suggesting wear. The second stone was an oval, flat sandstone rock with no apparent use marks. No other rocks or stones were recovered from this shell midden.

100

After the midden portion of Unit 8 had been completely excavated, the corresponding area on the western side of the unit that would have been located just outside the midden was removed (called Stratum 4). Interestingly, only a single charcoal bit and a few oyster shells were recovered from this area (see Table 9). This stands in contrast to the culturally rich soils found on the other side of the midden in Unit 4 (in Stratum 2 – see Table 7).

The shell midden was then removed from Unit 7 to compare this portion to that seen further north in Unit 8. A total of 30 Mockley sherds were recovered, and of those with recognizable surface decoration, seven were cord marked and one was fabric-impressed (Figure 44; see Table 8). However, none of the sherds were friable and crumbly like the large collection recovered from Unit 8. Two fire-cracked rocks were recovered from this portion of the midden, along with more charcoal and one small indeterminate bone fragment.

Figure 44: Portion of ceramic assemblage recovered from shell midden (Stratum 3) in Unit 7 at 18AN1285

101

Table 9: Total Assemblage from Unit 8 at 18AN1285

Stratum Material Type Count Weight 1 Earthenware Mockley, cord-marked 4 0.413 Earthenware Mockley, unidentified surface treatment 8 0.412 Lithic Chert, fire-cracked rock 1 1.462 Shell Oyster, orange encrusted 13 2.763 Shell Oyster 3 1.051 TOTAL STRATUM 1 29 6.101 2 Earthenware Mockley, cord-marked 7 0.8 Earthenware Mockley, unidentified surface treatment 20 0.44 Masonry Cut nail fragment 1 0.097 Masonry Possible brick fragment 3 0.087 Lithic Quartzite, shatter 1 0.2 Lithic Quartzite, fire-cracked rock 1 2.13 Shell Oyster, orange encrusted 4 0.175 Shell Oyster 15 9.8 TOTAL STRATUM 2 52 13.729 3 (Midden) Earthenware Unidentified, shell-tempered, black, small, pebble-shaped frags. 186 1.568 Earthenware Unidentified, shell-tempered, black, flat on both sides 5 0.449 Earthenware Unidentified, shell-tempered, black, adhered to oyster shell 1 0.578 Earthenware Unidentified, shell-tempered, reddish, thin-bodied 6 0.359 Earthenware Unidentified, possible surface treatment, tannish-exterior, burned 64 1.915 interior Earthenware Unidentified, shell-tempered, tannish-exterior, burned interior, 1 0.024 ochre inclusions Earthenware Unidentified, high sand content 5 0.079 Earthenware Very low-fired black clay ball, no temper 1 0.026 Earthenware Low-fired red clay balls 2 0.036 Lithic Friable sandstone, one black, one red 2 0.027 Lithic Quartz shatter 1 0.028 Lithic Quartz, rounded rock with deep pock in side 1 2.245 Lithic Quartzite, unidentified use, flat stone 1 6.348 Seed Possible charred nut 1 0.004 Shell Snail 1 0.01 Shell Oyster 80 86.0 Shell Oyster temper bits 7 0.043 TOTAL STRATUM 3 (MIDDEN) 365 99.739 4 Charcoal 1 0.002 Shell Oyster 11 13.5 TOTAL STRATUM 4 12 13.502 5 Earthenware Mockley, cord-marked 3 0.217 Earthenware Mockley, unidentified surface treatment 3 0.042 Earthenware Very low-fired reddish clay ball, no temper 2 0.009 Lithic Quartz, possibly thermally altered 2 1.717 Lithic Chert, possibly thermally altered 1 1.487 Charcoal 2 0.004 Shell Oyster 1 1.239 TOTAL STRATUM 5 14 4.715 6 Earthenware Mockley, fabric-impressed 1 0.129 TOTAL STRATUM 6 1 0.129

102

The layers underneath the midden were then removed in Units 7 and 8; this stratum (called Strat 4 in Unit 7; Strat 5 in Unit 8) was terminated at what appeared to be subsoil in each unit. A total of 20 sherds of Mockley ware were recovered from beneath the midden in Unit 7 (see Table 8). Of these, seven were cord-marked and 13 were too small or degraded to determine if a surface treatment existed. One charcoal fragment and one piece of quartz shatter were also recovered. The assemblage from the layer underlying the midden in Unit 8 consisted of only six Mockley sherds (three cord-marked), one very low-fired or unfired clay ball, two small charcoal flecks, and two possibly thermally altered stones (one chert, one quartz) (see Table 9).

Though the soil appeared to be subsoil, an additional cautionary 0.3 ft. thick arbitrary level (Stratum 6) was removed from Unit 8 and completely screened. Only one small Mockley sherd was recovered from this otherwise sterile layer, likely due to bioturbation, as it was recovered near a root. Excavation was halted in Unit 8.

As a comparison to the strats seen under the midden, the corresponding level (called Stratum 3) in Unit 4, located outside and to the east of the midden, was removed. Only one small, unidentified sand tempered ceramic sherd was seen here, but a number of Mockley sherds (n=5) were noted coming from the northwest corner (see Table 7). An oval, dark stain, full of oyster shell was partially exposed in this portion of the unit and labeled as Feature 1 (Figure 45). There was a second feature also noted at this level in Unit 7, called Feature 2 (Figure 46; see Figure 42).

103

Figure 45: Plan view of Feature 1 at 18AN1285

Figure 46: Profile of bisected Feature 2 at 18AN1285

104

Due to time and budget constraints, Feature 1 was not fully exposed; doing so would have required opening another test unit to the north of Unit 4. Instead, the portion of the feature exposed within the unit was removed. It was first bisected on a north-south axis, and the west half was removed; the east half was subsequently removed. The excavator noted that it was much deeper than she expected, plunging nearly half a foot into subsoil. It was also full of oyster shells that “pointed down”, as if they had fallen into the pit. This stands to reason, considering the proximity to the shell midden.

There were two layers within Feature 1. Layer A (10YR 5/3 brown sandy loam) contained the shells, along with seven sherds of Mockley ware (four cord-marked and three burned sherds that exhibited fireclouding, or burning on the broken, interior edges), eleven unidentified burned bone fragments, one quartz fire-cracked rock, and two very low-fired, delicate clay fragments (Figure 47; Table 10). Layer B was slightly redder than Layer A (10YR 5/6 yellowish brown coarse sand with no inclusions), and it appeared burned in comparison; however, no charcoal was noted (Figures 48 and 49).

Figure 47: Total ceramic assemblage from Feature 1 at 18AN1285, including (clockwise from upper left) very low fired clay fragments, cord-marked Mockley, and burned Mockley that exhibits fireclouding

105

Table 10: Total Assemblage from Feature 1 at 18AN1285

Stratum Material Type Count Weight (in ounces) A Earthenware Mockley, cord-marked 4 0.205 Earthenware Mockley, burned 3 0.159 Earthenware Very low-fired or unfired clay 2 0.046 Lithic Quartz, fire-cracked rock 1 0.478 Faunal Bone, burned 11 0.203 Shell Oyster 9 3.772 TOTAL FEATURE 1 30 4.863

Figure 48: North wall profile of Unit 4, including profile of excavated Feature 1 at 18AN1285

106

Figure 49: North wall profile of Unit 4 at 18AN1285 Feature 1 appears to be a pit of some sort where burning occurred, although the exact use of the pit is unclear. The presence of burned bone suggests the Indians may have been using this pit for cooking, and the three sherds of Mockley that are burned on the broken edge strongly suggests that a vessel broke and continued to be exposed to heat after breakage. The outer edge of the pit was reddened and compact, like it had been exposed to heat. Considering its proximity to the midden, the activity could have been related to the processing of oysters in some way.

Feature 2, located on the western edge of the midden in Unit 7, turned out to be a very shallow, small, circular hole that contained five sherds of fabric-impressed Mockley ware and one oyster shell (Figure 50). This feature was also bisected and later fully excavated, and could have been part of a post or another small pit used for an unknown purpose.

107

Figure 50: Fabric-impressed Mockley sherds from Feature 2 at 18AN1285 (top) and plan view of excavated feature (bottom) The Munsell colors and textures of the upper strata were nearly identical in all three of these units. Stratum 1 contained a 10YR5/2 grayish brown fine sandy loam and basically consisted of the upper humic layer (the O horizon). Stratum 2 (the A horizon, apparently unplowed) consisted of a 10YR 6/4 light yellowish brown sand with occasional oyster shell.

108

This sandy layer was noted to overlay a slightly darker 10YR 5/4 yellowish brown sand in some portions of the profile (Figures 51 and 52). Where it was noted, the soil from within the midden consisted of a 10YR 5/6 yellowish brown sandy loam with dense whole and fragmented shells. Subsoil (the B horizon) was generally a 2.5Y 5/4 light olive brown compact sandy clay, but a second, lighter strata that transitioned to subsoil was noted in the north wall of Unit 4, which consisted of a 10YR 7/3 very pale brown compact coarse sand (a possible E horizon) (Figures 53 and 54).

Figure 51: South wall profile of Unit 4 (to left) and Unit 7 (to right) at 18AN1285

109

Figure 52: South wall profile of Units 4 and 7 at 18AN1285

Figure 53: East wall profile of Unit 8 at 18AN1285

110

Figure 54: East wall profile of Unit 8 at 18AN1285

Excavation Block (Units 4, 7, & 8) Summary

A dense but shallow oyster shell midden was identified in the Unit 4/7/8 excavation block, and the artifacts recovered within and immediately surrounding it suggest it was created and used during the later part of the Middle Woodland time period of prehistory (ca A.D. 200 – A.D. 900). The midden was not fully exposed, but at minimum, the feature measures 10 ft. long by 5 ft. wide. This midden is situated on nearly the highest point of the peninsula on very well- drained soils.

Two pit features were identified on either side of the midden, to the east and west. The eastern feature, Feature 1, was the larger of the two, diving nearly 10 inches into subsoil. Burned, slightly reddened soil along the exterior of the pit and burned bone within it indicate something was certainly being cooked or fired here, despite the total lack of charcoal. Given its proximity to the midden (less than one foot from the edge), it stands to reason that the utility of this pit was related to the processing of oysters. The second feature, Feature 2, located to the west, was very shallow, small, and circular, and possibly represented the base of a post hole, although this is not certain. In contrast to Feature 1, which contained a few small burned and non-burned ceramic sherds and burned bone, Feature 2 contained five relatively large, unburned, fabric-impressed sherds of Mockley ceramics, two of which mend. There was no evidence of burning in Feature 2.

111

Nearly all of the ceramics recovered from the excavation block were shell-tempered, and of those that could be identified, all were Mockley. All of the sherds had a high sand content, although none appear to be sand-tempered per se. Rather, this is probably a result of the sandy clay that was being used to construct the vessels. Of those that could be identified, most sherds appear to be fragments of coil-constructed vessels.

Many of the Mockley sherds were cord-marked, which is a typical surface treatment for the ware type. However, a number of fabric-impressed sherds were also recovered, which is somewhat more unusual. While some variety in types were recovered from other units in the area, the overwhelming presence of Mockley from the midden area strongly suggests it was deposited during the later Middle Woodland, often referred to as the Selby Bay period (ca. A.D. 200 – A.D. 900). The Selby Bay knife found in the vicinity during the Phase II enforces this conclusion.

Perhaps the most interesting find from the excavation block was the over 260 partially- fired, friable clay sherds recovered from the shell midden in Unit 8 that indicate there was pottery manufacture taking place somewhere on site. These sherds seem to fall somewhere between clay and fully-fired pottery. When exposed to water, these fragments do not revert back to a clay-like state, nor are they impervious, like a fully-fired vessel. Rather, they begin to crumble and dissolve as if they were only partially fired. The sherd still adhered to an oyster shell demonstrates an effort at constructing a vessel by coiling the clay; three partial coils are still visible adhered to this shell. The presence of these friable sherds in the midden strongly suggests a failed attempt at pottery manufacture somewhere on site (Michael Stewart, personal communication, 2008). And their spatial location entirely within the midden of Unit 8 indicates they were deposited during a single event.

Unit 5

Unit 5 was placed near the southern tip of the peninsula to further investigation this portion of site 18AN1285 (see Figure 40). During the Phase II, Unit 3 was excavated in the vicinity and a portion of a dense, thick shell midden was excavated. A few Mockley sherds were

112

the only artifacts recovered from the midden, but the unit was smaller than most due to time constraints (2.5 ft. by 5 ft.).

To get a sense for the use of the southern portion of the midden, Unit 5 was placed just to the south and east of Unit 3. It measured 5 ft. square and ultimately ended up on the eastern edge of the larger midden. This provided for an interesting comparison about what was going on at midden’s edge versus inside the midden itself, as was seen in the Unit 4, 7, and 8 excavation block to the north.

The midden was located in the western half of the unit and it was first encountered once Strat 1, representing the upper organic layer (10YR 6/3 pale brown sand), was removed. Two undiagnostic sand-tempered wares were recovered from this layer, in addition to a number of whole oyster shells, three periwinkle shells, and a few charcoal bits (Table 11). Strat 2 was an interface layer excavated in the eastern half of the unit (2.5Y 4/3 olive brown fine sand with frequent oyster shells). This fully exposed the portion of the midden located within Unit 5, and produced only oyster shells and charcoal bits.

113

Table 11: Total Assemblage from Unit 5 at 18AN1285

Stratum Material Type Count Weight (in ounces) 1 Earthenware Unidentified Sand-tempered 2 0.045 Charcoal 3 0.008 Shell Periwinkle 3 0.087 Shell Oyster 23 15.0 TOTAL STRATUM 1 31 15.14 2 Lithic Possibly slag? 1 0.02 Charcoal 22 0.134 Shell Oyster 14 13.75 TOTAL STRATUM 2 38 13.904 3 (midden) Lithic Quartzite fire-cracked rock 1 6.411 Lithic Unidentified, probably cultural 1 3.146 Shell Oyster 54 65.0 TOTAL STRATUM 3 56 74.557 4 (midden) Lithic Quartz flakes (primary) 2 0.047 Lithic Quartzite flake 1 0.159 Lithic Rhyolite flake (tertiary) 1 0.11 Lithic Quartz shatter 1 0.39 Lithic Quartzite shatter 1 0.25 Shell Oyster 28 2.5 TOTAL STRATUM 4 34 3.456 5 Earthenware Mockley, net-impressed 2 0.379 Earthenware Mockley, smoothed 2 0.075 Earthenware Mockley, unidentified surface treatment 4 0.039 Lithic Quartz flake (secondary) 1 0.011 Lithic Quartzite flakes (primary) 5 0.58 Lithic Rhyolite flakes (tertiary) 3 0.04 Lithic Quartzite fire-cracked rock 2 0.313 Lithic Quartzite core 1 2.234 Charcoal 9 0.063 Shell Oyster 1 0.493 TOTAL STRATUM 5 30 4.227 6 Earthenware Mockley, unidentified surface treatment 1 0.375 Lithic Quartzite, thermally altered 4 0.066 Lithic Quartzite, fire-cracked rock 11 12.56 Charcoal 3 0.005 Glass Clear, flat 1 0.004 TOTAL STRATUM 6 20 13.01

114

The remainder of the midden was excavated as Strata 3 and 4. These layers within the midden were separated based on density of shell; strat 3 was far denser than strat 4. In sum, 82 whole oyster shells were recovered from the midden, in addition to a number of pieces of quartz, quartzite, and rhyolite debitage. Two pieces of FCR were recovered from Strat 3, while two quartz flakes, one rhyolite flake, one quartzite flake, one quartzite shatter, and one quartz shatter came from Strat 4. In contrast to the northern midden, no ceramic sherds were recovered from this midden. The soils in Strat 3 consisted of a 10YR 5/1 very dark grayish brown sandy loam while the soils in Strat 4 were mottled with a 10YR 5/1 very dark grayish brown sandy loam and a 10YR 5/6 yellowish brown coarse sand.

Strat 4 was stopped at the base of the midden on the western side of the unit (Figure 55). The archaeologists were unsure if they had reached subsoil, so they began to excavate this as an arbitrary level. In the end, this turned out to be another cultural horizon, plunging nearly a foot and a half below the base of the midden. While no ceramic sherds were recovered from the midden itself, a number were recovered from below and just to the exterior (east) of it. A total of nine shell-tempered Mockley sherds were recovered, all with the oyster shell temper leached out (Figure 56). This was in addition to charcoal and oyster shell, 17 quartzite fire-cracked rocks, a quartzite core fragment, five quartzite flakes, one quartz flake, and three rhyolite flakes (see Table 11). Oddly enough, one clear glass fragment was recovered from near the base of excavation. The excavator noted this probably came from wall scrapings. While there was some variety in the soil color and density at deeper levels, it was generally a 10YR 6/4 light yellowish brown sand at the top and a 10YR 5/6 yellowish brown sand at the bottom.

115

Figure 55: Unit 5, after excavation of shell midden (Strata 3 and 4), visible in north and west walls; facing north

Figure 56: Ceramic assemblage from Unit 5, Stratum 5 at 18AN1285; most were recovered from just to exterior (east) of midden 116

The excavators noted that most of the artifacts were recovered from the eastern portion of the unit, which would have fallen just to the outer edge of the shell midden. Similarly, a number of ceramics were also found on the exterior of the midden in the northern excavation block. This spatial comparison seems to suggest some sort of domestic activity taking place at the edge of both middens.

An arbitrary level was excavated at the base of the unit, and after about 0.2 ft. with no additional artifacts and increasing clay content (a 10YR 5/4 yellowish brown sandy clay), excavation was halted (Figures 57 and 58).

117

Figure 57: North (top) and west (bottom) wall profiles of Unit 5 at 18AN1285

118

Figure 58: Profiles from Unit 5 at 18AN1285 Unit 5 Summary

A shell midden was partially excavated near the southern tip of the peninsula on which site 18AN1285 is located. Unit 5 was strategically placed on the eastern edge of this midden, (the same one encountered in Unit 3 during the 2006 excavations) to more fully explore the interface between midden and inland deposits.

No ceramic sherds were recovered from within the midden, in contrast to the hundreds of sherds recovered from within the northern midden. Rather, all of the ceramics came from either above the midden in the upper organic layer, or just to the eastern edge of the midden. The ceramics from above the midden were sand-tempered, possibly suggesting a later ware type deposited during a Late Woodland visitation of the site, while all of those that came from midden’s edge were shell-tempered, most likely Mockley ware. A number of quartz, quartzite, and rhyolite flakes also came from this area, suggesting tool manufacture was taking place in this

119

area. And the 17 quartzite fire-cracked rocks strongly suggest a fire pit located near the edge of the midden, although no particularly burned areas were noted during excavation.

The high number of artifacts located at midden’s edge is a trend that was also seen in the excavation block to the north. A number of ceramic sherds were recovered from the eastern edge of the shell midden in that area, as well. The southern midden was far thicker than its northern equivalent, measuring nearly a foot deep in some places. This suggests more shell harvesting and processing was probably taking place near the tip of the peninsula, which stands to reason, as it is closer to the shoreline on three sides.

Unit 6

Unit 6 was placed in the far northern extent of site 18AN1285 because of a magnetic anomaly detected during the magnetometer survey (see Figure 39). Very few artifacts were recovered from this unit, and no features were identified (Table 12).

Table 12: Total Assemblage from Unit 6 at 18AN1285

Stratum Material Type Count Weight (in ounces) 1 Earthenware Whiteware 1 0.525 Lithic Quartz flake (tertiary) 1 0.012 Bone Avian 1 0.021 Charcoal 47 0.132 Shell Oyster 1 0.016 TOTAL STRATUM 1 51 0.706 2 Porcelain Hard-paste 1 0.007 Charcoal 14 0.02 Shell Oyster 10 3.0 TOTAL STRATUM 2 25 3.027 3 Lithic Quartz flakes (tertiary) 2 0.935 Charcoal 5 0.022 TOTAL STRATUM 3 7 0.957 4 Lithic Quartz flake (tertiary) 1 0.013 TOTAL STRATUM 4 1 0.013

The unit was excavated by natural layers into strata 1, 2, 3, and 4. Stratum 1, the upper humus, consisted of a 2.5Y 5/3 light olive brown sandy loam and produced one quartz flake, one bird bone, one oyster shell, one whiteware sherd, and 47 charcoal fragments (weight = 0.132oz).

120

Stratum 2, the A horizon, consisted of a 2.5Y 6/6 olive yellow loose sand and produced ten oyster shells, one porcelain sherd, and some charcoal. Stratum 3 was a transition layer between A and B (a possible E horizon?), and consisted of a 2.5Y 5/4 light olive brown loose sand and produced a few charcoal flecks and two nice quartz flakes from near the bottom. Stratum 4 seemed to be the B horizon, and only one quartz flake was recovered from near the top of the horizon. The B consisted of a dense 10YR 4/4 dark yellowish brown slightly sandy dense clay, and excavations were halted.

Unit 6 Summary

Considering no features or particularly notable artifacts came from Unit 6, the strong magnetioc anomaly is unexplained. Metal or other magnetically active trash could have been under the leaf mat in the vicinity, taking into account the few historic artifacts that were recovered from the unit. Judging by the profile of this unit and the presence of a possible E horizon, is seems likely that this portion of site 18AN1285 is unplowed.

No prehistoric ceramics were recovered from the unit, but four tertiary quartz flakes were found. This suggests the Native inhabitants of the site were utilizing this portion of the site at least in some limited capacity for tool manufacture. The historic artifacts found here are likely related to the occupation at nearby Wrighton Farm, site 18AN424, but considering the low density, they probably represent little more than casual discard.

Site 18AN1285 Discussion

The Camp Letts site (18AN1285) is interpreted as a seasonal camp site with the occupation ranging from the Selby Bay phase of the Middle Woodland time period (ca. A.D. 200 – A.D. 900) to the Potomac Creek phase of the Late Woodland period (ca. A.D. 1300 – A.D. 1700). Judging by the high percentage of Mockley sherds and the presence of a contemporaneous Selby Bay knife, it appears that the heaviest period of site use occurred in the Middle Woodland period. With only three small sherds definitively identified as Potomac Creek, it would seem the later occupation of the peninsula was slight and periodic.

121

A total of 1,699 artifacts were recovered from three years of excavations at the site. Of these, only 37 were historic artifacts. This included refined white earthenwares, porcelain, vessel glass, nails, slag, brick, and daub (Figure 59). It should be noted that of the 21 brick and daub fragments recovered, Lost Towns Project laboratory specialists could not be certain that many were actually bricks or simply a poorly fired prehistoric ceramic.

Figure 59: Total historic artifacts recovered from 18AN1285 These historic period artifacts are very sparse across the site and no historic features were identified. As such, it is highly unlikely that there was occupation on this peninsula after European contact. These few artifacts represent causal discard from nearby Wrighton Farm (18AN424) and from sporadic use of the land by people utilizing nearby Bear Neck Creek and the Rhode River.

The remaining 1,662 prehistoric artifacts can be generally classified into four broad categories: ceramics, lithics, faunal remains (shell and bone), and floral remains (charcoal). Table 13 provides a breakdown of these four categories.

122

Table 13: Total Prehistoric Artifacts Recovered from 18AN1285

Type Number Weight (in ounces) Bone 16 0.477 Lithic 98 84.9986 Charcoal 301 1.8875 Ceramic 495 19.936 Shell 751 531.456 TOTAL 1661 638.7551

By far the highest percentage of ceramics recovered from the site was shell-tempered, and for those that could be identified, nearly all were Mockley. Figure 60 presents the breakdown of the ceramic assemblage graphically.

Figure 60: Total prehistoric ceramics recovered from 18AN1285

In sum, a total of 94% (n=464) of the prehistoric ceramics recovered from the site were shell-tempered or Mockley. Adding those categorized as Townsend/Rappahannock ware to that total, classified as a Late Woodland shell-tempered ceramic that was a general progression of Mockley ware, the total shell-tempered percentage rises to 97%. The sand-tempered and Potomac Creek wares only account for about 3% of the total ceramic assemblage. The color of the sherds from the assemblage varied between gray, black, red, and orange.

123

Only about 8% of the sherds had visible surface treatment. Of these, the majority (n=53) were cord-marked, including one Potomac Creek sherd. However, seven sherds of Mockley ware were identified as fabric-impressed and six had net-impressions.

The curved fracture lines on many of the Mockley ceramics strongly suggests the vessels represented by these sherds were coil-constructed. The most obvious example of attempted coil- construction is seen on the partially fired ceramic sherd still adhered to an oyster shell, recovered from the shell midden in Unit 8.

The high percentage of shell-tempered wares, coupled with the 264 sherds of friable, crumbly, apparently failed shell-tempered sherds recovered from the midden in Unit 8, could suggest that some of the other Mockley sherds seen here could represent successful pottery manufacture that occurred on site. Regardless, the overwhelming percentage of Mockley sherds demonstrates the highest period of occupation during the later portion of the Middle Woodland time period of prehistory. The three Potomac Creek sherds (one cord-marked) illustrates that this peninsula was at least a moderately desirable location for later Woodland period Native Americans, as well.

A wide variety of lithics were recovered from Camp Letts, with the highest percentage being locally available materials (Figure 61). When based on artifact count, quartzite, quartz, ironstone, and sandstone combined account for 84% (n=75) of the total number of lithics recovered, while non-native chert and rhyolite only account for 16% (n=14) of the assemblage. When based on artifact weight, the percentage of native materials recovered from the site jumps to 90% (n=72.879 oz), while the non-native materials account for only 10% (n=8.367 oz) of the total assemblage.

124

Figure 61: Total lithics recovered from 18AN1285

It appears some tool manufacture was also taking place on the site. The percentages of lithic types seen in the assemblage of debitage (here defined as cores, flakes, chunks, and shatter) are consistent with the overall lithic assemblage, with quartz and quartzite making up the largest percentage (Table 14). A quartz Selby Bay knife and an undiagnostic rhyolite projectile point were the only two completed tools found on site. Of the flakes that could be identified, seven were primary (more than 50% cortex remains), four were secondary (less than 50% cortex remains), and 12 were tertiary (no cortex remains), indicating that lithic reduction was taking place at all stages of the process.

125

Table 14: Total Debitage and Tools from 18AN1285

Material Object Notes Count Weight Unit Stratum Quartz Knife Selby Bay 1 0.947 2 2 Quartz Chunk 4 0.946 2 2 Quartz Core 1 3.43 STP 70 Quartz Flake Unidentified 1 0.116 2 1 Quartz Flake Unidentified 4 0.138 2 3 Quartz Flake Primary 2 0.047 5 4 Quartz Flake Secondary 1 0.044 4 2 Quartz Flake Secondary 1 0.011 5 5 Quartz Flake Tertiary 1 0.012 6 1 Quartz Flake Tertiary 2 0.935 6 3 Quartz Flake Tertiary 1 0.013 6 4 Quartz Shatter 1 0.39 5 4 Quartz Shatter Primary 1 0.062 7 4 Quartz Shatter 1 0.028 8 3 TOTAL QUARTZ 22 7.119 Quartzite Chunk 1 0.036 STP 66 Quartzite Chunk 1 0.203 1 3 Quartzite Core 1 5.44 1 3 Quartzite Core 1 2.234 5 5 Quartzite Flake Primary 5 0.58 5 5 Quartzite Flake Tertiary – Thermally altered 4 0.066 5 6 Quartzite Flake Unidentified – possible scraper 1 0.159 5 4 Quartzite Shatter 1 0.25 5 4 Quartzite Shatter 1 0.2 8 2 TOTAL QUARTZITE 16 9.168 Rhyolite Point Fragment – base missing 1 0.17 1 1 Rhyolite Flake Unidentified 2 0.052 STP 66 Rhyolite Flake Tertiary 1 0.11 5 4 Rhyolite Flake Tertiary 3 0.04 5 5 TOTAL RHYOLITE 7 0.372 Chert Chunk Thermally altered 1 0.676 STP 71 Chert Flake Secondary 1 0.008 4 1 Chert Flake Secondary 1 0.085 4 1 TOTAL CHERT 3 0.769

126

Spatially, the highest number of flakes (n=20) from a single unit was recovered from the strat excavated just outside of the midden in Unit 5, located in the far southern end of the site near the point of the peninsula. Six pieces of debitage came from the Unit 4/7/8 block, but only one of these, a quartz chunk, was actually recovered from the midden itself. This seems to suggest lithic reduction was one of the activities that were taking place at midden’s edge.

The total count of 751 shells in the inventory is somewhat misleading. Considering the thousands of oyster shells encountered while excavating the two middens, in general, only whole oyster shells with “female” hinges were collected and taken back to the lab for processing; however, all shells that exhibited an encrustation of orange sediment were saved from the northern midden. And all shells other than oyster (including 30 snail shells and three freshwater periwinkle shells) were kept.

Only 16 bones were found at Camp Letts. This includes one bird bone from Unit 6 (possibly modern), three fragments of a charred mammal long bone from STP 66, located near Unit 1, one very small unidentified bone from the northern shell midden in Unit 7, and 11 burned mammal bones from Feature 1. The paucity of bones recovered from the site suggests one of two things: either excavations did not intersect with an area where the Native populations were discarding bones from their hunts, or the intervening centuries have caused all evidence of this type of trash to disintegrate.

When analyzing the artifact assemblage spatially, it is clear that more activity was taking place near the center of the peninsula (Figure 62). Granted, more excavation took place in this area (Units 1, 2, 4, 7, and 8) than closer to the point (Units 3 and 5), but a far greater number and variety of artifacts were seen closer to the interior. When shell is disregarded, 705 artifacts (weight = 64.459 oz) came from the units in the center of the landform, while only 90 (28.7 oz) were recovered closer to the point. Seventy-five prehistoric artifacts (weighing 1.727 oz) were recovered from the far northern Unit 6, but 66 of these were charcoal bits (Table 15; Figure 63).

127

260 260

27 240 7.5 240 26 7 25 24 6.5 23 220 220 6 22 21 5.5 20 19 5 200 200 18 4.5 17 16 4 15 180 3.5 180 14 13 3 12 2.5 11 160 160 10 2 9 1.5 8 7 140 1 140 6 5 0.5 4 0 3 120 120 2 -0.5 1 -1 0

100 Ounces of Ceramic 100 Ounces of Lithics

80 80

60 60

220 240 220 240

Figure 62: Distribution maps from 18AN1285

128

Table 15: Comparison of Prehistoric Artifacts from Three Excavation Areas at 18AN1285

Central Units Southern Units Northern Unit (1, 2, 4, 7, 8) (3, 5) (6) Count Weight (in oz) Count Weight (in oz) Count Weight (in oz) Bone 12 0.209 0 0 1 0.021 Shell 389 276.698 210 164.967 11 3.016 Charcoal 179 1.472 38 0.21 66 0.174 Ceramic 462 16.066 14 1.908 0 0 Lithic 47 48.889 36 26.338 8 1.532 TOTAL 1089 343.334 298 193.423 86 4.743

Figure 63: Comparison of prehistoric artifacts from three excavation areas at 18AN1285

The majority of the artifacts from the central units were ceramic sherds (n=462; 16.066 oz), while only 14 sherds were recovered from the southern units (1.908 oz). This highly skewed number is due in part to the high number of ceramic wasters found in the midden of Unit 8 from the failed attempt at pottery manufacture.

129

Lithics (including debitage and fire-cracked rock) seem to be more evenly distributed between the two areas (n=47, 48.889 oz in the center; n=36, 26.338 in the south), even though the only tools recovered from the site came from the center. However, more debitage came from the southern units (n=20) than the central units (n=17). Four quartz flakes were also recovered from the far northern Unit 6. This suggests that lithic reduction was taking place all across the peninsula, even though most of the activity was concentrated in the center.

Site 18AN1285 Summary

The Camp Letts site appears to have been occupied periodically and seasonally, probably by semi-nomadic bands of people. The heaviest period of site occupation seems to have occurred during the Selby Bay, or Mockley, phase of the Middle Woodland time period (ca A.D. 200-A.D. 800). The people who inhabited this area thousands of years ago were mostly interested in exploiting the rich natural resources of the Rhode River. The most overt remains of their time on this small peninsula that juts into Bear Neck Creek are the thousands of discarded oyster shells resulting in two middens. Oysters were an abundant resource in the area, and the meat could be smoked and preserved for transportation back to the larger base camp.

The native peoples were also engaged in rudimentary attempts at pottery manufacture during their stay at site 18AN1285. Over 260 sherds of crumbly, friable, shell-tempered bits of a material apparently mid-way between clay and pottery were recovered from the central shell midden. When exposed to water, these sherds do not revert back to clay, as they would if they were never fired, nor are they impervious to water, as they would if they had been fully and successfully fired. Rather, they absorb the water and continue to crumble apart. This, coupled with intentional shell-tempering in the sherds, strongly suggests there was an effort at producing ceramic vessels here on the peninsula (Michael Stewart, personal communication, 2008). A single sherd of this material that is still adhered to an oyster shell from the midden shows three, and perhaps four, coils, clearly demonstrating an attempt at coil-constructing a vessel. All of these sherds were recovered from the portion of the midden excavated in a single unit, possibly suggesting they were deposited during a single event. It is easy to imagine a person confronted with an unsuccessful pot tossing soft coils into the trash heap.

130

Two small pit-like features were excavated that were located at either edge of this central midden. Feature 1 was the deeper of the two, diving nearly a foot into subsoil. This small pit was lined with sterile, fire-reddened soil, and contained burned bone, fire-cracked rock, and both burned and unburned shell-tempered ceramics. The burned ceramics showed signs of charring on the interior and exterior, as well as the broken edges, suggesting a vessel broke and continued to char in the pit (see Figure 47). The pit later filled with oyster shells, possibly a result of the nearby midden slumping and settling. The second feature, called Feature 2, was located on the opposite side of the shell midden. It was very shallow (less than two inches deep), circular, and small (only about six inches in diameter). The only artifacts recovered from this feature were five sherds of net-impressed Mockley ceramics. The use or function of either of these pits is unconfirmed, but given their proximity to the midden, it stands to reason their purpose was directly related to the processing of oysters.

A second shell midden was tested during the 2006 and 2007 field season. This midden was located in the far southern end of the site, close to the point of the peninsula. Few artifacts were recovered from within this midden, and no evidence of the soft, friable ceramics were found anywhere in this area. However, a number of fire-cracked rocks and Mockley ceramics recovered from just outside the midden speak to activities taking place at midden’s edge. The highest number of debitage fragments recovered anywhere on site, including quartz, quartzite, and rhyolite cores and flakes were recovered from this area. This, coupled with the thicker and denser shell midden, seems to suggest a greater degree of oyster processing and tool manufacture on this side of the site, although no features were noted in the vicinity.

A single unit was excavated in the far northern portion of the site, and very few prehistoric artifacts were recovered from this area. While no prehistoric ceramics were found, four tertiary quartz flakes were recovered, which clearly indicates tool manufacturing across all of site 18AN1285. This unit did possess very dense, clay-rich subsoil only about one foot below the ground surface. Such clay could provide a source for the raw materials needed for the attempted pottery manufacture seen on the site.

The high percentage of locally available lithics found on the Camp Letts site and relatively sparse rhyolite is notable for a site of this time period. Stewart (1984) demonstrated 131

that rhyolite increases proportionately in assemblages located 80 to more than 100 miles from the sources in Western Maryland, demonstrating that this material was important in structuring intergroup relations during the Selby Bay phase of the Middle Woodland time period. Sites excavated in the area have long corroborated this theory. Jim Gibb (1997) reported a lithic assemblage that consisted of nearly 50% rhyolite at the contemporaneous Smithsonian Pier site (18AN284), located just south of Camp Letts on the Rhode River. Hettie Ballwebber (1994) reported almost 93% rhyolite in the assemblage of the nearby Luce Creek site (18AN143), a single component Selby Bay period shell midden. Here at site 18AN1285, rhyolite only accounts for 8% of the total lithic assemblage. One of only two completed tools recovered from the site was a heavily curated rhyolite point, with the other being a quartz Selby Bay knife. This profound scarcity of rhyolite on the site can possibly be explained by greater care and conservation taken with the material, given its difficulty in acquiring. Alternatively, the residents of this site could have simply been here on such a temporary basis that they were bringing little with them and mainly using the resources available locally.

Whatever the case, the Camp Letts site provides an interesting glimpse into the rich and diverse prehistoric past of Anne Arundel County. Evidence of pottery manufacture (even failed attempts) is rarely seen in the archaeological record, and the relative dearth of rhyolite found on site demonstrates that there is much still to learn about the people of the Selby Bay phase and the variations between sites representing this time period. Further exploration of the presence of rhyolite on Middle Woodland sites may offer new perspectives on how Native populations viewed, used, and lived at different site types, resulting in a refinement of site types for the period. Such analysis might also further refine temporal controls within the Middle Woodland period.

A notable lesson regarding excavation strategies at discrete seasonal oyster midden sites was learned from 18AN1285. Valuable data was recovered from the margins of the more highly visible oyster middens. Deposits at the periphery of the oyster midden features yielded significant information about the seasonal activities beyond the obvious procurement of oysters. While this was their primary activity while on the site, a full range of daily activities no doubt took place, yet such activities left only an ephemeral signature on the landscape. Excavation

132

strategies for similar small, temporary or seasonal camps should search out the ‘quieter’ cultural signatures, which may mean excavation at the margins of the obvious core of the site. As the knowledge base of the time period continues to grow, it seems clear that the findings from site 18AN1285 provide yet another important, and somewhat unexpected, voice to the continued discussion.

133

CHAPTER THREE: The Future of Cultural Resources on the Rhode River

This three-year effort has attempted to address broad questions and provide guidance for the effective stewardship of dozens of resources in the area. From the initial evaluation stage, to more in-depth study of select sites, much has been learned, both academically, and in a planning and management context. We have verified that the two most significant threats to sites in the area are development, and natural forces. Development can be managed through regulatory review, and enhanced through education programs and promoting citizen and landowner stewardship. Natural forces must be closely monitored, and when threatening the integrity of site, emergency salvage and study is warranted. While these are useful general strategies for managing the broad range of cultural resources in the entire Rhode River region, below are specific strategies for future protection, scholarly investigation, and long-term maintenance of the two sites discussed within this report.

Management Plan for 18AN339

This site is clearly a very significant resource and while there has been a Determination of Eligibility form completed, and the SHPO has concurred with the determination, the long- term study and stewardship of the site would benefit from the completion and submittal of a formal National Register nomination.

The site boundaries for 18AN339 encompass a broad continuum of resources, and were initially defined based upon archaeological investigations, archival documentation, and extant landscape features. One of the most significant aspects of this multi-component site is its intact historic landscape, which is very visually stimulating. Landscape features, such as topography, view shed, roads ad pathways, vegetation, are subtle, yet impart the sites significance to the public with minimal interpretation. Future cooperative efforts between the Lost Towns Project and SERC should include developing strategies for the preservation and interpretation of the sites visible archaeological features within a context of public access and interpretation.

Continued investigations of the 17th century loci, and continued discovery and evaluation of the 18th and 19th century features across the site are warranted, particularly as SERC looks toward making the site a centerpiece of their interpretive program in the coming years. Clearly, 134

the mansion ruins, which are deteriorating rapidly, deserve further study and stabilization. SERC has begun planning for active stabilization efforts, and hopefully, a systematic documentation project of the ruins will be included in those efforts.

Perhaps the most impressive thing about the 18AN339 site is the surrounding landscape, much of which has not been systematically surveyed. Initial evaluation has identified other sites on the 600+- acre parcel, and undoubtedly, there are copious significant resources that survive elsewhere on the parcel. Features and related sites outside the surveyed area should be sought, and further survey efforts should be mounted to identify resources “off the hilltop,” particularly undiscovered domestic and agricultural sites associated with the enslaved population that dominated the site throughout its colonial, Federal, ante- and post-bellum history.

The research potential at this site is enormous and its value heightened even more as it provides a public venue for the study and interpretation of a rich multi-component site. Such public access provides opportunities for education and for promoting stewardship of sites throughout the region. Most significantly, the sites’ recent incorporation into the SERC-owned land reunites a portion of Thomas Sparrow’s 17th-century parcel to its former bounds, where it has been disassociated since 1859. The site should continue to be studied and considered holistically, as part of a larger historic context of the region, as a part of a large inter-related tobacco plantation, and as a element of the current landscape and natural environment.

Management Plan for 18AN1285

As the YMCA has owned the camp property since 1902, and the land is under no immediate or known threat for sale or development, site 18AN1285 is well-protected. Based upon or findings, the site is of interest to the study of the Middle Woodland period, and should be preserved until such time that future investigation is warranted. The site is in a sensitive environment, along the highly erodible shoreline, thus we would recommend regular monitoring of the site and nearby shore for environmental degradation.

As a component of this project, we have been in close contact with the Camp Lett’s property managers, and hopefully have instilled an appreciation for the resources of which they

135

are stewards. Should circumstances allow, we would welcome the opportunity to offer additional educational programming in coordination with their ongoing camp archaeology program.

From an academic perspective, further consideration of the pottery-making component may be of great interest. Acquiring clay samples from the region, and conducting clay-sourcing studies would allow archaeologists to further explore the pottery making revelation. This site should be incorporated as a key representative site type within a multi-property National Register nomination on the Middle Woodland Phase in Anne Arundel County, MD. This site is ideally suited for preservation in situ and under Camp Letts stewardship should be well protected for future generations.

The Rhode River Region

As has been discussed in detail throughout Volume I and II of this series, the Rhode River drainage provides a unique microcosm of cultural resources that are representative of both prehistoric and historic resources throughout Anne Arundel County. There are varied threats and pressures on the sites throughout the watershed, such as development and natural forces. With the recent acquisition of the Kirkpatrick-Howat farm by SERC in the spring of 2008, more than one-third of the sites within the original study area are now owned, and under the thoughtful stewardship of a conservation-minded organization. SERC has expressed a clear commitment to partnering with cultural resources professionals to ensure a management strategy is employed that will both preserve, study, and educate the public on the unique synergy between historic land use patterns and modern day environmental stewardship.

The northern shores of the Rhode River has seen intensifying development and shoreline impact, yet on the whole, with open space uses, such as the Camp Letts acreage, and the Anne Arundel County-owned Beverly-Triton Park, the Rhode River area retains much of its historic and archaeological integrity as a region.

It is inevitable that some resources will be impacted or lost as time progresses, yet with a comprehensive system of regulatory review, monitoring, voluntary stewardship, and in some cases data recovery, the information these sites can yield will not be lost. Environmental degradation is a more difficult threat to manage, as sea level rises, and erosional actions continue

136

to impact the shorelines, where many of the Rhode River sites are found. There are only limited management strategies one can employ to mitigate these effects, though one useful strategy is to regularly monitor the condition of sites, and conduct emergency investigations should environmental conditions threaten the loss of site integrity. This innovative and comprehensive approach to developing a cultural resources management plan has been most successful, and has provided opportunity for studying a region from multiple perspectives. This three-year effort has already paid off as SERC has the tools at hand to effectively manage many of the Rhodes most significant sites. This planning and research effort will ensure that a majority of the archaeological resources in the region will continue to contribute to our broader understanding of the prehistory ad history of the Rhode River.

137

REFERENCES CITED

Ballweber, Hettie L. 1990 Preliminary Archaeological Reconnaissance of the Java History Trail, Smithsonian Environmental Research Center, Anne Arundel County, Maryland. Report to The Smithsonian Environmental Research Center, Edgewater, Maryland from Hettie Ballweber, Archaeological Consultant.

1994 Return to The Luce Site (18AN143). Maryland Archaeology 30(1): 1-16.

Cox, C. Jane, Lauren Franz, Erin Cullen, and Shawn Sharpe 2007a Survey and Limited Assessment of Archaeological Resources in the Rhode River Region, Anne Arundel County, Maryland. Report to the Maryland Historical Trust from The Lost Towns Project of Anne Arundel County.

Cox, C. Jane, Erin Cullen, Lauren Schiszik, Kelly Cooper, and Shawn Sharpe 2007b Assessment and Evaluation of Select Archaeological Resources in the Rhode River Region: Anne Arundel County, Maryland. Report to the Maryland Historical Trust from The Lost Towns Project of Anne Arundel County.

Dent, Richard J., Jr. 1995 Chesapeake Prehistory: Old Traditions, New Directions. Plenum Press, New York.

Gibb, James G. and Anson H. Hines 1997 Phase III Data Recovery at the Smithsonian Pier Site (18AN284/285), Smithsonian Environmental Research Center, Anne Arundel County, Maryland. Report to The Smithsonian Environmental Research Center, Edgewater, Maryland from James G. Gibb, Archaeological Consultant.

Gilsen, Leland 1978 Population Adaptation to the Chesapeake Bay: Estuarine Efficiency. Maryland Archaeology. 14(1-2):11-16.

Greenburg, Laurie and Amy Hyatt 1990 “Appendix I: History of the Smithsonian Environmental Research Center.” In Preliminary Archaeological Reconnaissance of the Java History Trail, Smithsonian Environmental Research Center, Anne Arundel County, Maryland by Hettie L. Ballweber.

Hanna, Susan D., Barry Knight, and Geoff Egan 1992 Marked Window Leads from North America and Europe. Historic St. Mary’s City, Maryland.

Hranicky, Wm Jack and Floyd Painter 1989 A Guide to the Identification of Virginia Projectile Points. Special Publication Number 17, Archaeological Society of Virginia, Richmond.

138

Lee, Byron A. 2004 John Contee and Java Plantation, Part II. Anne Arundel County History Notes 35(4):1-2, 7-9.

Luckenbach, Al and C. Jane Cox 2003 17th Century Lead Cloth Seals from Anne Arundel County, Maryland. Maryland Archaeology 39(1&2):17-26.

Noel Hume, Ivor 1969 A Guide to Artifacts of Colonial America. University of Philadelphia Press, Philadelphia.

Maryland Archaeological Conservation Lab (MAC Lab) 2008 Mockley Ceramics Details, Electronic Document, http://www.jefpat.org/diagnostic/Prehistoric_Ceramic_Web_Page/Prehistoric%20Ware%20Desc riptions/Mockley.htm, Accessed April 13, 2008.

Paschal, Herbert R. Jr. 1955 A History of Colonial Bath. Edwards & Broughton, Raleigh, North Carolina.

Potter, Stephen R. 1993 Commoners, Tribute, and Chiefs: The Development of Algonquian Culture in the Potomac Valley. The University Press of Virginia, Charlottesville.

Ruck, January M. 2008 Reintegrating Public History & Environmental Education: Preservation and Interpretation of the Ruin at Java Plantation, Edgewater, Maryland. Master’s thesis, School of Architecture, Planning and Preservation, University of Maryland, College Park.

Russell, Donna Valley ND First Families of Anne Arundel County, Maryland, 1649-1658: Volume 2, The Headrights. Catoctin Press, New Market, Maryland.

Schindler, William, III 2006 Middle Woodland Exploitation of Migratory Fish in the Delaware Valley. Unpublished Ph.D. dissertation, Department of Anthropology, Temple University.

South, Stanley A. 1977 Method and Theory in Historical Archaeology. Academic Press, New York.

Sparrow, Margaret W. 1990 The Sparrows of Sparrow’s Point. Maryland Historical Magazine 85:395-404.

Stewart, R. Michael

139

1984 Archaeologically Significant Characteristics of Maryland and Pennsylvania Metarhyolites. Prehistoric Lithic Exchange System in the Middle Atlantic. Jay F. Custer, editor, pp 1-13. University of Delaware Center for Archaeological Research Monograph Number 3.

Trostle, Michael F. 1981 Mount Clare: Being an Account of the Seat built by Charles Carroll, Barrister, upon his Lands at Patapsco. The National Society of Colonial Dames of America in the State of Maryland, Baltimore.

140

Appendix One: Artifact Catalogs for 18AN339 and 18AN1285

141

Appendix Two: Staff Qualifications

142

Appendix Three: Revised Site Forms for 18AN339 and 18AN1285

143

Appendix Four: Revised Determination of Eligibility Forms for 18AN339 and 18AN1285

144