SEAC: Featured Project
  • 3D Rendering of Shiloh Mound

    Southeast Archeological Center

    Cultural Resources National Park Service

SEAC: The Water Island Project

Archaeology in the Caribbean: The Water Island Archaeological Project¹

David G. Anderson
Southeast Archeological Center
National Park Service

List of Illustrations


From May through August 1998, extensive archaeological investigations were undertaken on Water Island, U.S. Virgin Islands (Slide WI-72), by a research team from the Southeast Archeological Center, National Park Service (SEAC). Analysis of the information collected is underway, toward the goal of producing detailed technical and popular reports on the research. The technical report, to be entitled "The Archaeology and History of Water Island," will be submitted to the Office of Insular Affairs and the Virgin Islands State Historic Preservation Office in draft form on 1 August, 2000. After review the technical report will be produced in final form, and used to produce a more popular volume, that should be released in 2001.

The fieldwork was undertaken because the federal government, which owns the island, is transferring it to the territorial government or to private ownership, action that would effect cultural resources. An earlier survey of the island, conducted by SEAC in 1992 (Wild and Anderson 1992), together with a brief follow-up visit in 1996, located a number of historic and prehistoric sites, including the remains of two major plantations, a smaller isolated structure, and an extensive World War Two fortification complex. Many of the ruins were in an exceptional state of preservation, and were superbly documented in the historic record, prompting the research program described here.

Project Location and Environmental Conditions

Water Island is located a half mile off the south coast of St. Thomas in the U.S. Virgin Islands, near Charlotte Amalie harbor (Slide WI-259). The island, which is volcanic in origin, is irregular in shape, covers 491.5 acres, and is approximately two miles long by a mile wide. The island is characterized by steep rocky slopes (Slide WI-173), a pronounced central ridgeline almost 300 feet high, and a highly indented coastline with numerous narrow bays and beaches (Slide WI-136). There are no permanent streams, but potable, if not very palatable, water is present in brackish ponds near the coast (Slide WI-194). Vegetation is semi-arid and ranges from dry tropical thorn scrub and cactus forests in the interior to mangrove/salt pond ecosystems in several of the coves (Slide WI-1179). A wide range of birds and lizards are found on the island (Slide WI-935).

A Brief History of Water Island

Water Island took its name from the fresh water it supplied to early settlers. Unlike many places in the Eastern Caribbean, where rainfall is the primary source of drinking water, fresh water ponds were present on Water Island, making it an attractive place for early seafarers and visitors.

Permanent European colonization of the Virgin Islands dates from 1672. One of the first things the Danes who arrived on St. Thomas did was to stock Water Island with cattle and goats. This livestock belonged to the Danish West Indian Company and was used to feed the colonists, and was also sold to passing ships and neighboring colonies.

Project historian David Knight (Slide WI-379) has found numerous references dating to the first half of the 18th century documenting the occasional use of Water Island for a variety of tasks, including ship repair and provisioning. Knight is summarizing this information, with George Tyson, another Virgin Islands historian, as part of a detailed overview of Water Island history that will appear in the final report.

Settlement on Water Island is depicted on a great many historic maps. The earliest we have been able to find, the 1719 Gerard van Keulen map (van Keulen map Image), depicts two open circles on either side of a knoll on the northern end of the island, near what is called "Albert de Ruyters bay." The circles may indicate houses, water ponds, or wells; they are near where both later 18th and 19th century wells and ruins have been found as a result of the archaeological fieldwork. Settlement on this knoll, now known as Carolina Point, is shown on subsequent maps from the late 18th century onward. In 1819 this plantation was named "Carolina Lyst" by the then current owner, Baron Lucas de Bretton, Sr.

Evidence for people living on Water Island dates from 1710 onward, although until 1769 occupation was sporadic. The earliest occupant appears to have been a sailmaker, William Breitz who took up residence with his family about 1710. Upon his death Albert de Ruyter obtained the property, and although he died in 1719, his name was immortalized by the mapmaker van Keulen, and is still used to describe the bay below Carolina Point.

Several other early families made use of Water Island for varying lengths of time. In the 1730s Elias Baile was living on the island, together with a number of slaves, apparently growing cotton. Baile and a number of slaves died in a smallpox epidemic in 1735, and while the property was worked by slaves for a number of years following the epidemic, it was abandoned about 1740. The Sprat Bay structure site, which dates from the 1730s, may be Baile’s plantation. The project historian, David Knight, suggests that the site may have alternatively been occupied by someone overseeing limestone/coral and ballast rock mining operations that are also reported occurring on the island in the early 18th century.

A number of interesting anecdotes have been found in the early records, including a detailed account of the hunt for a runaway slave hiding on the island, the sinking of Danish West India Company barque, as well as more prosaic descriptions of brief visits by watercraft of all sizes.

Detailed annual records of property ownership and resident population exist for the island for the period from 1769 to 1914 (Appendices). Water Island first appears in the land tax records in 1769, when the accounts state that Jean Renaud, a free mulatto, owned "an island" on the south side of St. Thomas. Eighteen slaves were shown as resident in that year. Renaud was one of 106 free colored men capable of bearing arms on St. Thomas at the time and one of just two free colored planters.

A map prepared by Oxholm in 1778 (Oxholm Map Image) shows the Renaud settlement where de Ruyter's house had been depicted on earlier maps. Renaud's settlement consisted of a single great house flanked on the south by a slave village with nine houses. At the time Renaud lived alone in the great house, while 31 slaves inhabited the village.

The only other structure shown on the 1778 map was a small building just inland from the beach at Sand Bay. This building appears to be a small structure complex found overlooking Sprat Bay in 1992, although the fact its mapped location was the middle of a dense swamp caused the survey crew no end of grief before they found it on a nearby hill in 1992.

In 1793 another free man of color, Peter Tamaryn, took over Water Island from Renaud. Tamaryn was the son of Mingo Tamaryn, the captain of the Free Negro Corps formed by the Danes in 1721 to help curb slave unrest. Mingo and his Corps helped quell the infamous 1734 St. John's slave rebellion, vividly described in the book Night of the Silent Drums (Anderson 1992) and were also used to track down runaway slaves throughout the Danish West Indies. Peter took command of the Corps when his father died in 1765. He apparently lived on Water Island for only a few years in the early 19th century.

In 1799, a second plantation, La Providence, was created on Water Island, when Tamaryn transferred 60 Danish acres to Jean Regis Pourier, another free man of color. In the unusually detailed 1805 census, there were 16 slave houses extending over four acres, and most 19th century maps show at least four main buildings. The same 1805 census indicates there were 17 slave cabins at Carolina Lyst in a two acre area.

For the next 60 years, the island's two plantations were independently owned and operated and, like other St. Thomas plantations, both changed hands frequently. They are often shown on maps of the period, sometimes with individual buildings depicted (H. B. Hornbeck 1835-1839 map; 1851 British Admiralty Hydrographic Office Map, Imbry and Sons 1855 map). In the 1860s they were reunited as a single property by Joseph Daniel, who was a master craftsman, merchant, and collector of customs. As a member of the Colonial Council during the 1870s, Daniel belonged to a radical faction that campaigned for greater self-government for St. Thomas. For the past several years one of his direct descendants, Louise Daniel Hutchinson, has been corresponding with the project team about her family history.

The number of free colored proprietors and occupants of the plantations on Water Island is noteworthy, something that may have been facilitated, or perhaps tolerated, by the colony's dominant ethic group, in part, because the island was in a marginal geographical location. The island's historic assemblages thus offer unusual opportunities for comparison with materials from white-owned plantations of the period.

Eighteenth-century land tax records do not provide much specific data on land use on Water Island, although animal husbandry and cotton production appear to have been important. Starting in 1803, fortunately, the St. Thomas land tax records began specifying the amount of land used for "Sugar", "Other Crops", and "Unused". In the early years of the 19th century the primary crop appears to have been cotton, at least until the 1820s, when the slave population began to decline. Thereafter, ground provisions, such as tanyas, sweet potatoes, pigeon peas, and plantains constituted the only crops cultivated, for sale in Charlotte Amalie as well as for consumption by the island's inhabitants. Throughout the nineteenth century, animal husbandry continued to be important, with most of the island given over to grazing cows, sheep and goats.

Maximum population densities on both La Providence and Carolina's Lyst occurred between 1801 and 1821, the era of cotton cultivation. During this time the number of free and slave residents on La Providence's population ranged between 40 and 65, and on Carolina's Lyst between 30 and 50. The peak year was 1815, when there were 111 persons living on the island. The resident population of both properties dropped rather sharply during the 1820s and from 1828 onwards was seldom more than 10 people. Following emancipation in 1848, it is likely that the slave quarters were permanently abandoned. The last recorded occupation of Carolina's Lyst occurred in 1861, although a few people continued to reside at La Providence until 1905. Given the archaeological evidence for a major fire at Carolina Lyst in the mid-19th century, this likely accelerated it’s decline.

In 1905, Joseph Daniel’s heirs sold Water Island to the Danish East Asiatic Company for $21,000. In 1917 the U.S. Government acquired the Virgin Islands from Denmark, although Water Island continued to be owned by the East Asiatic Company until 1944, when the U.S. Government acquired it for $10,000 through condemnation proceedings. Prior to World War II people from St. Thomas made occasional use of Water Island to grow crops, maintain animals, and in at least one case, as recounted by a local informant, unsuccessfully search for treasure.

In the early 1940s, the U.S. Government established a coastal defense installation, Fort Segarra, that included barracks, gun emplacements, watch towers, and underground bunkers (Slide WI-232), as well as an infrastructure of docks, roads, water, sewage and power systems (Military Facilities on Water Island 1952 map). When World War II ended, construction was abruptly halted and the Army's Chemical Warfare Division used the island for tests for several years, activity that made the 1992 survey crew particularly attentive. In 1950, the Army turned the island over to the Department of the Interior, who shortly thereafter leased the island to a private developer for 40 years at an annual rent of $3,000.00. The expiration of this lease, and the decision to transfer ownership of the property out of government hands, prompted the archeological investigations conducted since 1992.

Cultural Resource Management Considerations

Archeological and historical investigations on Water Island are being conducted under the terms of a Memorandum of Agreement (MOA) signed in December 1997 by the Office of Insular Affairs (OIA), the Virgin Islands State Historic Preservation Officer (SHPO), and the Advisory Council on Historic Preservation (ACHP), with concurrence by the Southeast Archaeological Center, National Park Service (NPS), and the Sprat Bay Corporation.

The sites examined in 1998 had been located during an intensive archaeological survey of Water Island conducted by SEAC personnel in the fall of 1992. This work was undertaken by OIA in compliance with the National Historic Preservation Act, as part of the transfer process (Wild and Anderson 1992). Eleven archaeological sites and the Fort Segarra bunker complex were documented, and Fort Segarra and eight of the sites were determined to be eligible for inclusion on the National Register of Historic Places.

At the request of OIA, these sites were revisited in August 1996 by David Anderson and David Brewer, two SEAC archaeologists, to assess their condition following recent hurricanes. As a result of this reassessment, an MOA was prepared, following lengthy discussions between the funding agency (OIA), and the SHPO, ACHP, and NPS as to whether any further archaeological investigations were warranted. Basically, OIA felt no further work beyond that conducted during the 1992 survey was necessary, and that preservation covenants were sufficient mitigation to protect the sites once they had been transferred to private ownership. The SHPO, ACHP, and NPS disagreed, and a program of limited data recovery coupled with strict preservation covenants was agreed upon and incorporated into the MOA that was signed in late December, 1997. SEAC received authorization from the Office of Insular Affairs to proceed with archaeological and historical research on Water Island project on 19 February 1998.

Project Research Design

In March and April 1998, a detailed research design for the Water Island project was developed in consultation with the Virgin Island SHPO, under the terms of the MOA. The research design underwent internal review by project team members, and was approved by the NPS Southeast Regional Archeologist. It was then submitted to The Honorable Beulah Dalmida-Smith, Commissioner/SHPO for the U. S. Virgin Islands on 20 April 1998. A letter approving the research design was sent to NPS by The Honorable Beulah Dalmida-Smith on 4 May 1998. A copy of the final research design and SHPO letter of approval, and a schedule for the proposed fieldwork, was submitted to Mr. Edgar Johnson, OIA Virgin Islands Desk Officer, on 5 May 1998.

Concurrent with the development of the research design, SEAC personnel obtained housing, vehicles, and equipment for use on Water Island, and lined up a team of some 25 specialists to assist in the fieldwork and subsequent analysis and reporting activity. Although the logistics were complex, because we had known for some time that the project would occur, informal planning had been underway for several months.

Archeological Fieldwork

Over 900 person days have been spent in fieldwork to date on Water Island. The fall 1992 intensive survey encompassed 147 person days, with much of the time spent chopping through the dense and prickly underbrush. The survey team was advised and assisted by Territorial Archeologist Elizabeth Righter, local historian George Tyson, whose work was summarized above, and local historic architect Fred Gjessing (Slide WI-69). The 1996 revisit by Anderson and Brewer occupied just six person days, enough time to revisit every site as well as examine a few locations on cliff faces that had been missed in 1992.

The 1998 fieldwork took place over 12 weeks, from May 10th to August 4th. The 27 person team spent just over 700 person days in field and laboratory activity, with crew sizes ranging from 6 to 14, and averaging about 11 each day (several Sundays were also worked on a voluntary basis). In addition, another estimated 50 person days in volunteer labor was obtained from local residents. During the fieldwork the project was visited by SHPO staff under the direction of Ms. Claudette Lewis, and by Mr. Allen P. Stayman and Mr. Edgar Johnson of OIA. In addition, over the course of the summer some 150 private citizens from Water Island and St. Thomas came out to see what was happening, and a number served as volunteers for one or more days.

The sites that were examined in 1998 included:

(1) Carolina Point Plantation (12-VAM-3-209)
(2) Ruyter's Bay (12-VAM-3-21)
(3) Tamarind Tree Bay (12-VAM-3-9)
(4) Elephant Bay (12-VAM-3-22 and 12-VAM-2-23)
(5) Sprat Bay Structure (12-VAM-3-208)
(6) Banana Bay (12-VAM-3-56)
(7) Banana Bay South (12-VAM-3-210)
(8) Fort Segarra Military Complex (no territorial site number currently assigned)
(9) Providence Point Plantation (12-VAM-3-211)

Archaeological investigations were conducted at sites #1 through 7, while a limited program of mapping and photo documentation was conducted at the remaining two historic properties, as per the project research design. Table 1 lists the area that was excavated at each site (Table 1). In all, a total of 180.75 square meters of fill were examined, almost twice the 100 square meters proposed in the project research design.

The fieldwork was directed to documenting the location, condition, and contents of each site. Activity included the clearing of underbrush; transit mapping of all features, units, and site boundaries; the establishment of permanent datums (stainless steel rods surrounded by concrete); the GPS documentation of these datums and other site features; and the excavation of dispersed test and larger block units. The excavations proceeded using standardized 0.5 m or 1 m units. All fill was dry screened through 1/4 inch mesh, and a number of soil, wood, charcoal, daub, or other samples were taken from profiles and features as warranted. A set of fine screens were used to dry screen soils for paleosubsistence remains in the field. Soil samples were taken from a number of areas prior to backfilling, which occupied two to three hours each day the last week.

No human burials were found or excavated during the project. Two possible human teeth fragments were found at Carolina Point Plantation, one each in fill above the floor of two houses in the slave village area. Additional fragments may be found during the analysis. These will be examined by a physical anthropologist to determine whether they are indeed human, and if so, their final disposition to take place in consultation with the Virgin Islands SHPO.

A total of 1369 color slides, 1440 black and white photographs, as well as almost 20 hours of 8mm videofilm were taken during the 1998 fieldwork. These included shots of the excavations, features, and detailed architectural information at all of the project sites. Some of these will be posted on the Southeast Archeological Center’s project web site.

A note about project organization is in order. The project director was in the field throughout the project, and spent much of the time digging and as the photographer. Supervision of the work at individual sites or site areas was, however, typically delegated to various members of the project team, who are also assuming primary responsibility for the reporting of investigations at their sites. Thus, Steven Kidd and Greg Heide supervised work in the slave village at Carolina Point (Slide WI-578), and Steven Kidd (Slide Morse7) will use this material for his Master’s thesis. Greg Heide directed the work at Banana Bay South, Mike Russo at Banana Point shell midden and in Elephant Bay (Slide WI-1217), Brinnen Carter (Slide WI-647) at Ruyters and Tamarind Tree Bays, and David Anderson and Holy Righter (Slide WI-1108) at the Sprat Bay Structure and the Carolina Point Great House. Site directors were, of course, advised and assisted by the various members of the project team. While the project director reserved a veto power, in practice this was never exercised. The project proceeded with a remarkable degree of consensus about approaches.

A rather remarkable array of archaeologists were involved in the 1998 excavations (Slide WI-689) (Slide WI-1058). Due to the length of the project, a core paid staff of about half a dozen people were in the field at any one time. Steve Kidd, Greg Heide, and Dave Anderson saw the project through from start to finish, while Tiffanie Bourassa, Brinnen Carter, Angela Davis, Joshua Kehrberg, and Thaedra Palmer were there for over 50 days. The remaining 20 team members were volunteers, either avocational or professional archaeologists, who donated from two to six weeks of their time in exchange for airfare, room, and board. These people, listed at the end of this report, included a remarkable array of talent, and explain why so much was accomplished.

A summary of the work undertaken at each site is described below. Detailed maps of each site are currently being prepared by Greg Heide, making use of transit and GPS data collected in 1998.

Carolina Point Plantation (12-VAM-3-209)
[List of Illustrations]

Carolina Point is located at the northwest end of Water Island, on a point of land between and overlooking Tamarind Tree Bay and Ruyter's Bay (Slide WI-1626). Early historic maps depict water ponds and possibly wells in these bays, and a great house and a number of outbuildings on Carolina Point. The location offers a spectacular view of St. Thomas Harbor, and is exposed to breezes from several directions (see Slide WI-72, previously referenced). The Oxholm map of 1779 is the most detailed, and depicts the great house and nine smaller structures to the south that presumably represent a slave village. This is probable, as the owner, Jean Renaud, and 33 slaves are listed in residence in the 1779 tax records for this plantation.

A dense artifact scatter and the ruins of a colonial era great house, together with a number of outbuildings and slave/servants quarters, was found on the top and slopes of Carolina Point. Historic artifacts and masonry wells were also found in the low-lying coves to either side, Ruyter's Bay and Tamarind Tree Bay. Although the fieldwork conducted at each of these sites is described separately, all three combine to form a single complex.

The fieldwork at Carolina Point started with a massive program of brush clearing that took more than a week to accomplish (Slide WI-32). Once the area was cleared, an intensive mapping program was initiated, with over 1000 elevation points collected. This work was accomplished under the overall direction of Chris Clement, assisted at various times by Greg Heide, Dan F. Morse, and Brinnen Carter (Slide WI-273).

The ruins included intact masonry foundation walls and wall rubble from a number of discrete structures, including a great house, a bake oven with small buildings to either side, and downslope a small house-like structure and the remains of a number of slave cabins (Slide WI-727). Many of the structures that were later examined, particularly in the slave village area, were completely covered with dense vegetation, and were not evident until clearing had occurred. This dense vegetation has protected these remains, with the result that the site’s ruins and archaeological deposits are in a remarkable state of preservation. Dense concentrations of artifacts were found on the floors in many of the structures, suggesting they are very much as they were when they were abandoned roughly 150 years ago.

A total of 96 discrete 1x1 m units were opened at Carolina Point (Carolina Point Contour Map Image). Sixty were opened in and near the Great House at the top of the hill, 55 in the vicinity of the Great House and 5 downslope in an outbuilding described as the "Overseers House." Units were placed in a number of rooms, and a major discovery was that the Great House appears to have burned about 1850 while fully furnished. A gray ashy lens covered the floor of one room in the southeast corner of the structure (Slide WI-465), in many units underneath well defined brick wall fall (Slide WI-1227). This ashy lens, upon excavation was found to be filled with artifactual debris (Slide WI-433). A china cabinet appears to have been present, as the remains of a great many reconstructable plates, platters, and other vessels were found in a fairly small area (Slide WI-450). In all, 24 1 m units were opened in this room, encompassing about 60-70 percent of the floor area (Slide WI-1566). Tammy Forehand, Kevin Eberhard (Slide WI-421), Phil Carr, Chris Clement, and Dan Morse did much of that work (Slide WI-1223).

The great house is approximately 30 meters long by 20 meters wide at its widest point. The foundation consisted of field stone set in mortar. The walls were plastered on the interior and exterior. Post hole remains extended into the masonry foundation, and drain gutter holes were also spaced along the walls (Slide WI-1269). A large cistern with an arched cover was located at the north end of the complex, and is the most dramatic surviving architectural feature (Slide WI-1299). A patio appears to occupy the western portion of the complex, given the discovery of small paving stones from probable walkways in a number of units opened in this area (Slide WI-347). The foundations of the eastern half of the great house appear to have been reconstructed and reinforced several times, with footers and wall additions, to prevent it from collapsing down the side of the hill (Slide WI-1552). A mass of red brick was found to extend appreciably above the wall line along this eastern wall that appears to be the remains of a major roof support (Slide WI-1488).

At the south end of the great house main structure, rubble remains of possible steps were observed (Slide WI-1496), and low rubble walls beyond the main foundation suggest part or all of the complex was enclosed within some type of fence (Slide WI-1294). Five 1 m units were also placed outside of the structure to the north, to document how yard areas may have been used.

Several units were opened in other rooms, but no evidence for burning was found, suggesting these rooms may be from earlier construction episodes that had been abandoned or dismantled by the time of the fire (Slide WI-442). One of the goals of the analysis will be trying to resolve the construction history of the complex.

Three excavation units were opened between the great house and the outbuildings located to the west, uncovering a rough stone paving that suggests a court yard or walkway (Slide WI-14). A moderately well preserved bake oven was found beyond the paving, constructed of field stone and yellow brick (Slide WI-48). Two outbuildings, located to the north and south of the bake oven, were present that were likely servants quarters and a kitchen (Slide WI-1024). They were constructed of brick with some field stone set in mortar and were plastered on the interior and exterior. Three 1 m units placed in the southern structure, by Brinnen Carter and Angela Davis, confirmed its probable use as a kitchen, given the discovery of a probable cooking area (Slide WI-720), as well as a frying pan and baking sheets on the floor (Slide WI-704). Directly down slope from these outbuildings were found a great many late eighteenth century through mid-nineteenth century artifacts (Slide WI-1649).

A small building was located off the southwest end of the great house that may have been a storeroom or an overseer or servant's quarters (Slide WI-519). Five units were opened in the so-called "Overseer’s House" off the southwestern corner of the great house by Brinnen Carter and Angela Davis (Slide WI-1421). Appreciable numbers of 18th century artifacts were found, suggesting this may have been the original residence of Renaud and Tamaryn. A portion of surviving wood from one of the wall posts collected in 1992 was sent to ethnobotanist Lee Newsom, who determined it was Fustic. Resembling ironwood, this wood is used for fence posts in areas of the Caribbean where most larger trees have been cut down. Downslope from this outbuilding a complete early 19th century case bottle was found in the dense underbrush, where it had likely been tossed 150 or more years ago (Slide WI-1644).

Three areas downslope were also examined, that appear to be the remains of slave cabins. This work was conducted under the general supervision of Steve Kidd and Greg Heide. The 1805 census had indicated 17 structures were once present in a ca. 2 acre area below the Great House. Wall lines, rock rubble piles, and artifactual debris was found throughout this area in 1998. (Slide WI-1472), and three particularly promising areas were selected for more intensive study. All three proved to be the location of structures.

In Structure 1, home of "Carolina Pete" the iguana (Slide WI-935, previously noted), was identified by the presence of well defined rock wall lines (Slide WI-417). Eight 1 m units were opened inside the structure and five 1 m units were opened outside it (Slide WI-418) (Slide WI-576). Plastered wall lines were found on the north, east, and south sides, and a probable doorway was found on the southwest corner (Slide WI-1162). The floor of this structure was of packed earth. A concentration of approximately 100 chert cores, from exhausted strike-a-lights and gunflints, was found in four units opened near the door.

Structure 2 appeared to be the remains of a long, L-shaped barracks (Slide WI-701). It had a packed earth floor, and well defined rubble wall lines (Slide WI-733). No internal partitions could be distinguished, suggesting that, if present, they were of wood or other perishable material. A total of ten 1 m units were opened, nine inside the structure and one outside. A number of iron tools were found on the floor of the structure (Slide WI-1398), as well as appreciable pieces of local earthenware pottery that is thought to have been made by slaves.

Structure 3 was characterized by a well defined plaster floor found at from ca. 20 to 70 cm below the surface in an area of heavy slopewash (Slide WI-1336). The area was chosen for excavation by Amy Young (Slide WI-605) and Steven Kidd on the basis of a fairly dense surface artifact scatter. No surface architectural features had been observed, and the discovery of the plaster floor was a surprise (Slide WI-656). A total of thirteen 1 m units were opened in this area, 11 to define the structure, and two outside it. The fill immediately above the plaster yielded unusually well preserved bone from fish and other animals, and soil samples were taken from many of the units for fine screening, to try to collect a good sample of these remains. Greg Heide (Slide WI-1475) and Steve Kidd put in appreciable voluntary time on the weekends to ensure that the structure boundaries were fully documented. They were able to delimit three of the four corners of the structure. Several additional units each were opened in the other two slave structures prior to closing down, thanks to some hard digging by Steve, Dave Anderson, and Dan Morse.

A program of archival research is currently being conducted by David Knight, the project historian, directed to documenting the colonial occupation of these properties, specifically the lives of the principal 18th century owners, and the people who worked for them. Carolina Point was owned by two prominent freedmen for almost 40 years, Jean Renaud (owner from 1769-1792) and Peter Tamaryn (owner from 1793-1806). Mr. Knight has been able to locate the wills of these men, as well as a great many other documents related to the use of the island.

The archaeological and architectural remains from the Carolina Point Plantation Great House and slave village area exhibit a remarkable degree of preservation. They are of exceptional significance, and the project team believes every effort should be made to preserve and protect them.

Ruyter's Bay (12-VAM-3-21)
[List of Illustrations]

Ruyter's Bay is located at the north end of Water Island, below and to the north of Carolina Point (Slide WI-76). A historic masonry well apparently associated with the plantation complex on Carolina Point was found in the cove in 1992 (Slide WI-78). The water level at the time of the survey was about 8 feet below the surface. Extensive shovel testing in the vicinity of the well located a number of 18th and 19th century historic artifacts, although none were found in undisturbed context more than a few meters from the well.

The well was about 75 meters back from the ocean. It was made of stone and mortar, and was plastered on the outside and the top 50 centimeters inside. Shovel tests opened around the well in 1992 found late 18th to mid-19th century materials. A small debris-rich late 18th century midden was also found in an adjacent cove to the north that may also be the remains of another plantation outbuilding.

The 1998 fieldwork included the clearing and photographic documentation of the well area, and the excavation of five 1 m test units around it (Ruyter’s Bay Contour Map). This work was conducted under the direction of Brinnen Carter. Again, 18th and 19th century historic artifacts were found, many in a dark midden lens found just below the surface on the east side of the well (Slide WI-741). This lens had a great many large rocks in it, and may be the remains of a path to the well area, or perhaps debris from the construction of the well. Given that the area was apparently a pond, this may have facilitated access. One unit opened at the side of the well on the west side showed the masonry was lined with mortar to a depth of more than 70 cm below the surface (Slide WI-1461). The masonry well is considered a significant historic resource that should be preserved and protected.

Tamarind Tree Bay (12-VAM-3-9)
[List of Illustrations]

Tamarind Tree Bay (Slide WI-80) is located along the northwest coast of Water Island, to the south and below Carolina Point and to the north of Elephant Bay (Slide WI-1597). An extensive scatter of 18th through 20th century ceramics were found in the central and northern part of the cove in 1992, together with the remains of a historic masonry well in the central part of the cove. The well is made of stone and mortar and is filled almost to the top with modern trash (Slide WI-239). Shovel tests opened next to the well yielded 18th and 19th century ceramics. In the 1930s a number of burials were excavated in the cove, of people of African descent. No grave markers or offerings, or human bones, or other evidence for burials was found in either 1992 or 1998, however. These historic remains appear to be associated with the Carolina Point plantation slave village.

During the 1992 survey, a strong chemical odor was noted in one shovel test in this same cove, in an area notably devoid of vegetation. Although no chemical weapons testing was reported to have taken place in this area, large amounts of 1940's period trash were found, including an ironstone plate sherd stamped with the initials for the U.S. Quarter Master Corps and dated September 16, 1940, providing a degree of chronological control unusual in archeological inquiry. This debris strongly suggested the military had been present in some capacity in this area, information that was relayed to the authorities overseeing clean-up activities on the island.

The 1998 fieldwork, under the general direction of Brinnen Carter, began with extensive clearing activity over a 40 x 20 m area in the vicinity of and to the north of the well (Slide WI-774). Ten 1 m units opened, four around the well and the other six at varying distances to the north and northeast (Tamarind Tree Bay Contour Map). Two 1x2 m trenches were opened on the north and east sides of the well, showing that it lined with mortar to a depth of about 70 cm, below which was bare rock (Slide WI-893). Builders trenches were identified in both trenches, indicating an appreciable area had been dug out and then refilled during the construction (Slide WI-851) (Slide WI-890).

The six units opened away from the well failed to find any evidence for human burials, suggesting they occur in another part of the cove. Historic artifacts were found to appreciable depths in most of these units, indicating appreciable mixing of the deposits, particularly in low areas of the site near the well, where land crab holes were common. The historic artifacts are probable debris from the nearby slave village, which may extend into this cove. A number of rock piles were noted on higher ground, but an excavation into one of them in did not detect well defined wall lines or submound features. Finally, a number of prehistoric ceramics were found in the two northernmost 1 m units, located some 30 m from the well, and on higher ground whose soils do not appear to have been as severely mixed. Only the Banana Bay South site has yielded more prehistoric sherds on the island, and fairly extensive prehistoric use of the area is indicated, adding to the importance of the site.

The masonry well at Tamarind Tree Bay is a significant historic resource and should be preserved and protected. The excavations conducted in 1998 examined only a small fraction of the bay area, yet discovered a previously unrecognized prehistoric component. In addition, the possibility exists that human burials and possible structures from the slave village are present in the cove. For this reason the entire cove area should be preserved and protected.

Elephant Bay (12-VAM-3-22 and 12-VAM-2-23)
[List of Illustrations]

Elephant Bay is located along the northwest coast of Water Island just south of Tamarind Tree Bay and north of Landing Bay (Slide WI-177). Late Saladoid and possibly Elenoid prehistoric components were identified at the site in 1992, with minor ceramic concentrations noted at the northeastern and central parts of the cove. The 1992 testing at the northeast end of the bay delimited subsurface materials within a 15 x 20-meter area, and testing near the center of the bay delimited subsurface prehistoric materials within a 15 x 40-meter area. The central component appears to date to the late Saladoid era (ca. A.D. 600), while the other assemblage is dated to later in the prehistoric sequence, possibly the Elenan period (ca. A.D. 800-1200).

Upon visiting the site in 1998, the soils in the northeast end was found to have been extensively disturbed by heavy equipment, precluding any likelihood of discovering undisturbed archeological remains. Several sherds were observed on the surface in the central part of the bay, however, and fieldwork was conducted in this area, under the direction of Mike Russo. Following extensive clearing, 14 0.5 m shovel test units were opened to delimit the stratigraphy, boundaries, deposit depth, and general content of the concentration (Elephant Bay Contour Map). Historic artifacts were found to great depths in many of the units opened, indicating appreciable disturbance and reworking of the deposits has occurred. Two larger 1 m units were then opened, and taken to well below a meter in depth (Slide WI-1174). Again, appreciable disturbance was indicated (Slide WI-1195). A number of shells were observed below 1 m in depth in one unit that were initially thought to derive from an Archaic midden, but the complete absence of any other artifacts suggests this is unlikely.

The archaeological deposits from Elephant Bay do not appear to be in primary context. A preservation covenant at this site may be unnecessary, although a more comprehensive examination of the entire cove area would likely need to occur before this could be argued unequivocally.

Sprat Bay Structure (12-VAM-3-208)
[List of Illustrations]

During the 1992 survey, a small scatter of eighteenth century era historic ruins and artifacts were located on the hill which overlooks Sprat and Sand Bays at the north end of the peninsula defining Sprat Point (Slide WI-1062). The site consists of a masonry foundation representing the remains of a small building with standing walls (Slide WI-99), an adjacent rock pile, and the foundation stones for a small ancillary structure about two meters on a side. A light scatter of 18th century historic artifacts was found downslope and to the north of the structure, that appears to represent sheet midden debris from the occupation. Diagnostic materials indicate the site was occupied in the early 18th century and was abandoned by ca. 1750 to 1760. The complex is unusual because it was not later reoccupied, and because of its excellent preservation. Only one other comparable early site, Zufriedenheit, has been investigated to date in the Virgin Islands, by Elizabeth Righter, who with David Anderson directed the excavations at the Sprat Bay Structure site in 1998.

On Oxholm's 1778 map a single structure is depicted just to the west of Sand Bay that may represent this complex, albeit mislocated some 200 meters north of its actual location (see Oxholm 1778 map noted previously). The hill crest near where the structure actually was found would have provided an excellent view of the approach to St. Thomas harbor.

A large rock pile was found three meters from the building, that was originally interpreted as a possible grave. Extensive debris was found downslope that likely reflects trash disposal, while sherds of African-Caribbean ware found in a cove below the site may be outlying debris or the remains of a slave habitation.

The 1998 fieldwork included extensive clearing, mapping, and the excavation of 40 1 m units and 45 0.5 m shovel tests (Sprat Bay Structure Contour Map) (Slide WI-1020). The shovel tests were dispersed to the north and northwest of the primary structure at 3 m intervals, and were used to place larger units. Historic debris was found in some quantity up to 25 m downslope from the primary structure, with most occurring between ca. 15 and 20 m. A concentration of boulders some 4 x 8 m in extent was located to the northwest of the structure, that may be collapsed foundation walls from a kitchen or some other type of outbuilding. Appreciable sheet midden, consisting of ceramics, glass, shell, and other historic artifacts, was found in this area, supporting this possibility.

The main Sprat Bay structure was found to be in an excellent state of preservation (Slide WI-1578). It consisted of a foundation 8.5 meters long by 5.4 meters wide, with a gap in the northeast corner that may represent a doorway. The walls were of field stone set in mortar that had been plastered on the interior and exterior. The floor had been plastered over a rock pavement (Slide WI-1153), and was intact in several areas. Several support posts were identified along the southern and eastern wall line, and two were excavated, and were found to extend 70 cm or more below the floor (Slide WI-1130). The small structure to the west of the main house was completely exposed by Josh Kehrberg and Steve Kidd, and was found to be about 1.5 x 1.5 m in extent (Slide WI-1115). Virtually no artifacts were found in any the units opened to delimit it, and its function remains unknown at the present. The site’s standing structural remains were extensively photographed and will be used to produce drawings of these remains.

The rock pile located to the east of the primary structure was piece plotted and then a 1 m unit was opened in the northern part of the scatter (Slide WI-973). This was excavated with great care by Dan F. Morse and Roger Michael (Slide WI-978). No pit or other features were found below the rocks, which appear to have been placed directly on undisturbed ground. The pile may be construction materials from the building of the house. Two other small rock piles further to the east were also tested, and also proved to be surface features (Slide WI-1159).

The artifact assemblage from the Sprat Bay Structure appears to predate 1760, since no sherds of either creamware or pearlware were found. The overall assemblage, in fact, appears to date from about 1730 to 1750, a period for which there is no record of settlement on Water Island. As such, the archaeological and architectural remains are the earliest known on the island from the historic period in good context. This fact, and the remarkable degree of preservation that exists, makes the site of exceptional significance. Every effort should be made to preserve and protect it.

Banana Bay (12-VAM-3-56)
[List of Illustrations]

In 1992 a small shell midden was found in Banana Bay, which is located at the north end of Water Island near Banana Point (Slide WI-129). A 12 meter section of eroding bank at the north end of the cove contained large numbers of conch or Strombus gigas shells mixed with water worn rock and coral. The midden deposits were as much as two meters thick in places (Slide WI-132). The absence of historic or prehistoric pottery prompted the collection of three of the Strombus gigas shells for dating, from the surface, the middle, and base of the deposits. The shell recovered from the base of the midden was radiocarbon dated at 1420 + 60 years, or A.D. 530 (Beta 58094). The shell taken from ca. 50 cm below the surface was radiocarbon dated at 800 + 60 years, or A.D. 1150 (Beta 58095). The shell recovered from the surface of the midden was radiocarbon dated to 740 + 60 years, or A.D. 1210 (Beta 58096). No artifacts or paleosubsistence remains besides conch, however, were found in the midden or elsewhere in the bay, suggesting the site was a special purpose shellfish gathering and processing area.

The August 1996 reconnaissance found that most of the shell deposits observed at the site in 1992 had eroded away, probably the result of storm damage associated with recent hurricanes. Accordingly, the remaining portions of the midden were examined during the 1998. field program, to document what was left, and how the deposits formed. This work was directed by Mike Russo (see Slide WI-1217 previously referenced).

The entire profile was cleaned and photographed (Slide WI-1069) (Slide WI-1073), and 1 x 3 m long stratigraphic cut was made into the central portion of the midden, perpendicular to the beach (Banana Point Contour Map). A 1 m unit was also opened in storm surge deposits just east of the midden area. The reworked shell in this unit was identical in appearance to the remains found in the surviving midden areas and in the stratigraphic cut, suggesting the surviving deposits were reworked.

Interestingly, a recent historic scatter of some 200 conch, opened with a machete, was found at the base of the midden (Slide WI-1212). Deposited since 1996, the scatter was piece plotted, photographed, and each shell was numbered and measured (See Slide WI-1217, previously noted). The scatter provides a valuable picture of modern conch processing, and suggests how local prehistoric conch middens may have formed. Measures of these modern conch will be compared to prehistoric conch recovered from sites on the island to assess whether conch populations were stressed from overexploitation as they are today. Comparison of processing techniques will also be made.

The archaeological deposits from Banana Bay do not appear to be in primary context. A preservation covenant at this site appears unnecessary. The possibility that an occupation area occurs in the nearby cove, however, cannot be unequivocally ruled out, particularly in light of the unexpected discoveries at Banana Bay South, as described below.

Banana Bay South (12-VAM-3-210)
[List of Illustrations]

Eighteenth and twentieth century artifacts were found just to the west of Banana Point at the northern end of Water Island in a small unnamed cove that was designated Banana Bay South by the 1992 survey team (Slide WI-139). A dark midden lens filled with artifacts dating to the early twentieth century as well as artifacts dating to the third quarter of the eighteenth century were found in a 15 x 30-meter area. The early 20th century materials were associated with extensive charcoal and slag-like debris, supporting the observation of a local resident that a metal boat had once been dismantled in this area. A small number of conch shells with holes punched in them to extract the meat were also found, in the storm surge near the front of the bay. These artifacts suggested possible prehistoric use of the cove.

The 1998 fieldwork, which was directed by Greg Heide (Slide WI-920), included extensive clearing in the central part of the cove, followed by mapping and the excavation of six 1 m units and 16 0.5m units (Banana Bay South Contour Map). A dense prehistoric midden made up of conch shell, shell tools, and ceramics was found extending to depths of more than a m in several units (Slide WI-919). Conch tended to be more common closer to the shore, while ceramics were more common in units opened near the back of the cove. The ceramic assemblage was the densest found to date on Water Island. Preliminary inspection suggests some of the ceramics may post-date A.D. 1200. The midden appears to have formed through a number of discrete conch collection and processing activities, as shell was intermittently distributed through the units (Slide WI-951). The site may represent a composite of a number of episodes like that producing the modern conch midden found at Banana Bay. The incidence of shell and ceramics in the 0.5 m units dispersed over the cove area support this possibility. Few 18th century artifacts, the reason the site was examined in 1998., were found in any of the units. The discovery of a dense prehistoric assemblage was unexpected, and demonstrates why cove areas need to be extensively examined.

The extent of the recent historic midden was traced with 0.5 m units, and a 1 m unit was opened to collect a sample of artifacts from it. The midden was found to be about 40 cm thick, and extending from ca. 30 to 70 cm below the surface. Several exploited conch were found below this, reflecting a continuation of the prehistoric midden. The prehistoric assemblage in Banana Bay South is the most extensive and diverse found on Water Island. The site is extremely important for understanding prehistoric use of the island, and should be preserved.

Providence Point Plantation (12-VAM-3-211)
[List of Illustrations]

The remains of Providence Point Plantation, a major eighteenth and nineteenth century settlement, are located along the main road that leads up the hill from Landing Bay. Providence Point Plantation was constructed in the last quarter of the eighteenth century, and detailed year-by-year historical records of ownership and occupation date from 1799. According to the historic record, at least four structures were associated with the plantation. The area in which the plantation is depicted in the historic record has been extensively developed, however, and a number of modern structures have been built in the area where the plantation complex buildings were situated. Portions of a few plantation structures remain, and have been either incorporated into new buildings, or have been preserved intact by the current owners. At present the only obvious remains are those of a bake oven in a good state of repair (Slide WI-118) (Slide WI-121), and a cistern that serves as the foundation for a house (Slide WI-126). Because the area of Providence Point Plantation has been extensively developed, no excavations were undertaken here in 1998.

Fort Segarra/World War II Military Complex
[List of Illustrations]
Fort Segarra Level 1 Floor Plan
Fort Segarra Elevation and Sections

During and shortly after World War II an extensive series of structures were built on Water Island by the military, including a massive underground fortifications complex at the south end of the island, on Flamingo Point. This complex includes two gun emplacements (Slide WI-161) serviced by an extensive underground bunker system (Slide WI-218) centered below an observation tower (Slide WI-159). The openings to the underground complex were densely overgrown in 1992 (Slide WI-164), picturesque vegetation that was removed in 1997 for a ceremony held at the complex, marking the transfer of land to the territorial government. The Flamingo Point complex provides an excellent shelter for island residents during hurricanes, and is in surprisingly good condition.

Other military architecture on the island include two ammunition bunkers/igloos that were built into the north side of Carolina Point, fortunately missing the plantation complex atop the hill (Slide WI-204). Two small concrete platforms were found, one on Sprat Point (Slide WI-169; Slide WI-170) and Banana Point (Slide WI-246; Slide WI 248) that may represent light or gun emplacements. The platform on Sprat Point has the date 1941 on it, suggesting it may be related to the military use of the island. General architectural plans of the bunker complex at Flamingo Point and the igloos at Carolina Point were made by NPS historic architect Dennis McCarthy during the 1996 reconnaissance (Fort Segarra Architectural Plan A; Fort Segarra Architectural Plan B). GPS mapping of these structures, together with limited photographic documentation, was conducted in 1998.

GPS mapping and photodocumentation was also conducted at a rock overhang on Banana Point (Slide WI-250), where a cache of wired-up military explosives was found and photographed by Anderson and Brewer during the 1996 inspection (Slide WI-252). These were removed shortly thereafter by the U.S. Navy and the FBI (Slide WI-254). Several 18th and 19th century historic artifacts noted on the floor and slopes of the rock overhang during the 1996 reconnaissance, but that were left in place, were also collected in 1998.

Analysis and Report Production

Analysis of the Water Island project collections is currently underway. A laboratory was established at one of the project houses during the fieldwork, where washing, sorting, and packing for shipping of project materials occurred. A vast quantity of artifacts and special samples were collected, weighing approximately 2000 pounds. Over half of the artifacts were washed and rebagged over the course of the fieldwork, and all were carefully packed for shipping. The assemblage encompassed 111 boxes weighing from ca. 10 to 40 pounds each. These were sent via first class postage, insured, to the Southeastern Archeological Center in Tallahassee, and all arrived safely within a few days of shipping.

The lab work was conducted largely by Tiffanie Bourassa, Jennifer Carter, and Phyllis Morse, assisted by Janice Burke. Over the course of the summer they not only washed, sorted, and rebagged close to half the massive assemblage, but they saw to it that the entire artifact collection was packed up for shipping. We placed the last artifact boxes in the mail July 30th, four days ahead of when tropical storm Alex was due to hit! Fortunately the storm turned north, and there were no problems with weather.

All 111 boxes of artifacts and equipment shipped out via the US Post Office arrived safely within a few days, as did all of the film and field notes that were hand carried. Special thanks are owned the staff of St. Thomas Communications at Crown Bay Marina, who helped with the shipping effort. All of the film has been processed, and we have finished cataloging the color slides and black and white negatives. There were almost 2000 of each, when the 1992 and 1996 field seasons were added in. Almost all of the shots came out, which is always a relief. These shots, plus the close to 20 hours of video taken over the course of the 1998 field season, and the 14 hours shot in 1992, means we have plenty of illustrations to work with down the line for the reports, meeting, and web presentations.

Greg Heide and Mike Russo have been working with the mapping data, and have produced a set of draft maps of all of seven project sites in Autocad and Surfer. These will be cleaned up appreciably over the course of the project, and will serve as base maps for subsequent analyses and presentations.

Mike Russo set up a formal protocol for the processing of the artifact collections with the Department of Agriculture’s Animal Plant Protection Service. Basically, soils from the Caribbean must be treated with care, to avoid contamination of our local soils by potentially dangerous microorganisms. All dirt from the project assemblages is being caught in fine mesh screen during washing, which is ongoing at present. By next spring we plan to have the entire assemblage washed, sorted, rebagged and reboxed, and cataloged and inventoried in preliminary form. The collections will then be ready for more detailed analyses, which will be conducted by various members of the project team, as well as by a number of outside specialists. The processing effort is being handled by Tiffanie Bourassa, Greg Heide, and Steven Kidd.

Data analysis and final catalog generation will be conducted by project specialists at a number of institutions, with the results placed into an electronic (Access©) database. A full catalog will be included with the final technical report, and with the collections when they are placed in a curation facility. The catalog will include the site number, catalog number provenience, field specimen number, and a brief technical description of the artifact(s). Stratigraphic information for each subsurface test excavated during the project will also be provided in the appendix to the final technical report. This information may also be posted on the project web site.

The project paper records have been duplicated on acid free paper, all slides and photographs have been cataloged and placed in curatorial mounts, and we are preparing to scan the slides for electronic storage on CDs. This effort will make assemblage processing and analysis easier, and will enable us to send information to specialists without fear of loss. Primary records are lost from too many projects, something we wish to avoid at all costs.

Subsets of the collection will go off to individual researchers for detailed analyses starting early next year. We are exploring geoarchaeological and zooarchaeological analysis options Sylvia Scudder and Irv Quitmyer at the Florida Museum of Natural History in Gainesville. One area we are looking into is the possibility of obtaining seasonality information from the conch shells recovered from several project sites. Shipping of ethnobotanical and other paleosubsistence samples to other project specialists should occur later this fall.

A web site will be posted in the near future describing the project, and including a number of slides of the fieldwork. We expect this site to evolve over the course of the project, and that it will foster information exchange. We expect to give a number of technical and popular presentations about the Water Island project over the next two to three years. Initial papers on the Water Island work have been presented already by Anderson and Ken Wild at the 1994 Society for Historical Archaeology (SHA) meetings, and by Anderson at the Southeastern Archaeological Conference on November 12th, in Greenville, South Carolina. We hope to have a full symposium on the project at the 2000 SHA meetings.

A Note on Project Staffing
[List of Illustrations]

Dr. David G. Anderson served as field director and principal investigator (Slide WI-197). Research specialists in the field included Dr. Phil Carr (prehistoric archaeology; lithic technology) (Slide WI-606); Dr. Chris Clement (historic archaeology) (Slide WI-1146; Slide WI-1223 previously noted); Dr. Dan F. Morse (prehistoric and historic sites archaeology) (Slide WI-1304; Slide WI-1223 previously noted), and Ms. Phyllis A. Morse (historic sites archaeology) (Slide WI-1024, previously noted), Dr. Virgil Noble (historic sites archaeology), Dr. Elizabeth Righter (historic archaeology (Slides WI-69, WI-1108 previously noted), Virgin Islands archaeology), Dr. Michael Russo (zooarchaeology, prehistoric archaeology) (Slide WI-1217 previously noted), Dr. Amy Young (historic archaeology)(Slide WI-596), Ms. Martha Zierden (historic archaeology) (Slide WI-790). In addition, a number of BA and MA level archaeological field and laboratory technicians were present, including Ms. Tiffanie Bourassa (Slide WI-698 previously noted, Slide WI-1058, previously noted), Ms. Janice Burke, Mr. Brinnen Carter (Slide WI-647, previously noted; Slide WI-698 previously noted, Slide WI-1058, previously noted), Ms. Jennifer Carter (Slide WI-519, previously noted; Slide WI-698 previously noted), Ms. Angela Davis (Slide WI-1024, previously noted; Slide WI-698 previously noted; Slide WI-1058, previously noted), Ms. Liz DeGrummond (Slide WI-698 previously noted), Mr. Kevin Eberhard (Slide WI-279; Slide WI-421 previously noted), Ms. Tammy Forehand (Slide WI-1223 previously noted), Mr. Greg Heide (Slide WI-698 previously noted, Slide WI-1058, previously noted; Slide WI-578, previously noted), Mr. Joshua Kehrberg (Slide WI-698 previously noted, Slide WI-1058, previously noted), Mr. Steven Kidd (Slide WI-698 previously noted, Slide WI-1058, previously noted), Ms. Regina Meyer, Ms. Emily Moss (Slide WI-279, previously noted), Ms. Thedra Palmer (Slide WI-790, previously noted), Mr. Bob Pederson, and Ms. Helen Savill (Slide WI-1058 previously noted). Mr. John Ehrenhard, the Director of NPS’s Southeast Archeological Center (Slide WI-1174, previously noted), and Ms. Janice Burke, SEAC Finance Officer, visited and inspected the fieldwork.

A number of people on St. Thomas and on Water Island helped the project tremendously, and deserve our thanks: Doug Armstrong, Merci Blosser, Clifton Boyd and the "Brighter Writers" students, Stepanie, Rachel, and Shane Brown, Sandy Castorena, Mary Coe, Jackie Coleman, Anita Davis, Carol Dudley, Bryan and Arlene Dooley, Dave Flaggerty (US Coast Guard), Mrs. Francis, Linda and Luci Guidley, David and Lolly Knight, Larry Mann, Kevin Mattock, Roger and Eric Michael, Susan and Butch Miller, Brad and Collette Monroe, Mari Morse, Joan and Paul Murray, Barry Nash, Chris Parker, Carl Ritter, Bill Shea, Harold and Clyde Tapp, Kenneth A. Turbo, Miki and Holland Van Gores, Michael Wall, "Water Island Dave," Arthur Watres, Dianne and Roger White, Wendy Yellen, and of, course, all the people at Honeymoon Beach who made Saturday evenings and Sunday afternoons so relaxing, including the owners and staff of the Lunch Box.

Historical investigations into what life was like on Water Island are being conducted by Virgin Islands historian David Knight. Mr. Knight is assisted by Mr. George Tyson, another Virgin Islands historian, who conducted the detailed historical investigations recounted in the 1992 Water Island survey report. These investigations have as one overarching goal the location of as many primary sources as possible about the people who lived on the island, and what their life was like.

Dr. Emily Lundberg, the principal authority on Virgin Islands prehistoric ceramics, has agreed to serve as a consultant in the analysis of prehistoric ceramic artifacts recovered from the project. Dr. Lee A. Newsom will examine the ethnobotanical remains, while Dr. Michael Russo will oversee the analysis of zooarchaeological remains. Ms. Tiffanie Bourassa, Mr. Greg Heide, and Mr. Steven Kidd, SEAC staff members and students at Florida State University, are working on the processing and analysis of the collections and records.


In conclusion, from our recent survey work, we have learned that Water Island has a well preserved historic archaeological record spanning the period of colonial settlement in this part of the Caribbean. This record, furthermore, is complemented by a superb set of historical records. Large numbers of slaves were working the island's two plantations during the years shortly before and after 1800, and, interestingly, their owners were themselves freed blacks. This provides an almost unparalleled opportunity to compare the archaeological remains from these sites with the assemblages from white-owned plantations and slave villages of the period.

Work is well underway on the Water Island Archaeological project. The fieldwork has been completed, and we believe it was unusually successful. A remarkable and extensive artifact collection was found, providing important information on past life on the island. Products of this project will include final technical and popular reports. We have recommended that these sites be preserved or, if this is not possible, carefully and comprehensively excavated. Hopefully Water Island will have the chance to offer further information about both black and white lifeways in the Virgin Islands.

References Cited

Anderson, John Lorenzo
    1992    Night of the Silent Drums. Mapes Monde, St. Thomas.

Wild, Ken, and David G. Anderson
    1992    An Archeological Inventory and Assessment of Cultural Resources on Water Island,
               U. S. Virgin Islands. Southeast Archeological Center, National Park Service,
                Tallahassee, Florida.


Appendix A:  Carolina Lyst Plantation Ownership and Occupational History 1769-1794

Appendix B:  Carolina Lyst Plantation Ownership and Occupational History 1795-1914

Appendix C. La Providence Plantation Ownership and Occupational History 1799-1914

List of Illustrations

        ¹Web pages adapted from a paper entitled "The Water Island Archaeological Project: Archaeology and History in the Eastern Caribbean," presented in the symposium "Current Research and Methodological Approaches to Cultural Resource Management Projects in the Southeastern National Parks” organized by Margo Schwadron and David Brewer, at the Annual Meeting of the Southeastern Archaeological Conference, Greenville, South Carolina, 12 November 1998.