Colonial Williamsburg Digital Library

Middle Plantation Ditch Feature

In comparison with the extensive historical and archaeological research on eighteenth-century Williamsburg, only scant traces of the seventeenth-century community of Middle Plantation are known. Among the few notable Middle Plantation- period sites recently studied are the John Page Site (Metz et al. 1998), the Nassau Street Ordinary (Levy 2000), the College Landing site (Edwards 1987), and Rich Neck Plantation (McFaden et al. 1999). A small number of isolated Middle Plantation-period archaeological features have also been encountered throughout Colonial Williamsburg's Historic Area in the course of investigations of later eighteenth-century sites such as the Shields Tavern site (Brown et al. 1990:37), the Nicholas-Tyler site (Noël Hume 1980:2), and the Charlton House site (Kostro 2004:34) (Figure 24).

In 1978, archaeologist Eric Klingelhofer identified a boundary ditch dating to the Middle Plantation-era at the Peyton Randolph site during his investigation of Structure A (Figure 25). Orientated along a linear northwest-southeast axis under Structure A's brick foundation (23 degrees east of south), the ditch cut into sterile subsoil and was sealed by a layer of dark brown silt loam believed to be buried topsoil dating to the period of Structure A's construction. The fact that Structure A was underpinned where its foundation's crossed over the ditch suggested that the ditch fill had not yet fully compacted at the time when Structure A was built. The ditch measured approximately four feet wide and fourteen inches deep. Only a small section of the ditch along the exterior of Structure A was excavated at this time. The only artifacts recovered from the ditch were a small quantity of nail fragments (Klingelhofer 1978).

This same ditch was re-exposed in 1982. From its known location under Structure A, excavators followed the ditch southeast along its orientation as it 32. Figure 24. Known Middle Plantation archaeological sites and features. Figure 25. Middle Plantation ditch under Structure A. 33. ran under nearby Structure B, Planting Bed 4 and the east wall of Structure E (see Figure 25). The portions excavated in 1985 contained brown to dark brown sandy loam (Munsell 10YR3/4). The only artifacts recovered were oxidized nail fragments and bottle glass. The feature continued along this orientation out of the excavation block. In association with the boundary ditch were three postholes. Although located at the same stratigraphic level as the ditch, the postholes did not form any meaningful pattern, and how they related to the ditch remains unclear. Further southeast along its projected orientation, the ditch was again encountered near the southwestern corner of Structure R, once again located at the lowest stratigraphic level, intruding only subsoil (Edwards et al. 1987:116).

Most recently, yet another section of the same ditch was discovered during the investigations of Structure N. Yet again the ditch was found at the lowest stratigraphic level, running under where the building's foundations later stood. The best-preserved portion of the ditch (28BJ-21-22) was in the northwest corner of the excavation block, preserved under eighteenth-century soil strata (28BJ-12- 13) and a marl walkway (28BJ-7, -9, -11). The ditch continued southeast out of the excavation block, but for how much further remains unknown (Figure 26).

Similar to prior experiences with the ditch, the fill consisted of dark brown (Munsell 10YR3/4) sandy loam, and contained few artifacts. The age of the feature was again inferred from the fact that it was found at the lowest stratigraphic Figure 26. Middle Plantation ditch under Structure N. 34. level and because it was not orientated in the usual north-south, east-west plan that most Williamsburg-era structures, roads, fencelines, or other man-made landscape features are aligned to.

The ditch at Peyton Randolph is most similar to the Middle Plantation-period ditches found at the Shields Tavern and the Nicholas-Tyler sites, interpreted as boundary markers delineating property divisions. Similar ditch features are also well-known from other seventeenth-century sites from the surrounding area including: Martin's Hundred (Noël Hume and Noël Hume 2001), Kingsmill (Kelso 1984), and Governor's Land (Outlaw 1990). In most of these examples, the ditches were similarly interpreted as property boundary markers or as internal property divisions. Excavations at the John Page site and the Charlton House site also revealed seventeenth-century ditches. These features were distinct from the others by their sloped bases which were hypothesized to have been designed for transporting water. Meanwhile the sloped ditches at the Page and Charlton sites are believed to be components related to small industrial complexes. As Middle Plantation grew more prominent in the late seventeenth century, it is to be expected that various industrial and trade related sites would have been established in the vicinity to support Middle Plantation's small yet growing population.

As already mentioned, the Peyton Randolph ditch's lack of a slope suggests it functioned as a boundary marker rather than as a mechanism for transporting water. Nevertheless, no matter what its function may have been, the identification of the early ditch feature at the Peyton Randolph site is an important clue regarding the seventeenth-century landscape at Middle Plantation. The ditch represents one of only a handful of features that can be reliably attributed to the poorly understood pre-Williamsburg period (pre-1699). The fact that ditches are among the most common seventeenth-century features discovered in Williamsburg suggests that they were significant features on the Middle Plantation landscape but have only been minimally considered by archaeologists. The growing database of Middle Plantation archaeological evidence is beginning to illustrate how in addition to the construction of structures, the residents of Middle Plantation were deliberately manipulating their environment via the excavation of these ditches for a variety of purposes in order to transform the area's wilderness into a landscape suitable for settlement that was consistent with the needs and pre-conceptions of the colonists.

East Storage Building (Structure N)

The original brick foundations of East Storage Building (Structure N) were first exposed during the excavations carried out by Francis Duke in 1938 (Figure 27). The following excerpt from Duke's report is a description of the foundations as he discovered them:

Separated by 3 feet from the nearest storehouse foundation and by 13 feet from the marl walk, remains of another outbuilding were found opposite the southeast addition top the second kitchen. Four 9-inch walls were traced, defining a building which measured about 10 x 14 feet. Paving of brickbats was found inside. Earlier work under parts of the 35. Figure 27. East Storage Building (1938 excavations). north and east walls indicates that the building was destroyed and rebuilt on the same site, or perhaps that the same building was raised to a higher level.

Two short 9-inch foundations project from the west wall south of center but are not bonded with the wall. They may have supported a small stoop, or a narrow flight of steps.

The brickwork in this foundation is of as early character as that in the first storehouse foundation. The east wall of this foundation lines with that of the third storehouse. Hence, the life of the building or buildings here indicated may have spanned that of all three storehouses.

This building cannot be positively identified but it may be the shed mentioned in Humphrey Harwood's accounts for 1777

(Duke 1939:10-11).

The storehouse referenced by Mr. Duke is actually the latest in a series of three dairy buildings (Structures P, Q, and R) superimposed on one another located just a few feet north of the East Storage Building. Duke suggests the brickwork in the East Storage Building's foundation matches the earliest of the three dairy buildings. Based on this observation, Edwards et al. (1988:162) suggested that the East Storage Building, along with Structure R, were constructed as part of the reorganization of the yard by Peyton Randolph following the addition of the mid-section of the main house in the 1750s.

36.

Figure 28. Reconstructed East Storage Building (on left) in the 1980s.

Drawing upon the archaeological evidence recorded in 1938, a Colonial Revival inspired reconstruction of the East Storage Building was completed ca. 1940. Originally, the replica building served as a kitchen, during the main building's restoration. Later the building was used as a garden shed, and during the four years of archaeological excavations of the Peyton Randolph yard from 1982-85, it served as a tool shed and field office for the archaeologists (Figure 28). Because the building was in use, no archaeological investigations of the building site were undertaken at that time. In early 2003, the reconstructed East Storage Building was razed as part of the most recent efforts to restore the Peyton Randolph property that began in the mid-1990s. Demolition of the building was necessary because the recent architectural research had shown that the Colonial Revival-styled building was inconsistent with current understandings of colonial outbuildings. In its place the construction of a new reconstruction was begun in December 2004, this time following the same level of historical authenticity made use of in the recent reconstructions of the Peyton Randolph kitchen, covered way, smokehouse, and other "new" outbuildings.

Prior to commencing with the second reconstruction, archaeological investigations of the site were carried out to search for additional evidence of the building that would be useful for its reconstruction. Removal of the sod exposed the top of the demolished reconstructed East Storage Building's brick foundation walls (ca. 1940) within a few inches below modern grade. The remnant foundation was rectangular, with a small brick pad attached to the west wall that supported a set of steps that led into the building. Both the foundation wall and the brick pad were set directly on subsoil. A builder's trench (28BJ-3) for the structure cut 37. Figure 29. William & Mary students excavating unstratified silt loam (Master Context 1). through a deep layer of unstratified silt loam (Master Context 1) that is attributed to the early twentieth century (Figure 29).

The unstratified silt loam averaged 50 cm in depth and extended across the entire area of the trench, terminating upon sterile subsoil. Artifacts recovered from the layer included materials of mixed antiquity, varying in age from the eighteenth through the early twentieth century (Table 1). The dates of the later artifacts coincide with Francis Duke's investigation, indicating the fill was probably deposited as backfill to cover his excavations in preparation for the Colonial Revival building's construction. In most cases when excavated holes are backfilled, the soil that was originally excavated from a hole is used to re-fill the hole. During the early part of Williamsburg's Restoration, however, it was not uncommon to bring in fill from other parts of town to backfill open excavations. At least one early photograph of a large dump truck depositing a load of soil in the rear yard of the Peyton Randolph property in the late 1930s documents this practice (Figure 30). Consequentially, any artifacts recovered from the suspected backfill cannot be reliably attributed to the property without corroborating their provenience by cross mending them with artifacts from undisturbed contexts. A labor intensive and time consuming process, no such cross-checking was carried out in 2003, thus the artifacts recovered from the backfill over the East Storage Building have not been included in the interpretation of the site.

Sealed below the 1940 backfill were the traces of the original East Storage Building. The initial evidence consisted of a pair of parallel linear features (25BJ-4) that cut two inches into the sterile subsoil located approximately 2 feet 38.

Table 1.
Artifacts from East Storage Building Excavations

Artifacts	No. of Fragments
Tin-Enameled:
English Delft	37
Faience	1
Unidentified Tin-enamelled	8
Coarse Earthenwares:
Black Glazed Redware	7
Border Ware	2
Buckley	1
Colono	2
Derbyshire	4
Iberian	1
N.Midland Slip	5
Pennsylvania	1
Red Sandy	1
Staff. Mottled	1
Yorktown/Local	3
Other	7
Refined Earthenwares:
Creamware	69
Pearlware	52
Whiteware	1
Other	5
Stonewares:
Dry-Bodied	2
English	2
Fulham	4
Nottingham	2
Staff. Brown	3
Westerwald	4
White Salt Glazed	20
Yorktown/Local	1
Porcelain:
Chinese	37
English Bone China	1
English Soft Paste	1
Porcelaineous	1
Other	1
Tobacco Pipes:
Bowls	7
Stems	29
Glass:
Wine Bottle	283
Case Bottle	1
Pharmaceutical	1
Table	33
Other Container	47
Faunal:
Bone	352
Shell	79
Personal:
Button	2
Buckle	1
Other:
Iron Alloy	25
Mirror Glass	1
Curtain Ring	1
Architecture:
Nails	216
Window	158
Plaster	9
Marble Fragments	1
Farm/Horse Tack:
Harness Buckle	1
TOTAL	1534

39. Figure 30. Dump truck at Peyton Randolph. to the east of where the foundations of the reconstructed building's west wall stood. The two trenches measured approximately eight inches wide, and were spaced one foot apart from one another forming a narrow berm of undisturbed subsoil between them (Figure 31). At right angles to the berm were two nine inch long projections of subsoil located to the south of the berm's mid-point. Both trenches were filled with the same soil matrix and variety of artifacts as the 1940 backfill, indicating that they were filled in as part of the backfilling of the site that took place between Duke's excavation and prior to the construction of the reconstructed building.

The fill date for the two parallel trenches suggests that they were most likely excavated by Duke's 1938 excavators. It was common practice at this early stage of archaeological research in Williamsburg to expose both the exterior and interior sides of foundation walls upon their discovery. This was done in order to document the extents of the walls, and to allow for the detailed recording of a wall's interior and exterior elevations. The parallel trenches are imprints left in the ground of their efforts. As the walls of the building were exposed, excavators dug just below the bottom-most course of brick in the wall, resulting in the parallel features observed in 2003. Furthermore, the two nine inch projections of subsoil formed at right angles to the berm were similarly formed as Duke's excavators dug around the footings for the steps leading into the building that Duke described. Although no bricks were found in situ between the two trenches, the width of the berm between them very closely matches the width of a 1½ coarse thick brick wall. Although physical evidence of the original East Storage Building had yet to be found, the evidence of Duke's excavations in this area provided strong evidence of the East Storage Building's original location that was just as informative as the original foundations would have been toward identifying the building's location.

Evidence of the East Storage Building's original east wall was considerably more substantial than the ephemeral evidence of its west wall. Located 40. Figure 31. Plan Map of East Storage Building features. approximately two feet to the east of where it had been reconstructed, the intact evidence consisted of an eight-foot length of the building's east foundation wall, including the building's northeast corner (Figure 32). Also present was the east wall's southeast corner, although the section of the east wall connecting the two corners had been displaced by the planting of a pecan tree in the twentieth century. The identification of the two corners also allowed for the documentation of the building's north and south walls, of which no other evidence survived. Unfortunately, none of the original builder's trench from any of the walls had survived which would have been very useful in determining a more precise age for the building.

From the above described evidence, the exact location of the East Storage Building could once again be defined. Missing still, however, was evidence of the brick bat flooring that Duke described within the building's interior. Undoubtedly, it was also removed during his investigation or the 1940 reconstruction. Duke also noted, "earlier work under parts of the north and east walls" of the building. However, no evidence of any differences in the wall's brickwork was evident. More than likely, this evidence was also destroyed during the 1940 reconstruction (Figure 33).

In addition to pinpointing the building's location, it was also hoped that a more accurate age for the East Storage Building could be determined. As already 41. Figure 32. East Foundation Wall, East Storage Building. Figure 33. Overview of the ruins of the reconstructed East Storage Building. 42. mentioned, however, no builder's trench or other archaeological evidence of the building's age was identified, leaving only the mention of the building in Humphrey Harwood's account book (1777) as the only hint of the building's age. How long before that date was the building built may never be conclusively known, although it is logical to assume, as Edwards et al. suggest, that the building was constructed as part of the general reorganization of the lot following Peyton Randolph's inheritance of the property in 1755.

Finally, no evidence of a specific function for the building was determined. The most common types of outbuildings in Williamsburg during the eighteenth- century were dairies and smokehouses, both of which have already been located for properties late eighteenth-century occupation. Humphrey Harwood listed the building as a shed in 1777, suggesting the building was designed probably for general storage rather than some other specialized function.

Marl Paving

The only soil stratigraphy relating to the East Storage Building's eighteenth- century construction and use were two lenses of eighteenth-century marl paving found to the west of the building (Figure 34). The exact relationship between the marl lenses and the East Storage Building is unclear because the relevant Figure 34. Marl Paving. 43. stratigraphic evidence was destroyed by Duke's excavations around the original building, although the artifacts embedded in the marl suggest the paving and the East Storage Building were probably contemporary.

The surface of the most recent marl lens (28BJ-7) was identified directly below the modern backfill (28BJ-2) that covered Duke's excavations. The lack of intervening stratigraphy indicates that the lens was exposed at the time the backfill was deposited, although no mention of it exists in Duke's report. The only artifacts recovered from the lens were three wine bottle fragments, an unidentified iron alloy fragment, and three animal bones. The small quantity of artifacts does not lend itself to accurately dating the lens, although they are generally of colonial date.

Below the first marl lens was a thin lens of ash (28BJ-8, -10), also with a small number of colonial-era artifacts embedded within it. Artifacts from the ash lens included: German Westerwald stoneware, tobacco pipe fragments, table and wine bottle glass, window glass, nail fragments, and animal bones. Based on these artifacts a TPQ of 1700 was calculated for the layer. A small fragment of Leighton's Patent glass (TPQ = 1864) was also recovered, but the inconsistent TPQ of the find as compared to the other artifacts, suggests that it was probably intrusive rather than actually accumulating during the ash lens' deposition.

The ash lens accumulated over the earlier layer of marl paving (28BJ-9, -11). The artifacts from the second layer of marl included wine bottle glass, window glass, and nails. In general, the artifacts are of eighteenth-century vintage, but again are not diagnostic of a more precise age. The most revealing evidence regarding the marl's age was its stratigraphic position. The fact that the lens was laid directly over the early eighteenth-century layer of Brown Sandy Loam (28BJ-12, -13)² indicates the lens was likely laid down around mid-century. This date suggests that the marl may be related to the East Storage Building which was presumably built around 1755.

The deposition of these layers is also an important clue towards reconstructing the grounds around the East Storage Building. The lenses accumulation over the early Brown Sandy Loam, and the early dates of the artifacts embedded within them, both suggest that the ground at the entrance to the East Storage Building was covered with the marl shortly after its construction. The identification of the later layer of marl overtop of the first marl (and separated by an accumulation of ash) further suggests a resurfacing of the original layer.

These findings, however, contrast somewhat with the findings of the 1982- 85 excavations. During those excavations, two layers of marl paving were also identified, but were attributed to subsequent occupations. The earlier layer (Marl Walkway #2), was attributed to the Hornsby occupation (1784-1800), and consisted of a broad area of well-packed marl paving located between the kitchen (Structure L) and the East Dairy complex (Structures P, Q, and R) to the north of the East Storage Building. The second layer (Marl Walkway #1) is 44. even later, dating to the Peachy occupation (1800-1860), and consists of only a narrow path of marl rather than the broad paving of the earlier layer (Edwards 1988:123-126).

It seems logical that the Marl Walkway #2, which covered a broad area across the yard, probably also extended to the door of the East Storage Building. Based on this assumption, it is tempting to attribute the early marl lens found in 2003 as a continuation of Marl Walkway #2 found in 1982-85. The artifacts, however, indicate otherwise, suggesting a date for the marl thirty years earlier than that attributed to Marl Walkway #2. The only way to know conclusively would be to expand the excavations and trace the full extent of the marl to see if the two sections match up to one another.

Pending further excavation, three possibilities exist that would explain the age discrepancy. The first is that the paving excavated found around the East Storage Building was not part of either of the two pavings found in 1982-85. Instead, only the area immediately around the East Storage Building was covered in marl shortly after the buildings construction, possibly to facilitate drainage around the building's entrance. A second possibility is that the paving was part of a previously unknown pathway leading from the main house and kitchen to the East Storage Building at mid-century. Given the extensive nature of prior excavations which would likely have encountered such a path if it existed, this scenario is probably the least likely possibility. The third possibility is that paving around the East Storage Building is indeed part of Marl Walkway #2, also known the Marl Yard spread identified by Edwards et al.

The differences in the dates between the paving excavated in 1982 and 2003 may be due to the very small area of paving that was excavated in 2003, resulting in too small an artifact sample to reliably determine the its age. Excavation of a larger sample would presumably result in a larger artifact sample that was better representative of the time period, and may have resulted in a later date that was more consistent with the evidence from past excavations. On the other hand, the lack of any intervening stratigraphy between the paving and the Brown Sandy Loam (TPQ = 1730) suggests otherwise. The marl's position directly over the layer of early eighteenth-century Brown Sandy Loam corresponds well with a mid-century deposition date for the marl. From this perspective, it is suggested that the age of Marl Walkway #2 is might actually be older then than previously determined. The age of the Marl Walkway #2 was originally estimated based on artifacts imbedded within the layer which were presumed to have been deposited at the same time as the marl. However, because the marl was laid as surface, it is possible that the marl was exposed for a long period of time during which artifacts had the opportunity to embed themselves into the marl under the constant foot traffic that could be expected within an area of constant activity. As a result, the artifacts recovered from the marl may actually represent artifacts accumulated during the uselife of the marl rather than having been deposited as part of the initial paving. The embedded artifacts could have found their way into the marl at any time during its use, thus the marl artifacts actually represent 45. a mixing of artifacts from the years when the marl was in use rather than from its deposition. A further complication to this scenario is the possibility that the marl may have been periodically swept clean of debris during its use. The implication that sweeping has on the artifacts is that the majority of artifacts that were deposited on the marl would have been swept away before they could be trampled into the matrix of the layer. Instead, the artifacts found in the marl might only represent the very end of the marl's usage, rather than its cumulative use life, or its beginnings.

Without further excavation and analysis, the evidence recovered in 2003 suggests at a minimum that the marl paving around the East Storage Building was put down after mid-century, and probably coincided with the construction of the East Storage Building. What remains to be seen is if the layer is part of the Marl Walkway previously found across much of the rest of the rear yard, and to when does the marl actually date to. What is clear, however, is that the paving was in place for a period long enough to warrant renewal with a second layer.

Old Dairy (Structure G)

Evidence of the Old Dairy (Structure G) was found directly north of the reconstructed Smokehouse (Structure H) in 1955 during cross-trenching of the property by architectural draftsman James Knight. At the time, the proximity of the then extant Skillman House prevented the full exposure of the building (Figure 35). Renewed archaeological excavations in 1982-85 re-exposed the foundations found by Knight, although the eastern end of the building was again not uncovered due to the proximity of the Skillman House foundations (Figure 36). Although 75% of the structure's footprint was exposed in 1980s, none of the intact eighteenth-century soil layers and features within or around the structure were excavated at the time due to scheduling and budgetary constraints. The fact that the site was not covered by the late eighteenth-century marl path (Marl Walkway 2) suggested that the building was still standing when the marl was laid down. Based on the available evidence from 1985, it was estimated that the Old Dairy was most likely built in the third quarter of the eighteenth century, during Peyton Randolph's ownership, and survived into the nineteenth century (Edwards et al. 1988:140).

The site of the Old Dairy was revisited in 2003 with the objective of determining whether the building could be more reliably dated to the Peyton Randolph's occupation of the property. A significant advantage in 2003 was the ability to expose the entire structure simultaneously for the first time including the section previously sealed below the Skillman House's foundations. In the interim between the excavation in the 1980s and those in 2003, the Skillman House was demolished, thus permitting access to the Old Dairy's eastern end for the first time (Figure 37).

In spite of the two previous archaeological investigations, the construction of the Skillman House, and the installation of modern utilities, the remains of the Old Dairy were found remarkably intact (Figures 38 and 39). The evidence 46. Figure 35. Old Dairy excavations (1955). 47. Figure 36. Overview of the Old Dairy (on right) (1985). Figure 37. Old Diary foundations following initial exposure in 2003. 48. Figure 38. Plan map of Old Dairy features. Figure 39. Old Diary site excavated. 49. consisted of a square brick foundation measuring 10 x 10 feet with a small area of brick paving within the northwest corner of the building's interior. Sometimes known as a "single dairy," this type of dairy had a single interior space, as opposed to a "double dairy" which had two rooms (Smith 1982:20-21). Both types were common forms found on the eighteenth-century sites in the Chesapeake.

Excavations further revealed the Old Dairy's brick foundation was laid directly upon subsoil without the benefit of either a footing or builder's trench. Within the foundation, the bricks were not laid in any consistent pattern or bond, and included both English bonded and row-locked bricks. Due to changes in the natural topography, the northeastern corner of the foundation was underpinned with an additional course of brick to compensate for the changes in the ground slope (Figure 40).

The remnant of interior brick paving (28HB-94) was located in the building's northwestern corner. Although only a small area of the paving survived intact, the paving probably covered the entire area of the building's interior when the building was in use (Figure 41). The paving consisted of bricks and brick bats of varying sizes and shapes, including at least one trapezoidal-shaped "compass" or "well" brick, a specialized type manufactured specifically for use in the lining of a well (Figure 42). The inconsistency of the paving material and the variety of brick sizes suggests that paving was done with salvaged or left over building materials (Edwards et al. 1988). One possibility is that the compass brick was Figure 40. Brick underpinning, Old Dairy. 50. Figure 41. Archaeologist Buddy Paulett examines a remnant of brick paving in the Old Dairy. left over from the construction of the nearby well, located only a few feet to the west of the Old Dairy.

The identification of the brick paving within the Old Dairy was an important clue toward interpreting the building's function, as dairies typically had their interior's paved in order to help keep their interiors cool (Smith 1982:39). Cool temperatures were necessary to separate the cream from the milk and to prevent dairy products from spoiling. In addition to the brick flooring, the interpretation of the building as a dairy was further reinforced by the recovery of fragments of an earthenware pan used in the processing of milk and cream found on the floor of the building (Figure 43). Eighteenth-century milk pans were large open pots with sloping sides, typically made of coarse earthenware clay and sealed with a lead glaze to allow them to hold liquid. After milking a cow, the milk was poured into a milk pan and placed in a dairy to cool. As the milk cooled, the cream would rise to the top, and the milk would settle to the bottom. Once the cream and milk separated from one another, the cream was processed into butter or cheese, and the milk was stored in another vessel for drinking or cooking, often a stoneware jar. Although most often used in the processing of diary products, milk pans were also used for a variety of other purposes including clothes washing and cooking.

The soil upon which the Old Dairy foundation and floor were laid was a brown sandy loam (Master Context 9) dating to the early eighteenth century which had been previously been identified across much of the rear yard (see 51. Figure 42. Compass brick used in Old Dairy interior paving. Figure 43. Milk pan recovered from Old Dairy. Edwards 1988:133-134). Two thin lenses of ash, both with substantial quantities of artifacts (Master Contexts 5 and 7) were also identified under the southwest corner of the dairy foundation, but lying over the brown sandy loam. The fact that these layers were all found under the brick foundation and floor meant that their accumulation also had to pre-date the building's construction. The most recent artifacts recovered from the layers included fragments of locally produced coarse earthenware closely resembling the work of Yorktown potter William Rogers. Rogers began production around 1725, thus the layer with the Rogers pottery had to date to that year at the earliest (Table 2). Other artifacts included delftware, colonoware, borderware, white salt-glazed stoneware, Fulham stoneware, Chinese porcelain, tobacco pipes, wine bottle glass, pharmaceutical glass, table glass, nails, window glass, and faunal remains. All the artifacts consistently date to the early eighteenth century, and are most likely associated with the occupation of the lot in the 1720s when William Robertson owned the property, but the individual buildings were probably rented out to tenants.

The Rogers-type pottery recovered from the pre-Dairy layers also indicates that the Old Dairy was probably built after the property was sold to Sir John Randolph in 1724. The lack of artifacts post-dating the mid-1720s from under the building suggests that its construction likely occurred not long after that date. The construction of the building during Sir John's occupation (1724-1737), is at least a quarter century earlier than previously assumed (see Edwards et al. 1988:142).

Further evidence of the Old Dairy's early date is its orientation. In comparison with the orientations of the surrounding buildings, all of which are attributed to the second half of the eighteenth-century, the Old Dairy is slightly askew. The variation in orientation between the Old Dairy and the rest of the buildings suggests 52.

Table 2.
Artifacts from Pre-Dairy Contexts

Artifact	No. of Fragments
Earthenware:
Delftware	34
Yorktown	13
Colonoware	26
Border Ware	11
Stoneware:
WSG	10
Yorktown	2
Staffordshire Brown	1
Fulham	5
Dry-Bodied	1
Porcelain:
Chinese	10
Tobacco Pipe:
Bowl	11
Stem	11
Glass:
Wine Bottle	297
Pharmaceutical	30
Container	33
Table	9
Architectural:
Window Glass	15
Window Lead	1
Nail	103
Marble	1
Other:	22
Faunal:
Bone	678
Shell	17
Eggshell	1
TOTAL	1342

they were built at different times, while buildings with like orientations were probably built around the same time. Moreover, according to archaeologist Andrew Edwards, who directed the 1982-85 fieldwork at the site, the brickwork in the Old Dairy foundations closely resembles the brickwork found in the foundation of the nearby second kitchen (Structure J) (Andrew Edwards, personal communication). Structure J was built in the first quarter of the eighteenth century, facing onto North England Street, but was later incorporated into the Randolph kitchen. The similarity between the two buildings suggests that they were both built around the same time. Whichever building was first, the latter one appears to have been built to match the former's orientation.

Determining how long the Old Dairy was in use from the archaeological evidence was complicated by the lack of a discernible demolition layer to mark the building's deletion from the landscape. Instead, the duration of the building's use was inferred from the soil layers of ash and kitchen debris post-dating the building's demise. These layers were found across the interior of the building, and covering over portions of the building's foundations. The fact that they were found directly over the foundation ruins indicated they were deposited shortly after the building was torn down. Numerous kitchen ceramics dating the second and third quarters of the eighteenth century were recovered from the layers, along with an abundant quantity of animal and fish bones (Table 3). Assuming that the building was constructed in the late 1720s, the artifacts recovered from the 53.

Table 3.
Artifacts from Post-Dairy Contexts

Artifact	No. of Fragments
Refined Earthenware:
Astbury	2
Red Agate	1
Creamware	11
Pearlware	3
Whiteware	1
Earthenware:
Delftware	147
Yorktown	52
Colonoware	84
Borderware	19
N.Midlands Slip	4
N.Devon Gravel	1
Red Sandy	2
White Sandy	1
Derbyshire	1
Buckley	1
Iberian	2
Staffordshire Mottled	2
Other	9
Stoneware:
Dipped White Salt-Glazed	2
White Salt-Glazed	28
Fulham	15
Yorktown	1
Dry Bodied	1
Westerwald	2
Nottingham	5
Other	7
Porcelain:
Chinese	49
English Bone China	3
English Soft Paste	2
Fireclay:
Crucible	4
Tobacco:
Bowl	22
Stem	82
Glass:
Wine Bottle	823
Case Bottle	20
Pharmaceutical	52
Container	24
Table	29
Architectural:
Glass	155
Nails	465
Delft Fireplace Tile	7
Window Leads	3
Other:	83
Faunal:
Bone	1063
Marine Shell	46
TOTAL	3336

layers of ash and kitchen debris indicate the building stood for an approximate thirty-year period lasting into the 1750s or 60s.

The reason(s) for the lack of a demolition layer associated with the Old Dairy is not entirely clear. The most likely possibility is that as part of the Old Dairy's demolition, any salvageable materials were removed from the site to be reused elsewhere on the property as part of the rear yard's reorganization. The simultaneous construction of so many outbuildings after Peyton Randolph inherited the property would have required a large quantity of materials, thus the reuse of anything salvageable would have been an effective way to reduce the financial and resource strain of multiple construction/renovation projects.

54.

Figure 44. Kitchen vs. table ceramics.

The layers of ash and kitchen debris that covered the Old Dairy ruins likely originated from the nearby kitchen. The origin of the layers is inferred from the fact that most of the ceramics recovered from the post-dairy contexts were related to food storage and preparation rather than table wares which would have come from the house rather than the kitchen (Figure 44). In addition, a horizontal distribution of the artifacts from these layers indicates a general pattern of increasing density from the southwest to the northeast corners of the excavation block (Figure 45). This corner was the closest to the kitchen building. Refuse from the kitchen was being dumped over the foundation ruins, possibly to fill the depression left by the building's demolition. This would have been an effective way to dispose of unwanted trash and would have efficiently filled an unwanted hole. In addition, by filling in the depression, the ground was also made ready for the laying of a brick path that would extend out from the smokehouse and to the nearby well.

One artifact that particularly stood out among the post-dairy artifacts was a metatarsus (foot) bone from a chicken with its spur removed (Figure 46). The vast majority of chickens raised in eighteenth-century Williamsburg were for food or egg production, although cocks (males) were also kept for fighting. The removal of spurs was one of several surgical modifications done to prepare cocks for fighting along with the removal of the comb and wattle (dubbing), the shortening of feathers (trimming), and the sharpening of beaks. The reason for removing the naturally formed spurs was to fit artificial ones that could be attached with leather straps to the cock's foot. The artificial spurs were made of silver, steel, or iron and were much larger, and thus much more lethal than natural ones. Alternatively, metal spurs could be fashioned to fit over the natural ones, or if metal spurs were not available, the tips of natural spurs were sometimes sharpened to a point with a knife. By making the cock spurs more lethal, the duration of cockfights were hastened, thereby increasing the number of fights that could take place in an afternoon or evening (Powell 1988:23-25, 63).

55.

Figure 45. Distribution of all artifacts from post-dairy contexts.

Figure 46. Photograph of modified chicken metatarsus.

56.

The long history of cockfighting extends back at least 2500 years, probably originating in southeast Asia. From there it diffused to China, the Middle East, Western Europe and the British Isles. The cockfight, however, did not spread to any great extent to sub-Saharan Africa or the Americas until after European efforts to colonize these regions (Dundes 1993:23-24). In colonial America, cockfighting was widely practiced, particularly among men in the southern colonies (Carson 1958:151-164; Powell 1988). In 1724 Hugh Jones commented, "The common planters leading easy lives don't much admire labour, or any manly exercise, except horse-racing, nor diversion, except cock-fighting, in which some greatly delight"(Jones 1956:84) .

Cockfights were hardly an exclusive pastime of colonial Virginia's well-to- do planters. The poor and middling were also avid enthusiasts, as were many enslaved African Americans (Powell 1993:362). Elkanah Watson's description from 1787 of a cockfight in Southampton County, Virginia illustrates the ethnic and socio-economic diversity of its devotees from Watson's point of view as a wealthy planter: "in the center of which was arranged a large cock-pit; surrounded by many genteel people, promiscuously mingled with the vulgar and debased"(Watson 1856:261-262) . In light of this co-mingling, Richard E. Powell suggested that cockfighting represented a "community practice" in which "men of different social classes co-operated through the year-round cycle of husbandry and conditioning of cocks … and provided an element of shared experience for urbanites and agriculturalists"(Powell 1993:262) . In contrast, it is interesting to consider the eighteenth-century cockfight as an opportunity for enslaved African Americans to safely challenge the existing social order. As anthropologist Clifford Geerz has observed among the twentieth-century Balinese, the cockfight "is like playing with fire only not getting burned. You activate village and kingroup rivalries and hostilities, but in 'play' form, coming dangerously and entrancingly close to the expression of open and direct interpersonal and intergroup aggression"(Geertz 1973:440) . In the context of eighteenth-century Virginia, the cockfight may have similarly functioned to "activate" racial, social, or economic "rivalries and hostilities" brought on the enslavement and coerced labor of Africans.

The participation of enslaved African Americans in cockfighting also extended to their employment as fighting cock managers on behalf of the gentry. A poultry breed book kept by the Cocke family of Surry County, Virginia mentions several enslaved African Americans engaged in the supervision of fighting cocks (Powell 1993:370). The management of fighting cocks was a highly specialized occupation that required detailed knowledge of breeding, diet, veterinary medicine, surgical modification, physical training and an understanding of the effective use of artificial cockspurs (Powell 1993:364-367). By acquiring the specialized knowledge that related to cockfighting, African Americans were taking hold of skills that were prized by both the gentry and the enslaved which could in tern be capitalized on to improve their standing in social order of white society, as well as, within the internal enslaved community.

Cockfights were often publicized well in advance of the event taking place. Announcements published in newspapers like the Virginia Gazette regularly indicated 57. that cockfights often took place at taverns or near public places such as courthouses. First-hand accounts further indicate that cockfights often attracted a considerable amount of gambling among the spectators, both free and enslaved (Carson 1958:151-164). In Williamsburg, at least one area tavern is known to have hosted cockfights according to a notice published in the Virginia Gazette:

On Tuesday next will be fought, at the George and Dragon, in Williamsburg, a Match of Cocks, for Ten Pistoles the first Battle, Five Pistoles the Second, and Two Pistoles and a half the Third, andc. As Likewise several other Matches.(Virginia Gazette, February 14, 1750/1).

In seventeenth- and eighteenth-century England, cockfights took place on circular stages about twenty feet in diameter, surrounded by a barrier to keep the birds in the ring (Carson 1958:162). For Virginia, however, there are no comparable descriptions of cockpits, although excavations behind Marot's Ordinary (later Shields Tavern) in Williamsburg revealed an unusual circular feature dating to the first quarter of the eighteenth century that may have been the remains of a cockpit (Brown et al. 1990: 59-60).

For those excluded from taverns, or for those who couldn't afford them, cockfights were also informally held in a variety of alternative venues. For enslaved African Americans, due to restrictions on time and resources, cockfights probably took place not far from where they were either living or working, such as within slave quarters, within outbuildings or in nearby woods and fields. At Nomini Hall on Virginia's Northern Neck, for example, Philip Vickers Fithian noted that he witnessed "a Ring of Negroes at the Stable, fighting Cocks" (Fithian 1957:96). The lack of any diagnostic archaeological features to mark where cockfights took place, however, has limited archaeologists' ability to place these fowl battlegrounds on the colonial landscape.

Archaeological evidence of enslaved African Americans keeping fighting cocks was also found at the Rich Neck slave quarter site. Excavations recovered a modified metatarsus from an eighteenth century context that is very similar to the example from the Peyton Randolph site (Steve Atkins, personal communication). While documenting the presence fighting cocks within slave quarter compound, no evidence was found indicating where the actual cockfights may have been held.

Ethnographers such as Geertz were among the first researchers to propose that cockfights are more then just entertainment or distractions from everyday life. As already noted, Geertz suggested that cockfighting "is fundamentally a dramatization of status concerns"(1973:437) . Alan Dundes, using a psycho-analytic approach, concludes that "cockfighting cannot be understood without being seen as an exemplar of a more comprehensive paradigm involving male gladiatorial combat"(1993:32) . Archaeological evidence from the Levi Jordan plantation in post-bellum Texas is also suggestive of the symbolic status of fighting cocks within the plantation's African American community. Excavations within the 58. plantation's political leader's cabin recovered a pendant made form the spur of a fighting cock. The pendant was the central element of a necklace comprised of glass beads, several ebony rings, an elaborately decorated metal ring, and a large number of amber beads that marked this individual's status in the community (Brown 1994:115). In a similar example, cut marks on the tarsometatarsus of a great horned owl found at a slave quarter at George Washington's Mount Vernon suggests the removal of the talon (claw) from the leg. The talon could then be used as an object to be suspended around an individual's neck-possibly as a status indicator. In that same assemblage, a raccoon baculum (penis bone) with evidence of having been tied with a cord around its posterior end may have been used a symbol of fertility or masculinity (Atkins 1994:84).

Although no evidence of a cockpit was found at the Peyton Randolph site, the recovery of the sawn spur is strong evidence that someone in the household was at least keeping fighting cocks. The fact that the bone was found within the yard suggests that the animal's keeper was probably one of the enslaved African's working on the property either for themselves or in the service of Peyton (women almost never participated). By extension, it is also possible that cockfights may have been held somewhere on the property, possibly within one of the outbuildings on the lot. Regardless of where the cock fights were taking place, the recovery of the cock spur is potentially rare physical evidence of enslaved African American social life in Williamsburg. Equally intriguing is the possibility that the sawn cock spur was used to symbolically mark an individual's relative status within Williamsburg's African American community. Documentary records sometimes indicate how different enslaved persons were perceived or favored by the slave owning class (e.g., naming a favorite servant), but an emic perspective on status within the enslaved community is rarely recognized either in the documentary or archaeological records.

Smokehouse (Structure H)

In the course of the excavations at the Old Dairy, a brick path (28HB-13) associated with the adjacent smokehouse (Structure H) was also found overtop of the kitchen debris and the Old Dairy foundations (Figure 47). Remnants of the path were also exposed in 1984, although at the time they were mistakenly interpreted as part of the Old Dairy's floor. Upon closer inspection in 2003, it was realized that the bricks were indeed not part of the floor, but were laid overtop of the Old Dairy foundations as part of a path leading out from the adjacent smokehouse to the nearby well. The fact that the bricks were laid over the Old Dairy foundations could only mean that the path was put down after the Old Dairy had been demolished. In addition, the path was also laid over the kitchen debris layers. Artifacts recovered from the contexts directly below the path include fragments of creamware, a refined earthenware ceramic type not found in Williamsburg until 1762. The fact that the creamware was recovered from below the path indicates the path had to have been laid down after that date.

The path was a significant discovery that had implications relating to several aspects of the smokehouse's interpretation. First of all, the path is an important 59. Figure 47. Brick path leading out of the Smokehouse. clue toward understanding the building's orientation. Previous archaeological and architectural investigations of the building had failed to determine the location of the building's entrance. As a result, during the planning of the building's reconstruction, it was necessary to estimate the location of its entrance from comparisons of other late eighteenth-century smokehouses in Williamsburg. These comparisons overwhelmingly suggested that the most likely place for the door would have been through its south elevation, facing toward the main house and kitchen. The discovery of the path extending out from the building's north elevation, however, indicates the opposite. Rather than facing the house, the building's entrance was located along its northern elevation, facing away from the main house and kitchen.

The discovery of the path, and the resulting relocation of the door to the north side of the Smokehouse, also had implications for refining the age of the building. Previous excavations had determined only a broad date range for the construction and use of the Smokehouse. Excavations within its interior indicated only that the building was built sometime between 1720 and 1779. Based on the evidence from the Old Dairy and brick path excavation, however, the construction of the smokehouse can be narrowed to after mid-century and only after the demolition of the Old Dairy. Datable artifacts recovered from under the path indicate that the path was not laid until after 1762 at the earliest. By linking the brick path with the operation of the smokehouse, the construction of the smokehouse can similarly be narrowed to after 1762. Previously, the two buildings were assumed 60. to be contemporary to one another, although now it is clear that the two buildings could not have been standing at the same time since the Old Dairy would have blocked access to the smokehouse's door.

Lumber House (Structure F)

The recent excavations of the Old Dairy also turned up possible evidence relating to another neighboring building. Directly to the north of the Old Dairy foundations, the fragmentary remnants of a brick wall consisting of a single course of bricks laid as headers oriented east-west was discovered. Initially, the wall was thought to be part of the ruins of the Old Dairy, although how it related to the building was unknown. It was subsequently realized that the wall did not relate to the Old Dairy at all, and instead was part of an addition or shed attached to the east side of the neighboring Lumber House (Structure F). Located to the northwest of the Old Dairy, the Lumber House was first excavated by James Knight in 1955. Initially constructed sometime after 1763, the building was remodeled and expanded in the nineteenth century. A new foundation was laid as part of the rebuilding, although portions of the original foundation were incorporated into the renovation (Edwards et al. 1988:53).

Renewed archaeological excavations of the Lumber House site in 1996 found three partially intact foundation walls relating to the north, east and west walls of the original structure. Completely absent was any evidence to pinpoint the south wall's location. Neither the south wall, nor the building's corners were ever identified; although significant quantities of brick rubble were found along the southern end of the building which likely marked the wall's the approximate location.³ Unfortunately, the intrusion of a later posthole, and the placement of a modern concrete walk had significantly disturbed the southern end of the building, and as a result, the precise location of its southern foundation wall has not been more conclusively determined except for its general location indicated by the brick rubble (D.A.R. 28BD Field Notes).

Interestingly, the conjectural south end of the Lumber House is directly west of the east-west wall found north of the Old Dairy (Figure 48). The alignment and proximity of the two suggests that the wall uncovered north of the Old Dairy may be a continuation of the Lumber House's south wall for a shed or lean-to attached to the east side of that building. Additional excavation to further explore the possibility of an attached shed to the Lumber House might yield further evidence of the shed's existence. Mitigating the usefulness of further exploring this area is the fact that the area, however, was also the site of the early twentieth-century Skillman House, the construction and demolition of which in all probability destroyed any archaeological evidence of a shed. Nevertheless, the alignment of the Lumber House's south wall with the wall north of the Old Dairy does indeed suggest the strong possibility that the two are part of the same wall.

61.

Figure 48. Relationship between the Lumber House foundation and hypothesized shed foundation.

East-West Fence Line

In addition to digging around the foundations of colonial-era outbuildings, excavations were also carried out to explore the possibility of a colonial-era fence that may have stood between the reconstructed East Dairy (Structure Q/R) and the reconstructed Smokehouse (Structure H). Excavation of a pair of exploratory trenches ultimately revealed a line of four contemporary postholes post-dating 1762 between the two outbuildings (Figure 49). Excavation of two of the posts revealed that both post-dated to 1762, temporally linking them with Peyton Randolph's reorganization of the property in the 1750s and 60s.

All four fence line postholes were found buried under several inches of twentieth century landscaping fill; and two of the posts were disturbed by twentieth- century features. The posts were irregularly spaced, although the fact that they were all in a single line suggests that all the posts belonged to the same fence. Beginning at the east end of the line, the first post (28BJ-47,-48) was located at the northwest corner of the East Dairy. The post measured 66 cm across from north to south, and 48 cm from east to west. The eastern end of the feature, however, was buried under the eastern baulk of the excavation trench, preventing the full measurement of the posthole's east-west dimension. The feature was either square or rectangular shaped, depending on how far east the feature continued under the 62. Figure 49. East-west fenceline. baulk. In profile, the posthole had straight sides and a flat bottom (Figure 50), at a depth measuring 35 cm below its top edge. The posthole fill consisted of single homogenous layer consisting of a brown (Munsell 10YR3/2) sandy loam that cut a lens of clay and oyster shell (28BJ-53) that dated to after 1762. Artifacts from the posthole fill included: English Delftware, white salt-glazed stoneware, Fulham stoneware, wine bottle glass, tobacco pipe stems, animal bones, window glass, and nails. A TPQ of 1730 was assigned for the feature based on the variety of dip-molded wine bottle glass recovered, although the fact that it cut the 1762 lens of clay and oyster shell indicates that it too had to date to after that year.

The eastern edge of second post (28BJ-63-64) was located 110 cm (3.61 feet) to the west of the first post. The interval between the estimated centers of the two posts was 185 cm (6.07 feet). The post was rectangular shaped with rounded corners, measuring 70 cm across from north to south, and 80 cm from east to west. Unlike the first post, the full extent of the second post was exposed in the trench which permitted the full measurement of all its dimensions. Similar to the first post, the profile of the second post also had straight sides and a flat bottom; its fill consisted of a single homogenous layer of brown sandy loam; and it cut the same clay and oyster feature cut by the first posthole (see Figure 50). Artifacts from the post included English delftware, a single tobacco pipe stem, a single wine bottle fragment, and two animal bones. A TPQ of 1650 was assigned for the feature based on the type of wine bottle glass. Like the first post, however, the fact that the post was stratigraphically located above the oyster and shell lens that had a 1762 meant that the fence post also had to date after that date.

The third post (28BJ-65-66) in the line was located 180 cm (5.94 feet) to the west of the second post. On its east end, the brick edging for the nineteenth- century brick path that led down the center of the rear yard covered the post's edge. On its west end, a recent utility trench cut through the feature. Assuming that the preserved portion of the posthole represented its approximate center, the interval between the centers of the second and third posts measured 230 cm (7.54 feet). From north to south the post measured 60 cm across, and 28 cm from east to west through what remained of the post. Time constraints prevented its excavation, although the feature fill appeared to be the same brown sandy loam that was found in the previous two postholes.

63.

Figure 50. Fence posts 1 and 2 profiles.

The utility trench that impacted the third post's western edge was known to be several feet wide, and likely cut through the forth fence post in the line. As a result, the excavation of the area where the forth post was probably located (between five and eight feet west of the third post) was abandoned given the likelihood that little archaeological evidence remained of the post. Efforts shifted toward defining fifth post's location. Remnants of posthole number five (28BJ- 93-95) were found at the southeastern corner of the reconstructed smokehouse. The recent reconstruction of the smokehouse also destroyed all traces of the post's western edge. Like all the other posts, the remaining portion of the feature was rectangular with rounded corners. The post measured 53 cm from north to south, and 70 cm from east to west. The distance between the third and fifth posts measured 396 cm (12.98 feet), leaving room for a post halfway between the two which would have left an average interval of 198 cm (6.49 feet) between the third and forth posts, and the forth and fifth posts. Time constraints also prevented the excavation of the fifth post.

The fence across the rear yard of the property is an important component of Peyton Randolph's organization of property that had been unrecognized until now. Prior to the fence's discovery, the archaeological evidence of Peyton Randolph's extensive re-organizational efforts had been primarily architectural evidence unearthed in 1982-85 relating to the outbuildings. The fence is the first non-architectural evidence associated with Peyton Randolph's reorganization of 64. the property when he inherited it from his parents. The archaeological and architectural evidence suggests that when Peyton became the head of the household, he decided to enact a series of changes to the property possibly in an effort to mark its change of ownership.

The link between changes witnessed in the archaeological or architectural records and inter-generational household succession has been studied by historical archaeologists since the 1970s (Beaudry 1999: 120). More recently, archaeologists have further recognized that the succession of a particular property from parent to offspring can be the catalyst of landscape change, and that the majority of these changes typically occur within five to ten years before or after the transfer of a the property occurs (Groover 2004: 25-43). This certainly appears to be the case at the Peyton Randolph site where a large number of improvements to the main house, outbuildings, and yard all appear to have occurred within a few short years of one another following Peyton's inheritance.

While household succession may be the catalyst for change, the reasons behind these changes are diverse. Groover (2004) proposes individual autonomy, technological innovation, and the adoption of new architectural styles as the potential reasons for the changes. In explaining the changes they observed in the archaeological record at the Peyton Randolph site, Edwards et al. (1988) suggested that Peyton Randolph, as the new owner of the property, made a concerted effort reorient his house toward the emerging power center of Williamsburg, combined with an attempt to Georgianize his house in accord with current stylistic trends in architecture and landscape design. Edwards et al. further interpreted the back yard as a single space organized symmetrically along the same Georgian architectural and landscaping principles that guided the expansion of the main house. They suggest that the path leading from the back door of the house and through the yard served as a central axis to which the yard's outbuildings were symmetrically planned around. Because the timing of the house's expansion was coincident with the changes in the locations of outbuilding and landscape features in the rear yard, it was logical to conclude that the layout of the rear yard was similarly influenced by Georgian ideals.

The placement of the fence through the middle of the yard, however, does not fit with the symmetrical ideals clearly evident in the main house's renovation. The fence divided the backyard into northern and southern halves, within which related outbuildings were clustered. Within the area to the south of the fence, the portion of the yard closer to the Randolph's home, were the outbuildings associated with the maintenance and service of the main house, and the open space between them likely served as a work yard. The layout of the yard and buildings suggests that it was planned as a work area where all the daily chores related to serving the Randolph family could be effectively attended to, including food processing and storage, household laundry, and cooking.

Meanwhile, the area north of the fence was reserved for building's associated with plantation or agricultural related work. The buildings located to this side of the fence similarly reflect the agrarian specialization of the north half of the yard. The outbuildings located here include: the Lumber House, Granary, and 65. various sheds and storage buildings, all of which are more closely associated with agricultural rather than domestic tasks.

Rather than organizing the yard according to aesthetic ideals, its division along related tasks suggests that practical considerations of spatial economy, environment, and social control appear to have bad a larger influence on how the yard and buildings were laid out than previously recognized. Similar patterns of spatial differentiation and specialization of work spaces have also been documented on rural plantations in the area (Epperson 1999). Like his rural counterparts, by dividing the yard in the way, Randolph accomplished two related goals. The separation of household and agricultural tasks would theoretically increase the efficiency of the servants and slaves assigned to those tasks by limiting the number of distractions. The division also increased the Randolph family's level of control over the mobility and interactions of their enslaved workers. Rather than asserting social or political legitimacy through symbols imbedded in architecture and landscape, Randolph's authority was established through his ability to compartmentalize the spaces that enslaved Africans and African Americans were living and working in.

The lone outbuilding that initially does not appear to conform to the proposed sub-division of the yard and outbuildings into household and plantation related work areas is the smokehouse. As a building designed purposefully for the curation and storage of meat products, it might be expected that a smokehouse would have been located within the south yard's domestic complex. At the Peyton Randolph site, the smokehouse was located along the boarder of the two yards; with its door facing north into the agricultural portion of the lot. The reason for smokehouse's door on its north side may have been so that unsmoked meat products could be easily brought into the smokehouse from the agricultural yard where the animals were butchered without having to trespass into the domestic yard. At the same time, however, the brick path that led from the smokehouse to the kitchen and well also tied the building back into the south yard's domestic complex. In this case, the smokehouse occupied an intermediate position within the yard, one that is accessible to both sides of the lots, but not fully part of either.

Clay and Oyster Shell Feature

In addition to the fenceline evidence, several additional important features were located between the two outbuildings. At the eastern end of the fenceline, located between the first and second posts was a thin lens of olive brown (Munsell 5Y5/3) mottled clay (28BJ-53) (Figure 51). A small number of artifacts were recovered from the lens including: creamware, pharmaceutical glass, wine bottle glass, animal bone, roofing slate, window glass, and nails. Creamware is not found in Williamsburg prior to 1762, indicating that the lens' accumulation had to occur after than date as well. However, the most striking aspect of the feature was the dense concentration of 47 oyster shells imbedded within the feature matrix, probably representing the refuse from the preparation of a meal.

66.

Figure 51. Clay and oyster shell feature.

Figure 52. Marl paving along the fence.

67.

Marl Paving II

Between the smokehouse and the second fence post was shallow marl feature (28BJ-60-61, -88) cutting into subsoil (Figure 52). The northern edge of the marl-filled feature corresponded precisely with the location of the fenceline, suggesting that the fence was standing when the marl was put down. The marl continued south under the trench's baulk for an unknown distance. Two sections of the marl were excavated, one at its eastern extent, the other at its western. The section in-between was disturbed by more recent intrusions. Uninterrupted, the marl would have measured 5.22 meters (17.14 feet) from east to west. Excavation of the marl resulted in the recovery of a number of artifacts, including English delftware, creamware, Fulham stoneware, Chinese porcelain, wine bottle fragments, window glass fragments, nails and oyster shell. A TPQ of 1762 for the marl fill was determined from the creamware fragment, further temporally linking the marl with the fence that was also installed after that date.

Edwards et al. (1988:179) identified a broad area of marl paving east of the kitchen complex, and south of the smokehouse dating to the Hornsby occupation (1783-1800). The marl features identified in 2003 could be continuation of that paving. Alternatively, the earlier TPQ of the marl suggests it could be part of an earlier as yet undocumented paving episode. Regardless of its date, what is clear is that the fenceline between the smokehouse and the east dairy was standing at the same time as the marl paving was put down.

Iberian Jar Fragments

The final find in the area of the fenceline was a very large portion of a large eighteenth-century Iberian storage jar, found lying over the thin clay lens with the oyster shells that dated to after 1762. The fragment was subsequently broken in situ into several smaller fragments, that when mended represented nearly one- third of a complete vessel (Figure 53). A complete example of the storage jar would have stood four feet tall and measured three feet across its width.

The combination of the initial fragment's large size, and the fact that it appears to have been broken in situ, suggests that the fragment might have been intentionally placed at that location, rather than being indiscriminately deposited. In most instances, ceramic fragments become part of the archaeological record as trash that is dumped into a convenient hole or is scattered in a surface midden. In both of these scenarios, the pottery is usually broken up into small fragments prior to deposition, and is further broken up when subjected to regular foot traffic. Accordingly the large size of the initial fragment prior to its in situ breakage suggests that it did not enter the archaeological record as the result of trash disposal, and was not trampled upon, and more likely found its way into the archaeological record as the result of some other process.

The location of the find may be the key to understanding its origins. The pottery fragments were found northwest of the first posthole and the corner of the East Dairy. The distance from the corner of the East Diary to the center of the pottery fragments measured approximately 64 cm, about the same distance 68. Figure 53. Iberian jar in situ (top) and reconstructed jar (from Noël Hume 1991:143) (bottom). Figure 54. Iberian storage jar with thickened rim and vestigial arched handles, coarse redware streaked externally with white slip, and lead-glazed on the inside. Common in the period c. 1745-80 Ht. 32½" 69. that the eve of the East Dairy's roof extended out from its wall. During heavy rains, rain water cascading off the roof the building would slowly erode the ground surface directly under a building's eve forming a trough or shallow ditch. These erosion features, often referred to as "drip lines," are sometimes found archaeologically around the perimeters of buildings. Over time, if this process is allowed to continue, the continued erosion of the soil away from a building's foundation would expose the building to possible water damage. Gutters that carry rain water off the roof and away from its foundation are a common method to avert this problem today, as they were in the eighteenth century. Another, albeit less sophisticated, way to cope with this problem is to place a large stone or other object along drip line, or against an exposed foundation. The stone or like-object would deflect falling water away thus preventing the erosion of the ground surface around a building. The large fragment of Iberian pottery that appears to have been carefully laid at the corner of the building may have been such an attempt to repel rain water falling off the roof of the East Dairy away from its foundation.

Front Yard Features

The final area investigated in 2003 was the front yard of the Peyton Randolph property between the house and Nicholson Street. Two large exploratory trenches were excavated in the front, one along the property line between lots 207 and 237, and the second just to the west of the main house's front door (Figures 54 and Figure 54. Trench A, front yard. 70. Figure 55. Trench B, front yard. 55). Unfortunately, no features that dated to the eighteenth century were identified in either trench. The oldest features were two posthole complexes related to a late nineteenth/early twentieth century picket fence that ran parallel to the front of the house. Evidently, the combination of modern landscaping, road grading, previous archaeology, and utility work had obliterated virtually all archaeological evidence of the front yard's eighteenth-century appearance.