The Palace Lands Site (site 44WB90) was once inhabited by a small group of enslaved Virginians from c. 1747-1769. During this period, the site was part of John Coke’s 200-acre plantation. The site’s inhabitants were most likely a family who shared a two-room dwelling with a central chimney and a root cellar. This group also constructed two post-and-rail fences and a series of three ditches adjacent to their house. Despite intensive surveys of the project area, archaeologists discovered no other features. There was, however, a natural ravine just northwest of the dwelling that had been filled with 18th-century refuse.
The site is named for its association with the Governor’s Palace Lands. John Coke died in 1767, and willed the plantation and a number of enslaved blacks to his son Samuel. Samuel continued to operate the plantation until c. 1769 when he put the land, enslaved blacks, and livestock up for auction. At about this time, the Governor’s Council purchased the 200 acres and it became part of the Palace Lands. The last two royal governors undoubtedly hired enslaved blacks or assigned their own to work within this portion of their landed estate. There are no sources, however, that indicate that these enslaved individuals occupied the site.
The artifact assemblage reveals that two subsequent occupations took place in the site’s vicinity. The first occurred during the late eighteenth century to early nineteenth century. The second dates to the late nineteenth century and is represented solely by artifacts recovered from the plowzone. Still, the site’s features and the artifacts retrieved from feature contexts are associated with Coke’s enslaved Virginians.
Site 44WB90 is located within Colonial Williamsburg’s (CW) Visitor Center complex north of the Historic Area. The site is situated on a terrace bounded 50 meters to the west by the Cascades Motel and 25 meters to the east by Route 60. Due to planned renovations of the Visitor Center, which have since taken place, CW’s Department of Archaeological Research (DAR) commenced with a survey of the area in 1996. The Phase I and II surveys were followed by a Phase III data recovery of the site in 1998-1999.
A chain of title from 1704-1904 has been traced for the Palace Lands site (Franklin, n.d.). The land on which the site is located was once a 300-acre lot owned by Mary Whaley in 1704 (YCRR 1704). The York County parcel was half-a-mile from Williamsburg. Its most prominent feature was Capitol Landing Road, a major thoroughfare which passed through the property from town and to Queen’s Creek and Capitol Landing. During the colonial era, Capitol Landing was a busy site of trade and shipping; enslaved Africans entered the colony at this site. The property would be described with reference to Capitol Landing Road until at least the late nineteenth century.
Whaley’s property was divided, sold and willed a number of times before John Coke purchased it by 1747. At the time, the lot consisted of 200 acres. Although a record for the transaction has not been located, Coke more than likely bought the land from a bricklayer by the name of John Baskerville who was deeded the land in 1742. Another deed dated for August 17, 1747, concerning a 52-acre parcel describes the parcel as bounded by “Mr. Coke’s Line”. It also has a sketch showing the 52 acres with “Mr. Coke’s Land” indicated for the property to the north which is the 200-acre lot in question (YCR, 5:212-216). Thus, by August 17, 1747, Coke held title to the 200 acres.
Coke was a silver and goldsmith, as well as a tavern owner, who came to Williamsburg from Derbyshire, England, in 1724 (Bullock 1931; Daniel 1946:11). He owned five lots in town on Nicholson Street by 1755, and the house that he and his family resided in is still standing. It is referred to today as the “Coke-Garrett” house. Since the Palace Lands site was inhabited by c. 1750, Coke apparently established his plantation not long after acquiring the land. There are no records to suggest that he leased the land or hired laborers, so Coke’s enslaved Virginians must have provided the labor on his plantation.
Coke’s (1768) estate inventory lists nine enslaved blacks: five men and four women. This was a relatively large slaveholding for a Williamsburg household. Although five-sixths of the families owned slaves in town, a high percentage of the slaveowners were of modest means who owned one or two individuals (Tate 1965:55). Rather than simply suggesting that Coke was wealthier than many of his neighbors, these numbers indicate that he needed enslaved labor for other than domestic service, the work performed by most of Williamsburg’s enslaved population. Coke likely assigned domestic tasks and tavern duties to some of his enslaved Virginians, while the rest were put to work at his plantation. Although we know the names of these nine individuals, it is unclear at this time which ones lived at the Palace Lands site. Coke or his sons probably supervised work at the plantation since it was less than a half-a-mile away and easily accessible via Capitol Landing Road which was adjacent to their house.
Work at Coke’s plantation consisted of raising livestock and crops, and felling wood for fuel. This is indicated by the items listed in his inventory (Coke 1768). Coke’s estate included 24 head of cattle and 10 calves, oxen, and one sow. He also owned five horses, several of which may have been kept at the plantation. Work-related tools listed in his inventory include six axes, seven hoes, three spades and a pair of sheep shears. There is also an entry for “79 barrels of” valued at ?35, but the inventory is torn at the entry. Presumably, the barrels held grains that were raised at the plantation.
Coke died in 1767 and divided his estate between his two sons, Samuel and Robey, and his wife Sarah. Robey inherited the town property and to his son Samuel he willed “my plantation containing 200 acres, more or less, lying on both sides of the Main Road which leads from the city of Williamsburg down to the Capitol Landing commonly called Queen Mary’s Port, to him and his heirs forever” (Daniel 1946:12). Coke’s slaveholding was divided between his sons. Samuel continued to operate the plantation, but only for a short time. On January 12, 1769, Samuel and Sarah announced the following (Virginia Gazette 1769):
To be SOLD by publick auction, on Thursday the 2d of FEBRUARY next, at the late dwelling-house of JOHN COKE, deceased, in Williamsburg, ALL his HOUSEHOLD & KITCHEN FURNITURE, several valuable SLAVES, with the stocks of CATTLE, HORSES, and SHEEP; also a quantity of CORN and FODDER. At the same time will be sold, or rented, a plantation lying on both sides of the road to the Capitol landing, containing upwards of 200 acres; it is exceeding good land, and in order for cropping. Credit will be allowed for all sums above five pounds until the 20th of October next, the purchasers giving bond and security to:
SARAH COKE, Executrix. SAMUEL COKE, Executor. The HOUSES in Williamsburg will be rented at the same time, on reasonable terms.
Although a record for the sale of the plantation has not been located, the Governor’s Council eventually purchased Coke’s plantation. This probably took place at the 1769 auction, but it is certain that the property was part of the Palace Lands by 1773. In a deed dated to that year, a 100-acre parcel is described as “bounded by the Governor’s Land”, which was the same land formerly of the Cokes (YCD, 8:343-347).
As early as 1618, the appointed governors of Virginia enjoyed the use of a landed estate that produced additional income through leasing the land to tenants (Gibbs 1980). Soon after Virginia’s seat of government moved from Jamestown to Williamsburg in 1699, the Council acquired a 75-acre tract for the governor’s residence. Two additional land purchases would enlarge the tract to 364 acres by c. 1769-1773. If Coke’s former parcel was part of the Palace Lands by 1769, the last two royal governors had the privilege of its use. Norborne Berkeley, Baron de Botetourt, served as governor from 1768 until his death in 1770. His replacement, John Murray, 4th Earl of Dunmore, was the last royal governor. Dunmore began his residency at the Governor’s Palace in 1771 and ended it abruptly when he fled the Palace four years later for the safety of a naval ship anchored on the York River.
Both Botetourt and Dunmore kept a bevy of staff and laborers busy. Both had a number of indentured servants and enslaved blacks, and they also hired free and enslaved blacks to work within the Palace and its surrounding grounds. Patricia Gibbs (1980) notes that the Palace Lands estate was subdivided into a number of utilitarian areas that each served to support the governors’ households. The largest portion of the estate was referred to at the time as the Palace “park”; Coke’s former plantation acreage was designated park land. The park included woodland, meadows, pastures and orchards (Gibbs 1980:5).
Gibbs’ (1980) research on the Palace Lands demonstrates that the park was used as pasturage for the governors’ livestock, to raise crops, and for felling trees for fuel. Enslaved blacks, whether hired or owned, were most likely assigned to work in the park, but specific mention of this in the historical record has not yet come to light. Botetourt’s slaveholding numbered seven adults and one child, and it appears that these individuals worked mainly within the Palace and its adjacent grounds. Dunmore, however, owned three plantations and his Loyalist Claim (for losses sustained during the War) includes 57 enslaved blacks (Willis et al. 1998:356-357). It is possible that some of these individuals were responsible for raising crops and caring for Dunmore’s 154 head of cattle, 150 sheep, and other livestock and horses kept at the park (Hood 1991:298).
With Dunmore’s departure from the city in 1775, and the arrival in 1776 of the newly-elected governor, Patrick Henry, the Palace Lands park was set aside to house American troops. Following the removal of the capitol to Richmond, The College of William & Mary acquired the 364-acre Palace Lands tract through an act of the General Assembly in 1784. The property would be deeded and willed multiple times for the next 120 years. The most recent deed dates to 1904, when Dr. Van F. Garrett sold the property to the Southern Land Company. It is referred to as “Garrett Farm” in this deed.
Excavation history, procedure, and methods
Prior to site excavation, the project area was surveyed and tested in 1996. During the Phase I survey, archaeologists dug 135 40-cm shovel tests at fifteen-meter intervals across the future excavation site and beyond the site’s boundaries (Pickett 1997:5). There were 23 positive shovel tests and archaeologists recovered 18th-century artifacts from three of these. The Phase II survey and testing focused on the area where these early artifacts were discovered. Archaeologists dug 16 75 × 75 cm test units at ten-meter intervals and eight 75 × 75 cm test units at five-meter intervals. They also excavated two 1 × 1 m units and one 2 × 2 m unit. Artifacts dating to the third quarter of the eighteenth century were evident in a number of test units, and archaeologists also uncovered a portion of a brick chimney foundation (Cooper 1997). It was decided at this point to proceed with excavation.
Excavations of the site took place during June and July of 1998-1999. The site was mechanically stripped of young trees and dense ground vegetation at the early stage of the excavation. The plowzone extended 11-56 cm below grade across the site. In addition to plowing, recent construction activities also impacted the site. The expansion of Route 60 and the construction of a house adjacent to the site sometime during the 20th century led to heavy disturbance of the plowzone layer along the site’s eastern and southern bounds. Machine push piles of redeposited subsoil and gray silt were discovered on top of and mixed with plowzone. These activities did not appear to have disturbed any of the excavated features, although evidence of cultural activities that might have extended east and south of the excavation unit were probably destroyed.
The general excavation unit measured 20 x 32 meters; archaeologists dug a total of 75 2 x 2 m and two 1 x 2 m units. In addition, ten 1 x 1 m test units extending to the north and west of the excavation were also dug as part of a limited site survey in 1999. The plowzone, redeposited subsoil and silt layers were removed by shovel-shaving to subsoil, although three transects were machine graded to subsoil (along 990E/997E-1013E, 1004N-992N/995E-997E). In general, a 25-percent sample from each 2 x 2 m unit was dry-screened through 1/4” mesh.
All features were hand trowelled save for five postholes that were shoveled out following the last field season. Most of the feature fill was dry-screened through 1/4” mesh although deposits within F01, F04, and F05 were wet-screened through 1/16” mesh. Flotation samples were collected from every feature deposit. Phytolith and soil chemistry samples were selectively collected from features.
The major features associated with the Palace Lands site include a brick chimney foundation, a rectangular sub-floor pit, a series of three ditches, and two fencelines defined by postholes (Table 1). A ravine that had been filled during the 18th century was identified in at least two of the 1999 test units (contexts 263 and 271). There were also twenty miscellaneous features uncovered at the site, including treeholes, unidentifiable features, and animal burrows.
Table 1: Major features, The Palace Lands Site.
Feature Description Feature Number Associated Contexts Feature TPQ
Fill: 12, 50, 53, 54, 58, 129; Wood lining: 55; Feature cut: 13
Brick Chimney Foundation
Mortared brick: 4;Feature cut: 7
No Date Assigned (NDA)
Fill: 5; Feature cut: 6
Fill: 19, 30, 40,47,48,49, 132, 204, 205, 206,207, 210, 211, 213, 214, 216, 217, 221, 234, 276-279, 282, 295, 298, 318; Feature cut: 20
Fill: 51, 274, 275, 296, 310, 319; Feature cut: 273
Fill: 248, 299-304, 311-317, 320, 325; Feature cut: 249
No feature number
Posthole 21/22c + postmold 255/256c; Posthole 33/34c + postmold 257/258c; Posthole 35/36c + postmold 251/252c + postmold 253/254c; Posthole 38/39c + postmold 342/343c; Postmold 41/42c; Posthole 43/44c + postmold 266/267c; Posthole 45/46c + postmold 268/269c; Posthole 56/57c; Postmold 66/67c; Posthole 68/69c + postmold 340/341c; Posthole 228/229c + postmold 336/337c; Posthole 230/231c + postmold 338/339c; Posthole 74/75c + postmold 305/306c; Posthole 141/142c + postmold 307/308c; Posthole 145/146c + postmold 143/144c; Posthole 280/281c (not excavated); Posthole 283/284c + postmold 321/322c; Posthole 335/334c; Posthole 287/288c + postmold 332/333c; Posthole 289/290c + postmold 329/328c; Posthole 291/292c + postmold 330/331c; Posthole 293/294c + postmold 327/326c
Posthole TPQ of 1787; Postmold TPQ of 1775
No feature number
Posthole 76/77c + postmold 194/196c; Posthole 82/83c + postmold 198/199c; Posthole 92, 225, 226, 93c + postmold 193/195c; Posthole 94, 297, 309, 95c; Posthole 107/109c + postmold 125/240c; Posthole 100/101c; Posthole 235/236c + postmold 237/238c; Posthole 323/324c + postmold 344/345c; 152/153c
Posthole TPQ of 1720; Postmold TPQ of 1775
Note: Twenty miscellaneous features (treeholes, animal burrows, and unidentified features) are not included in this list.
Note: Posthole and postmold “cuts” are indicated by a “c” following the context number.
Summary of research and analysis
As with nearly every site that has been discovered within a stone’s throw of CW’s Historic Area, CW scholars have conducted previous archival research that is relevant to this project that dates back as far as the 1930s. Thus, biographical information on John Coke, Botetourt, and Dunmore is related in various historical and architectural reports; most of these are available on-line (Buchanan 1961; Bullock 1931; Daniel 1946; GPHN 1930). There are also three additional works that contain significant information regarding the Palace Lands: Graham Hood’s (1991) book on the Governor’s Palace, a CW resource book (Willis et al. 1998) on slavery in Virginia, and Patricia Gibbs’ (1980) research on the use of the Palace Lands acreage.
The following research summary attempts to combine the evidence gathered from both the archaeological and historical records in order to interpret the site’s chronology. In all, three phases have been determined for the Palace Lands site (Table 2).
During Phase I, enslaved Virginians owned by John Coke initially occupied the Palace Lands site from c. 1747-1769 when Coke operated a plantation upon the land. It was this group that was responsible for constructing the dwelling, fences, and ditches. In the absence of an architectural footprint, the dwelling may have been a loghouse with a dirt floor. Architectural historian Willie Graham suggested that the house could also have been either wood-framed or a loghouse supported on wood piers. This two-room structure had a central chimney that heated one room, while the root cellar was located in the second room. The size of the structure can only be estimated as about 10 ft. wide by 20 ft. in length.
Some time after settling the site, its residents constructed the post-and-rail fences which were built in alignment with the house. Eighteenth-century artifacts were recovered from most of the postholes along both fencelines which indicate that cultural activities at the site preceded the construction of both fences. The posthole TPQs (see Table 1) suggest that the south fence may have been built first. In any case, site inhabitants took more care in preserving the north fence by digging ditches along it in order to drain water away from the posts. The fences and ditches clearly extended further in both directions, so accurate measurements of their lengths are not possible. The portion of the north fence that was uncovered measured about 100 ft., as did the line of three ditches.
Table 2: Site chronology, The Palace Lands Site.
Phase Date Identity of Occupants
Enslaved Virginians belonging to John and then Samuel Coke who lived at the site.
Late eighteenth century to early nineteenth century; TPQ 1787
Possibly tenants of Samuel Smith McCroskey’s who lived in the site’s vicinity.
Mid to late nineteenth century; TPQ 1880
Possibly tenant farmers who lived adjacent to the site and leased the land from Robert M. Garrett, Van F. Garrett, or both.
Note: The phases discussed in this table were assigned by the P.I. (Maria Franklin) and are not related to the DAACS Phases discussed on the Chronology page.
The presence of a dwelling emphasizes the domestic nature of the site, as does the artifact assemblage. Among the finds are ceramic tablewares and teawares, and wine bottle and pharmaceutical glass. It seems clear that the site was occupied by a kin-related group that included a mother and her child or children. Gender-related objects typically associated with women during this era such as beads and sewing implements were found at the site. The latter includes scissors, thimbles, sewing needles, and hundreds of straight pins. The presence of a female child or children is suggested by doll parts and two small thimbles that could only fit a child’s finger. Some other notable finds include the nozzle portion of a bone enema syringe, bone fan fragments, a slate pencil, a whirligig (child’s toy), and three finger rings (one is metal with a silver wash, and two are carved bone). In contrast to the domestic-related artifacts, work-related objects constitute a small percentage of the assemblage despite the fact that the site was part of a plantation. The artifacts associated with work include a rake head, draw knife, and whetstone. There are harness buckles and hooks which do suggest the use of draft animals (or horse-and-carriage transport) at the site. Yet Coke’s inventory provides better evidence for the kinds of work that was performed at the site than do the artifacts.
The Phase I occupation ended in c. 1769 with the sale of Coke’s plantation. There are no sources that suggest that Coke’s enslaved Virginians were sold to the Council and continued to live on the land. Instead, during the late eighteenth to early nineteenth centuries, another home was settled somewhere in the site’s vicinity (see Table 2). A small number of ceramic ware types (primarily pearlware and American blue-and-gray stoneware) representative of this period were found within a handful of features. The site’s occupants were probably tenants who leased the land from Samuel Smith McCroskey who owned the Palace Lands from 1790-1816. It is not known at this time whether McCroskey owned enslaved blacks.
The feature MCDs all cluster around the 1760s, which suggest that they were filled during Phase I. Yet pearlware and American stoneware sherds also appeared in one ditch (F04), one posthole, and two postmolds. Moreover, a Virginia half-penny (TPQ 1773) was retrieved from the root cellar. What this evidence suggests is that during Phase II, McCroskey’s tenants continued to use the fences, and tossed their refuse into features. Repair posts were evident along both fencelines and these mends may have been made during Phase II. The fences rotted in place some time after 1775 and the Phase II occupation ended after 1787.
Ceramics and glass fragments recovered from the plowzone indicate that the site was once again reoccupied during the late nineteenth century (see Table 2). A seriation of the ceramic assemblage for this phase, which includes whitewares, yellow wares, and ironstone wares, produced a MCD of 1912. Moreover, the glass assemblage includes colorless, non-leaded glass (1864 TPQ); some of these fragments are pressed glass tablewares, embossed glass, or Ball Mason canning jars. Decalcomania whiteware sherds provide a TPQ of 1880 for this phase.
In the late nineteenth century, the site was part of Garrett Farm. Dr. Robert M. Garrett was deeded the property some time prior to 1866 and upon his death his son, Dr. Van F. Garrett, inherited the farm land in 1883. Both Garretts lived in town. In fact, their residence was formerly that of John Coke, hence the “Coke-Garrett” house. Thus, the land was probably leased to tenants. The tenants were evidently a family since the assemblage points to a domestic occupation. The location of their homestead is not known since archaeologists did not identify any features associated with Phase III. Further, since the artifacts were distributed across the excavation unit and were found in Phase I and II tests beyond the excavation, it can only be stated that the Phase III settlement was somewhere near the site.