Colonial Williamsburg Digital Library

Builder's Trench

In an effort to more specifically date the building, several sections of the Bassett Hall builder's trench were examined. A builder's trench is formed as a result of the excavation of the original cellar hole, corresponding to the dimensions of the building. Once the cellar hole is excavated, the foundation or cellar walls of a structure are seated within the bottom of the hole. After the walls are erected, the narrow gap between the exterior of the foundation wall and the exterior bank of the cellar hole is backfilled. It is this backfilled gap that is known as a builder's trench. Often, during the backfilling process, contemporary artifacts, usually consisting of construction debris, are deposited in within the builder's trench. The analysis and dating of these artifacts can sometimes provide a date for the filling of the builder's trench, and thus a TPQ for the construction of the house.

Portions of the Bassett Hall builder's trench (master context M1) were examined along the north elevation and south elevation of the original portion of the structure. In each of these areas, the uppermost section of the builder's trench had been partially truncated by later repair trenches, utility lines, and landscaping activities. On the east and west gable ends of the building, previous landscaping and utility lines had destroyed almost all evidence of the original builder's trench leaving intact only those sections of the trench that were hidden under the relatively shallow chimney bases (Figure 16).

The archaeological excavation of a builder's trench is a slow and difficult process owing to the depth of the feature, as well as the narrow width of the feature. As a result Figure 16. Profile of Bassett Hall builder's trench under east gable end chimney base. 32 the difficulties associated with the trench's excavation, and the time constraints imposed by the construction schedule, only three sections of the Bassett Hall builder's trench were excavated. On the east side of the north porch a 2 meter (6.56 foot) long section, and on the west side of the north porch a 1.2 meter (4.0 foot) long section, were excavated (Figure 17), while along the rear of Bassett Hall a 1 meter (3.28 feet) long section of the builder's trench was excavated east of the later dining room wing addition. In each section, the narrow width and extreme depth of the builder's trench prevented the full excavation of the lower half of the trench.

Figure 17. Profile of the north elevation builder's trench, west of the front porch.

In each of the excavated sections and in the profiles of the builder's trench under each of the chimney bases, the builder's trench fill consisted of a compact, yellowish-brown (Munsell color 10YR5/6) clay which measured between 25 cm (0.8 feet) and 45 cm (1.47 feet) at grade, and gradually tapered down to only 5 cm (0.16 feet) at the bottom of the cellar foundation wall, at approximately 183 cm (6.0 feet) below grade.

From within the three excavated portions of the builder's trench a total of thirty-three artifacts were recovered. Seventy-six percent of the recovered artifacts were architecturally related materials, consisting of small fragments of window glass and nails. Fragments of brick and plaster were also noted within the builder's trench fill, but not collected. The remaining twenty-four percent of the artifacts were a mix of different types of artifacts including two fragments of ceramics, one case bottle glass fragment, one wine bottle glass fragment, and one tobacco pipestem (with a 5/64th-inch bore diameter), a single lead alloy shot, and a fragment of unidentifiable hardware. The two ceramic fragments included a single sherd of coarse earthenware of undetermined origin and date, and a single sherd of Japanese porcelain.

Ultimately, the excavation of the builder's trench proved to be inconclusive with regard to determining the construction date of the house. Neither of the ceramic fragments could be attributed to a tightly-constrained time period, and the recovery of 33 only a single tobacco pipestem precluded the reliable determination of a viable date for the feature fill based on pipestem analysis, on account of an inadequate sample size.

Scaffolding Postholes

The builder's trench was not the only construction-related feature encountered at Bassett Hall. Other features included several small postholes along the exterior of the structure, which appear to be the locations of vertical support posts for scaffolding used during the framing and siding of the building. Each of the suspected scaffolding postholes was located 64cm (2.1 feet) north of the house foundation along the house's north elevation. Additionally, each of the postholes was cut into either subsoil or pre-construction soil layers, and was sealed by the earliest occupation layers.

The scaffolding postholes were identified on either side of the front porch, suggesting that it may have extended across the entire north-elevation of the structure (Figure 18). The easternmost scaffolding posthole (01AC-184-186) was roughly square-shaped, measuring 37 cm (1.21 feet) x 40 cm (1.31 feet) with a flat bottom. The feature fill consisted of light olive brown (2.5Y5/4) sandy loam soil. Once the fill was excavated, a 24 (0.79 feet) x 28 cm (0.92 feet) rectangular postmold became evident in the bottom of the posthole. A surprising sixty-nine artifacts, a large number for a posthole, were recovered from both the mold and posthole fill, although none of the artifacts were diagnostic to any specific date range. Similar to the artifacts from the builder's trench, eighty-eight percent were architectural, consisting of window glass and nail fragments. Non-architecturally-related artifacts included porcelain, wine bottle glass, and bone. Brick and plaster fragments were also noted from within the posthole fill.

The next scaffolding posthole (01AC-146-147) was located 2.5 meters (8.2 feet) to the west of the easternmost posthole. The posthole was square-shaped, measuring 30 cm (0.98 feet) across. The bottom of the feature was flat, and similarly showed no evidence of a postmold. The feature fill consisted of light olive brown (2.5Y5/4) sandy loam, and contained only five artifacts, including a fragment of window glass, two nail fragments, and two fragments of unidentifiable iron alloy hardware.

On the west side of the front porch, a third possible scaffolding posthole feature (01AC-227-228) from this line was identified. It was spaced roughly 5 meters (16.4 feet) west of the nearest posthole on the east side of the porch (01AC-146-147). Midway between these two postholes, another posthole may have once also existed, thus making the interval between the scaffolding posts a uniform 2.5-meters (8.2 feet). However, if such a post had once existed, the construction of the north porch in the 1930s has long since destroyed any evidence of it. The surviving posthole, west of the porch, was a small square-shaped feature measuring only 22 cm (0.72 feet) x 26 cm (0.85 feet), with a flat bottom. Once again, no evidence of a postmold was evident. Unfortunately, the majority of the feature had been previously destroyed by boxwood plantings in the twentieth century. The small amount of remaining feature fill consisted of a light olive brown (2.5Y5/4) silt loam with brick inclusions, but contained no artifacts.

Figure 18. Scaffolding postholes along north elevation.

Similar to the builder's trench, the scaffolding postholes are features relating to the construction of Bassett Hall, and are among the earliest features on the site. Although a date could not be established from the small quantity of artifacts recovered from the posthole fills, the relative age of the postholes is indicated by their stratigraphic position below the site's early occupation layers.

Early Shed Addition

On the west side of the house, two large and deep postholes were identified (Figure 19). Both postholes were sealed by early occupation layers or features, suggesting that they either pre-date or are contemporary with the earliest phase of the house's occupation. The southern posthole (01AC-93-94) measured 65 cm (2.1 feet) x 45 cm (1.48 feet) and was 50 cm (1.64 feet) deep, with straight sides and a flat bottom. The fill consisted of a dark brown (10YR3/3) silty clay with inclusions of brick, but no artifacts. The center of the posthole was located 2.2 meters (7.22 feet) west of the west wall of the house. The northern posthole (01AC-127-128), measured 70 x 55 Figure 19. Shed posthole features along west gable end. 36 cm (3.3 x 1.8 feet) and was 43 cm (1.41 feet) deep with straight sides and a flat bottom. The fill was a mottled olive brown (2.5Y4/4) sandy clay with brick inclusions. No artifacts were recovered from the fill. The center of the posthole was 2.4 meters (7.87 feet) west of the west wall of the house, while the interval between the two postholes measured 4.8 meters (15.75 feet).

Although the posts are slightly out of alignment from one another, their comparable size and relative stratigraphic position suggests that they are associated with one another. It is likely that they may have supported the frame of an early post-in-ground shed attached to the west side of the original portion of the house. Their stratigraphic position also indicates that the shed addition was part of the earliest phase in the development of the house. The shed may have been removed in order to install the new larger chimneys. Similar sheds added to the gable ends of houses are common throughout the Tidewater during the colonial era.

Early Occupation Layers

In addition to the builder's trench, the scaffolding postholes, and the shed postholes, several occupation layers related to the earliest occupation of Bassett Hall were identified along the west, east, and south elevations of the house. Unfortunately, modern utility installations and landscaping have destroyed these layers along the north elevation. To the west of the west end chimney, a layer of dark brown (10YR3/3) silt loam had accumulated (01AC-75 and 01AC-97) directly over subsoil. Ground disturbance along the foundation walls in this area has destroyed the nature of the relationship of the layer to the structure's builders trench. The depth of the layer measured 19 cm (0.62 feet) to 25 cm (0.82 feet) thick, and contained a total of one hundred fifteen artifacts. A little more than a quarter of the artifacts were architecturally related, consisting of window glass and nails. The remaining artifacts consisted of domestic sheet refuse, including ceramics, glass, animal bone, and casting waste. The majority of the ceramics were fragments of either Chinese porcelain or English delftware. Single sherds of Westerwald stoneware and coarse earthenware were also recovered (Table 1), along with the bowl of an imported tobacco pipe. Wine bottle and case bottle glass accounted for most of the glass from the layer. All the artifacts were typical of those found on late seventeenth- and early eighteenth-century historic sites, but a more precise date range could not be determined.

Located directly south of the southeast corner of the original structure were two early eighteenth-century occupation layers. The layer was deposited directly over subsoil, and consisted of a dark yellowish brown (10YR4/4) loamy sand, with a moderate quantity of charcoal inclusions (01AC-350). The layer covered an area 1.5 x 2 meters (4.92 x 6.56 feet), and sloped from west to east, ranging in depth from 0 to 5 cm (0.16 feet). Similar to the occupation layers on the west side of Bassett Hall, modern ground disturbances have obscured the relationship of the layer to the structure's builder's trench. One hundred ten artifacts were recovered from the layer, forty-five percent of which was architecturally related debris consisting of nail fragments, window glass, and window leads. Domestic sheet refuse artifacts accounted for fifty-five percent of the total, and included ceramics, glass, animal bone, and tobacco pipes.

The small quantity of ceramics (see Table 1) included Chinese porcelain, white salt glazed stoneware (TPQ=1720), English delftware, and Frechen brown stoneware. The glass artifacts included fragments of wine bottle glass, case bottle glass, table glass, and pharmaceutical glass.

Deposited on top of 01AC-350, and covering virtually the same area, was a 6 cm (0.20 feet) thick layer of light yellowish brown (10YR6/4) loamy sand (01AC-323). The types and quantities of artifacts recovered from the layer were very similar to the artifacts from the layer below. Architecturally related artifacts (nails, window glass and one window lead) made up thirty-seven percent of the total 167 artifacts found in this context. The remaining domestic artifacts included ceramics, glass, bone, tobacco pipes, and a small number of copper alloy artifacts. Among the ceramics from the layer, Chinese porcelain was the most common, with English delftware, Fulham stoneware, and white salt-glazed stoneware recovered in lesser quantities (see Table 1). A TPQ of 1720 was attributed to the layer, once again based on the recovery of fragments of white salt-glazed stoneware. Glass fragments from wine bottles, case bottles, and tablewares and pharmaceutical bottles were also found.

On the east side of the house, four distinct early eighteenth-century occupation layers and one pathway were identified. However, as has been the case in each of the early occupation layers around the rest of the house, the relationship between the 38 layers and the house had been obscured by modern disturbances. The earliest of the occupation layers was an 8 cm (0.26 feet) thick layer of light brownish gray (10YR6/2) sandy loam (01AC-170) deposited over subsoil. This layer contained eighty-five artifacts consisting primarily of domestic refuse while the remaining artifacts were composed of architecturally related debris, consisting entirely of nails and nail fragments. The domestic artifacts were made up of ceramics, glass, animal bone and a single piece of unidentifiable iron hardware. The ceramics found in the layer included Fulham stoneware, coarse earthenware, Yorktown-type stoneware, Chinese porcelain, and English delftware (see Table 1). A TPQ of 1725 was attributed to the layer based on the recovery of the Yorktown-type stoneware. All the glass from the layer was from wine bottles, except a single fragment of case bottle glass.

Deposited directly on top 01AC-170 was a 7 cm (0.23 feet) thick linear patch of crushed mortar (01AC-158). The area of the crushed mortar was approximately 150 cm (4.92 feet) wide and was orientated at a roughly 20-degree angle west of grid north, extending through the length of the excavation unit (Figure 20). The dense concentration of the crushed mortar along a distinctly linear course suggests that it was the remnants of a path or walkway situated on the east side of Bassett Hall. The vast Figure 20. Plan view of crushed mortar path around the east gable end. 39 majority of the 139 artifacts recovered were architecturally related, consisting of window glass and nail fragments. The few domestic artifacts included fragments of white salt-glazed stoneware (TPQ=1720), Chinese porcelain, English delftware, wine bottle glass, an unidentifiable piece of iron hardware, and a single fragment of animal bone (see Table 1).

On either side of the pathway, small deposits of soil were identified. The deposits were clearly deposited after the crushed mortar path had been laid down, and were probably contemporary with the use of the path. Situated to the east of the pathway, and overlapping the path's eastern edge, was a thin (5 cm/.16 feet) layer of pinkish colored (2.5YR6/2) ash and sand (01AC-36). The ash and sand fill may have been deposited in order to fill in a depression or "pothole" along the edge of the crushed mortar path. The majority of the one hundred twenty-eight total artifacts were architectural. Nails and nail fragments made up the majority of these artifacts, and window glass fragments composed the remainder. The domestic artifacts included local and imported coarse earthenwares, white salt-glazed stoneware, Westerwald stoneware, Yorktown-type stoneware, English delftware, tobacco pipestems, wine bottle glass, colored and leaded glass fragments, unidentifiable iron hardware, and animal bone. One of the white salt-glazed stoneware fragments had a dot/diaper/basket press-molded decoration along the edge of the plate, a pattern not produced until after 1740.

Along the western edge of the pathway, and also slightly overlapping the path's margin, was a small deposit of light olive brown (2.5Y5/3) loamy sand (01AC-165), ranging in depth from 2 (0.065 feet) to 8 cm (0.26 feet). In comparison to the path (01AC-158) and the ash and sand deposit (01AC-36), far fewer artifacts were recovered, totaling only forty-two. Architectural artifacts accounted for forty percent of the total, included nail fragments, window glass, and window leads. Domestic refuse accounted for sixty percent of the artifacts from the layer, and included ceramics, glass, animal bone, and unidentifiable hardware fragments. Ceramics represented in the layer included white salt-glazed stoneware, Westerwald stoneware, and Chinese porcelain. All the non-architectural glass from the deposit was wine bottle glass. A single fragment of an iron alloy horseshoe was also recovered from the deposit. The layer had a TPQ of 1720, once again based on a fragment of white salt-glazed stoneware.

The final early eighteenth-century occupation layer (01AC-71) extended to the east and to the south away from the east gable chimney, and sealed the mortar path (01AC-158) and the two contemporary soil deposits along side the path (01AC-36 and 01AC-165). The layer consisted of a brown (10YR4/3) sandy loam soil and a large quantity of burned oyster shell. Other evidence of burning included several concretions of melted glass and dirt. Architectural debris accounted for thirty-three percent of the one hundred nine total artifacts recovered from the layer, and included nails, fragments of window glass, and a window lead. Domestic refuse accounted for the other sixty-seven percent of the artifacts and included ceramics, glass, animal bone, and miscellaneous other artifacts. The ceramics were North Midlands slipware, white salt-glazed stoneware, an unknown variety of stoneware, Chinese porcelain and English delftware. Six fragments of wine bottle glass and five case bottle glass fragments 40 were also recovered. Among the miscellaneous artifacts were a straight pin and a fragment of worked bone that may have been part of a musical instrument. Although the latest artifact from the layer was white salt-glazed stoneware (TPQ=1720), the stratigraphic relationships point to a post-1740 deposition of the layer.

Analysis of the artifacts from these early occupation layers has revealed several interesting insights into the earliest occupation of Bassett Hall. The most important characteristic of the layers is that they all date to the second quarter of the eighteenth century. Remarkably, only a single artifact (molded white salt-glazed stoneware fragment), out of a total of 898 finds from the early occupation layers, definitely post-dates 1740. The remaining contexts all have TPQ's of either 1720 or 1725, or did not contain any datable artifacts. The proximity of the layers so close to the house strongly suggests that they are related to the occupation of the house. Accordingly, the archaeological evidence suggests an occupation of the house dating to as early as the second quarter of the eighteenth century.

Further analysis of the artifacts can help to identify the activities that took place on the site as the layers were accumulating. Architectural artifacts (nails, window glass and window leads) were compared to domestic artifacts (ceramics, vessel glass, animal bone, and personal effects) from each of the early occupation layers around Bassett Hall (Figure 21). The architectural artifacts most likely would have accumulated during periods of construction or renovation. Alternatively, the domestic artifacts most likely accumulated during the occupation of the house. Their accumulation was the result of the disposal of household trash as sheet refuse in the yard of the dwelling house, often accumulating over an extended period of time.

From six of the eight early occupation layers, domestic artifacts accounted for a larger percentage of the total assemblage. On average, the ratio of domestic artifacts to architectural artifacts favored domestic by a ratio of 2:1. Expressed in a different way, the percentage of artifacts related to domestic refuse averaged sixty-six percent Figure 21. Comparison of architectural debris vs. domestic refuse from the early occupation layers around Bassett Hall. 41 from these six contexts, compared to thirty-four percent for architectural debris (Figure 22), suggesting that domestic refuse disposal was the primary agent influencing the accumulation of artifacts within the early occupation layers. Accordingly, the artifacts suggest that the house already existed and was occupied while the layers were forming.

Figure 22. Architectural debris vs. domestic refuse.

In contrast, artifacts from the remaining two contexts exhibited the opposite pattern. In both of these contexts, the percentage of architectural debris significantly exceeded domestic refuse. On average domestic refuse accounted for nineteen percent of the assemblage, while architectural debris accounted for an average of eighty-one percent (Figure 23). Contexts with the highest percentage of architectural material included the crushed mortar pathway (01AC-158) and the deposit of ash and sand Figure 23. Architectural debris vs. domestic refuse. 42 (01AC-36) that was contemporary with pathway. The high percentage of architectural material from these two contexts suggests that they may have been formed during a period of either construction or renovation to the dwelling house. The contexts were deposited over a layer of domestic sheet refuse, indicating that the two contexts were deposited after the house had already been occupied, suggesting that the accumulation of architectural debris was probably not part of the initial construction of the house, but rather from a repair or renovation to it.

As already discussed, previous researchers have speculated Philip Johnson originally constructed that Bassett Hall around 1760. However, the results from the waterproofing excavations suggests that the original portion of Bassett Hall was constructed and occupied in either the 1730s or 1740s, which predates Johnson's acquisition on the property. Alternatively, the archaeological evidence for the houses construction and early occupation coincides with Thomas Bray II's ownership of the property, after he inherited the property from his cousin David Bray II in 1731.

Chimney Construction

The 2001 archaeological excavations also revealed significant information regarding the construction of Bassett Hall's two large exterior chimneys. Once the various soil layers around the east and west end chimneys were removed, it became immediately evident the chimneys were not part of the building's original construction. The excavations revealed that the construction of the chimney bases, and the construction of the cellar walls, were clearly two distinct and unrelated events. The chimney base brickwork and the cellar foundation wall brickwork are clearly not tied into one another (Figure 24). Examination of the brickwork indicates that the cellar foundation walls were already in place when construction of the chimney bases began, and in order to accommodate the new hearths, holes in the brickwork of the cellar foundation walls had to be made. Once the cellar foundations walls had been breached, the chimney bases were constructed matching the width of the openings into the building. The remaining gaps between the chimney base and the cellar foundation were filled with mortar. Because no hearth was planned for the basement, the chimney base was not extended down to the bottom of the cellar foundation wall. Instead, the new chimneys were set relatively shallow into the ground, extending only 60 cm (approximately 2 feet) below grade, and as previously mentioned, the chimney bases were built over the builder's trench for the cellar foundation wall. The brickwork for the chimney was laid in English bond with rubbed brick corners, on both the inner and outer corners, and bonded with a shell-based mortar. According to Colonial Williamsburg architectural historian Michael Bourne, the quality of the chimneys may be characterized as "superior" (Bourne, 2001, personal communication).

An intact builder's trench (01AC-76, -78, -109, -77) related to the construction of the chimneys was located for the east chimney (Figure 25). Although a nineteenth-century repair trench (01AC-32-33) had destroyed the upper portion of the chimney builder's trench, the bottom portion of the builder's trench had survived intact. The trench width varied from 20 cm (0.66 feet) along the south face of the chimney base to 48 cm (1.57 feet) along its east face. The trench had vertical sides and a flat bottom, and a fill that consisted of three overlapping lenses of yellowish brown (10YR5/6) 43 Figure 24. Interface between the west chimney base and the foundation wall. Figure 25. Profile of east chimney builder's trench. 44 sandy clay loam with inclusions of brick and mortar. A total of twenty-two artifacts were recovered from the trench fill, primarily consisting of nail fragments. Glass, animal bone, and a small fragment of English delftware were also recovered from the fill, but none of the artifacts were considered datable. The only clue to the age of the chimney was based on the recognition that the original excavation of the chimney builder's trench was cut through a post-1740 stratum of burned oyster shell and sand (01AC-71). Because the layer cut by the builder's trench post-dates 1740, so must the east gable end chimney.

Presumably, the two gable end chimneys were constructed at around the same time. Unfortunately, however, landscaping around the building during the restoration of Bassett Hall in the 1930s had destroyed almost all evidence of the builder's trench for the west chimney, except for the very bottom 15 cm (0.53 feet) to 20 cm (0.66 feet) section of its builder's trench (01AC-85-86). The remnants of this feature only cut subsoil, and not any early cultural layers. The fill consisted of a single lens of dark yellowish brown (10YR4/6) clay with fragments of brick and mortar. The only artifact recovered was a single non-diagnostic nail fragment. Assuming, however, that the two gable end chimneys were built around the same time, it may be assumed that the chimney was erected after 1740.

Although the excavation of the west chimney builder's trench turned out to be inconclusive with regard to ascertaining an age for the chimney's construction, its excavation did reveal an interesting additional feature. The removal of the builder's trench fill exposed a curious line of bricks in course (01AC-87), apparently forming a brick wall, seated on subsoil within the builder's trench cut (Figure 26). The extant line of Figure 26. Plan view of the brick wall within the west chimney builder's trench. 45 bricks was made up of both whole bricks and brick bats, but with no evidence of any mortar between them. The line of brick was located 33 cm (1.1 feet) west of the existing west chimney base and was oriented parallel to the existing chimney base. The northern extent of the line coincided with the northern edge of the chimney builder's trench, and extended south from that point for 140 cm (4.6 feet). The southern extent of the line was destroyed by boxwood plantings. Deposited between the chimney base and the line of bricks, and also within the chimney builder's trench cut, was a pile of brick and mortar rubble (01AC-90). Unfortunately, no artifacts were found in association with the brick rubble, and the function of the small stretch of bricks within the builder's trench remains unknown.

The two gable end chimneys appear identical, but during the excavations to underpin the chimneys, at least one major difference was identified. On the east end, the base is formed by a large solid rectangular brick pad. The west side contained no such pad; instead the chimney base was U-shaped with the open portion of the "U" facing into the building. The area within the interior of the "U" was filled with loose brick and mortar rubble, and capped with a solid brick pad at the level of the fireplace hearth on the first floor. The reason for the difference between the chimneys is unclear. One possible explanation may be that the construction of the "U-shaped" chimney base was an effort to save on expensive building materials.

Archaeology has clearly demonstrated that the two end chimneys were not constructed as part of the original foundation, and also that they were not added to the building until an unknown date after 1740. The fact the existing chimneys are not original to the building foundation ultimately raised new questions regarding the earliest phases of Bassett Hall's architectural history. In particular: where were the original chimneys? And why were they replaced?

Chimneys in colonial Virginia were typically located on the gable ends of houses, usually with at least one chimney on either end. End chimneys were constructed on either interior or exterior of the building's foundation walls. Alternatively, in some colonial examples, a building may have had only one large central chimney. In order to determine what type of chimneys were originally part of Bassett Hall, Colonial Williamsburg architectural historians examined Bassett Hall's cellar walls and ground level floor joists. The examination, however, did not locate any evidence to indicate the former presence of an interior central chimney or interior end chimneys (Bourne, 2001, personal communication). The implication of the architectural evidence is that the structure's original chimneys most likely would have been located on the exteriors of the building's gable ends. However, the archaeological traces of these early chimneys were non-existent. Most likely, the original chimneys were in the same location as the existing chimneys and were completely demolished in order to build the new ones. In the process of demolition and the subsequent construction of the existing chimneys, all evidence of the original chimneys was eradicated.

The archaeological record is similarly silent on the explanation as to why the new chimneys were constructed. Nevertheless, a possible explanation, can be inferred from an architectural study of the existing structure. As already discussed, the limited architectural evidence suggests that the original structure was only one-and-a-half stories tall, and was later raised to two stories (Buchanan 1980). Assuming that the 46 original structure was only one-and-a-half stories, the original chimneys would have been servicing a much smaller structure, and thus the original chimneys were likely to have been proportionally smaller as well. However, once the building was raised to two stories, the original chimneys would have been inadequate for heating the larger building. As part of the house's expansion, the original chimneys would have been dismantled, and the larger existing chimneys would have been built in their place. Given this set of circumstances, it is not unbelievable to imagine that any part of original smaller chimneys would have been entirely concealed by the construction of the larger second set of chimneys in their place.

Dining Room Addition

By 1782, the year in which the Frenchman's Map of Williamsburg was produced, a dining room had been added to the south (rear) of Bassett Hall, giving the building a "T-shaped" plan (Figure 27). Architectural evidence suggests that the dining room was originally 1½-stories, with an A-framed roof. Regrettably, the original date for the construction of the dining room wing has escaped comment in the records pertaining to Bassett Hall. Through archaeology, however, it was hoped that it would be possible to determine a date for the dining room addition. Unfortunately, a 1930s waterproofing of the dining room's cellar walls destroyed all remnants of the wing's builder's trench along its east wall. On the west side, the dining room was expanded two feet (61 cm) to the west as part of the restoration, and as a result, access to the west wall builder's trench was not possible without undermining the existing structure. Consequently, no determination based on the builder's trench could be made regarding the construction date for the dining room.

Figure 27. Detail of the Frenchman's Map showing the T-shaped plan of Bassett Hall.

Although the builder's trench for the dining room had not survived, or was inaccessible, its age could be inferred from indirect evidence. In particular, a long ditch (M2), oriented parallel to the dining room, was identified 180 cm (5.9 feet) west of the dining room's west wall (Figure 28). As the ditch approached the house, the ditch made a sharp and distinct ninety-degree turn to the west around the house. From there, the ditch followed along the exterior of the house and turned ninety-degrees back to the north, and continued around the southwest corner of the house. Unfortunately, the total extent of the ditch could not be traced, as the northern end of the ditch feature had been destroyed by nineteenth- and twentieth-century landscaping, and the southern end extended out beyond the bounds of the project area. Nevertheless, it was clear that the ditch clearly snaked around the house, rather than continuing through it, thus post-dating the house's construction.

Figure 28. Photograph of the ditch feature (M2) along the west side of Bassett Hall.

The ditch was 50 cm (1.64 feet) to 60 cm (1.96 feet) wide, and 13 cm (0.43 feet) to 25 cm (0.82 feet) deep. The ditch had a flat bottom and sloped sidewalls. The walls may have once been straight sided, but have subsequently eroded (Figure 29). The ditch feature may have served as an open drain around the perimeter of the house to draw water away from the house. No drainage ditch was located along the east side of the house, suggesting that the water problem was been limited to the west side.

Figure 29. Profile of early ditch feature (M2) along the west side of Bassett Hall.

The ditch contained two distinct soil layers overlaying one another. The bottom fill (M2) consisted of an accumulation of dark grayish brown (10YR3/2) silt loam. One hundred one artifacts were recovered from the fill, the vast majority of which were fragments of window glass. The percentage of architecturally related artifacts (window glass and nails) accounted for ninety-one percent of the artifacts from the lower fill. The high proportion of architecturally related material in contrast to the low quantity of domestic material suggests that the fill was accumulating during a construction/renovation period - possibly the construction of the dining room wing. The remaining artifacts consisted of a small quantity of ceramics and glass. All the ceramics from the ditch feature were fragments of English delftware, which does not have a meaningful TPQ.

The upper layer (M3) of fill in the ditch feature was a layer of brown (10YR5/3) silt loam with a heavy quantity of brick fragments. Three hundred fourteen artifacts were recovered from the layer. Architectural and domestic artifacts were found in nearly equal proportions. The dramatic decrease in architectural material in the upper fill suggests that this layer was formed while the dining room was in use. The architectural material from the upper fill was primarily window glass, followed by nails and a small number of window leads. The domestic artifacts include proportionately large quantities of wine and case bottle glass fragments and a small number of ceramic artifacts. The ceramic ware-types from the layer included molded white salt-glazed stoneware teapot fragments, Chinese porcelain, English delftware, and Westerwald stoneware. Based on several fragments of a molded white salt-glazed stoneware, a TPQ of 1740 was established for the ditch fill.

In summary, the location and orientation of the ditch suggests that it had been excavated after the construction of the original portion of the house, and probably after the addition of the dining room. The ditch functioned as an open drain in order to draw water away from the house foundations during which time soil accumulated within the open ditch, eventually filling it completely. The artifacts from the ditch indicate 49 that it was filled around 1750. Accordingly, the construction of the dining room occurred around mid-century as well, possibly coinciding with Elizabeth Bray Johnson's inheritance of the property in 1753.

East Bulkhead

The original entrance into the basement located under the dining room was gained through a bulkhead located along the east side of the addition. The archaeological excavations exposed the badly damaged ruins of the original bulkhead, the majority of which had been destroyed during the 1930s waterproofing of the dining room foundation, and the construction of a modern bulkhead entrance. The original bulkhead sustained further damage from the installation of several modern utility lines, and a recent backhoe excavation along the modern bulkhead footing. Consequently, only a very small section of the original bulkhead's north wall had survived intact (Figure 30). The archaeological excavations exposed a portion of the bulkhead's original builder's trench (01AC-296-297), which cut into subsoil and was backfilled with yellowish brown (10YR5/6) clay. No artifacts, save for a few mortar fragments were recovered from the builder's trench.

The bulkhead's brickwork clearly exhibited two periods of construction. The original portion of the bulkhead consisted of several courses of bright orange bricks bonded together with shell mortar, and set directly onto subsoil. The second area of brick was mortared into place directly on top of the original brickwork. These bricks were brownish-red in color, and also held together with a shell mortar. The second layer of brickwork Figure 30. Profile of the original (east) bulkhead entrance. 50 may represent either a repair or modification to the original bulkhead. Eventually, the bulkhead entrance in the east wall was relocated to the west side of the dining room, and the east bulkhead entrance was filled in. The TPQ of the fill over the east bulkhead is 1762, based on the presence of a fragment of creamware.

Middle to Late Eighteenth-Century Occupation Layers

Several occupation layers dating to the second half of the eighteenth century were identified to the east of the east chimney. An 8 cm (0.26 feet) thick layer of dark brown (10YR3/3) sandy clay with plaster (01AC-68) was uncovered that sealed the thin stratum of burned oyster shell and sand (01AC-71) previously discussed. The 60 artifacts from the layer included both architecturally related material and domestic refuse, and in virtually equal proportion. The architectural material included window glass, nail fragments, a window lead, and a door pintle. Domestic refuse included English delftware, black-glazed redware, Whieldon-type refined earthenware, white salt-glazed stoneware, Westerwald stoneware, a tobacco pipe stem, wine bottle glass, case bottle glass, animal bone, and a buckle. The TPQ for the layer is 1760, based on the recovery of the Whieldon-type refined earthenware fragments. A mean ceramic date for the layer was calculated at 1749.

On the west side of the dining room wing, a layer of brown (10YR4/3) silt loam (M4) was evident across the excavation area, and sealed the two layers of ditch fill (M2 and M3). The layer was highly dissected by a number of late nineteenth- and early twentieth-century utility lines. The layer ranged in thickness from 5 cm (0.16 feet) to 20 cm (0.66 feet) deep across the west rear excavation area. The layer was formed through the natural accumulation of soil and the deposition of household sheet refuse into the rear yard of house over a lengthy period of time-probably lasting throughout the eighteenth century. The long duration of the of the layer's formation is reflected in the quantity and diversity of artifacts recovered from the layer.

The layer (M4) contains the highest concentration of eighteenth-century artifacts found at the site. Some 1541 artifacts were recovered from the layer, with architectural finds accounting for forty-seven percent of the total artifact assemblage, while domestic refuse accounted for fifty-three percent. Most of the architectural artifacts were window glass fragments, with nails and window leads accounting for the rest.

Among the domestic refuse, glass accounted for the majority of the domestic refuse, and over third of the total assemblage. Wine bottle glass, case bottle glass, pharmaceutical glass, table glass were all represented in the assemblage. One hundred ninety-one ceramic fragments were found, and included thirteen different ware types. White salt-glazed stoneware fragments were by far the most common, accounting for forty-five percent of all the ceramics from the layer. Creamware, Chinese porcelain, and English delftware, were all also found in substantial quantities. A variety of other ceramic ware types were recovered in lesser quantities, and are individually listed in Table 2.

Standard archaeological dating techniques resulted in a variety of dates for M4. The TPQ for the layer was 1820, based on the presence of a small number of porcellaneous ceramic fragments. More than likely these were intrusive artifacts, and thus have skewed the TPQ date for the layer much later than actual date of the layer's 51

Using the techniques pioneered by J.C. Harrington (1954) and Lewis Binford (1962), the imported tobacco pipe stem fragments were employed as an additional method of determining the occupation dates for the layer. The bore diameter of each of the pipestems was measured using a set of graduated drill bits. Fifteen out of sixteen of the tobacco pipestems from the layer had bore diameters of 5/64ths of an inch. Using Harrington's method, based on histogram chart that assembles the different sized pipestem bore diameters into temporal groupings, the distribution of pipestems from M4 most closely matches the period of 1710 to 1750. Alternatively, the application of Lewis Binford's straight-line regression formula to determine a mean date for the layer's occupation results in a date of 1732.

Of the four suggested dates calculated above, the mean ceramic date (1766) is probably the most accurate as a mean date for the accumulation of the layer. A date of 1766 is supported by the relative abundance of popular third-quarter eighteenth-century ceramics in the layer. White salt-glazed stoneware was a popular ware type in the second and third quarters of the eighteenth century, and was overwhelmingly abundant in comparison to other ceramic ware types in the layer. Creamware was another very popular ware type, but not produced before 1762, and was also fairly abundant in the layer. The combination of white salt-glaze stoneware and creamware accounts for sixty-two percent of all the ceramic artifacts. Although M4 was building up throughout 52 the second half of the eighteenth century, the high percentage of these two ware types compared to other ware types strongly suggests that the layer was forming primarily in the 1760s and 1770s.

West Bulkhead

At the end of the eighteenth century, the bulkhead entrance into the basement of the dining room was relocated from the east side to the west side. The partial remains of the west bulkhead were examined in 2001 (Figure 31). The builder's trench for the west bulkhead was cut through the layer of sheet refuse (M4), indicating that the bulkhead was installed after the layer had formed. Excavation of the builder's trench unearthed wine bottle glass, window glass, nails, and a window lead, but failed to recover any datable artifacts to further delineate a construction date for the west bulkhead. A description of Bassett Hall indicates that the west bulkhead was still in service into the twentieth century (Campbell 1933:8), however, during Bassett Hall's restoration, the entrance into the basement was relocated back to the east side. In 2001, approximately 50 percent of the west bulkhead had survived intact, while the other 50 percent had been previously demolished during the building's 1930's restoration. The west bulkhead had also undergone significant repair in the nineteenth century, masking most of its original masonry.

Figure 31. Photograph of west basement bulkhead entrance.

Silt Loam with Brick Rubble

Sealing both the west bulkhead's builder's trench and the layer of brown silt loam (M4) was a layer of brown (10YR4/3) silt loam with brick rubble (M5). The brick rubble consisted of small to medium sized fragments of brick, but no whole bricks were recovered. A total of eight hundred fifty-six artifacts were also recovered from the layer. Architecturally-related finds accounted for sixty-six percent of the total number of artifacts, and consisted of fragments of window glass and nails. The increase of architectural debris over domestic refuse may be related to repairs or modifications to Bassett Hall (i.e., the relocation of the basement bulkhead). Domestic refuse once again included a variety of glass fragments from wine bottles, case bottles, pharmaceutical containers, and table glasses. The domestic refuse also included a number of fragments from different ceramic ware types (Table 3). Sixty-four percent of the ceramics were creamware. The remaining ware types included small numbers of Chinese porcelain, pearlware, white salt-glazed stoneware, Westerwald stoneware, German stoneware, Buckley coarse earthenware, and English delftware. A TPQ of 1775 was attributed to the layer based on the presence of the pearlware fragments, while the mean ceramic date was 1774.

Repair Trench

A discontinuous narrow trench (M6) along the building's foundation was located the north and south elevations of the building's original section (Figure 32). The trench was only 40 cm (1.31 feet) wide and 18 cm (0.59 feet) deep, and cut the building's original builder's trench. Excavation of the trench may have been in order to gain 54 Figure 32. Photograph of the late eighteenth-century repair trench along the north elevation of Bassett Hall. access to the exterior of the foundation walls for repairs to the brickwork. After the repairs were completed, the trench was backfilled. The archaeological re-excavation of the trench revealed several distinct thin lenses of different fills within the trench fill (Figure 33). Included within each of the fills of the repair trench were substantial quantities of both brick and mortar. Four hundred forty-six artifacts were recovered Figure 33. Profile of the repair trench along the south elevation of Bassett Hall. 55 from the repair trench fills, eighty-three percent of which was architecturally related material. The architectural finds consisted of a large quantity of window glass and nails. Domestic refuse comprised the remainder of the artifacts from the repair trench. The domestic refuse included ceramics, glass, animal bone, tobacco pipestems, unidentifiable iron hardware, one button, and a scrap of lead. In spite of a single sherd of whiteware (TPQ=1805), thought to be intrusive, the rest of the artifacts, and the stratigraphic position of the trench, suggests that trench was filled in the late eighteenth century.

Porch Footing

The final eighteenth-century feature is a footing (01AC-152) along the north elevation of Bassett Hall for a small porch (Figure 34). The footing was orientated perpendicular to the foundation wall of the house, and located 5.6 meters (18.37 feet) west from the northeast corner of the house. Only the bottom course of the footing remained, extending 70 cm (2.3 feet) out from the house, the remainder had been demolished during the installation of a water drain-pipe that ran parallel to the front of the house. No evidence of the companion footing for the west side of the porch was located. Most likely any remnants of the west footing were demolished beyond recognition as a result of landscaping and utility line installations. The small surviving portion of the east footing was 35 cm (1.15 feet) wide (1½ bricks) and bonded with shell mortar. A Figure 34. Photograph of the north porch footing. 56 sand and clay filled builder's trench (01AC-148-149) was located along the east side of the footing, but the only artifacts from the builder's trench fill were six non-diagnostic nail fragments. Because of the lack of datable artifacts, the age of the footing is somewhat speculative. It clearly post-dates the late eighteenth-century repair trench (M6) along the north elevation of the house, and was bonded with shell mortar, indicating that the likely construction date for the footing was ca. 1800. Early twentieth-century photographs of Bassett Hall depict a large two-story porch attached to the front of Bassett Hall (Figure 35). The dimensions of the porch in the photograph appear to match the dimensions of the footing discovered archaeologically. However, the two-story porch in the photograph is believed to be a late nineteenth-century addition, according to a description of the house made in 1933 (Campbell 1933:2). It is possible, however, that the porch shown in the photograph replaced an earlier porch, and had been built on the earlier porches footings.

Abundant archaeological evidence of substantial repair and renovation to Bassett Hall in the late eighteenth-century was recovered in all areas around the house. Among the repairs and renovations included, the repairs to the cellar foundation walls, the construction of a new basement bulkhead, and the construction of a front porch were all completed at this time. The late eighteenth-century date of these repairs coincides with the ownership of the Bassett Hall by Richard Corbin. Not surprisingly, the account books of several area merchants and craftsmen indicate that Corbin was spending a lot of money on building repairs and new furnishings between 1795 and 1796.

Figure 35. Early twentieth-century photograph of the two-story north porch.

Marl Pathway

Among the nineteenth-century features were the heavily disturbed traces of a marl pathway or apron (M8) on the west side of the dining room. Large patches of marl were located around the foot of the west bulkhead entrance, suggesting that the marl may have formed a path leading to and away from the bulkhead which is thought to be of late eighteenth-century origin. Accordingly, the marl path would have probably been laid down soon after the bulkhead was completed. The marl lining the path varied from 5 cm (0.16 feet) to 12 cm (0.39 feet) in depth, but the extent and direction of the path is impossible to determine because of the heavy amount disturbance. A TPQ of 1850 was determined on the basis of presence of cut nails imbedded into the path, suggesting that marl path was still in use at mid-century. A total of two hundred five artifacts were recovered from contexts identified as part of the path. Architectural debris and domestic refuse were recovered in equal numbers. Among the ceramic fragments, creamware was particularly abundant, accounting for thirty out of the forty-eight ceramics. In spite of a TPQ of 1850, the high percentage of creamware (TPQ=1762) from the path contexts suggests that it was probably originally put down at a much earlier date. A likely scenario is that the path and the west bulkhead are roughly contemporary with one another, most likely dating to the turn of the century. Meanwhile, the recovery of the small number of mid-nineteenth-century artifacts (i.e., whiteware, porcellaneous ceramics, and cut nail fragments) suggests that the use of the path continued to be used until around the middle of the nineteenth century. An occupation date range spanning the first half of the nineteenth century for the marl path corresponds with the ownership and occupation of the house by Burwell Bassett.

Repair Trench

Limited archaeological evidence of building maintenance and repair during the nineteenth century was also unearthed during the excavations. A repair trench (M7) was found along the cellar foundation walls of the original portion of the house (Figure 36). The trench measured 45 cm (1.48 feet) wide but was relatively shallow, extending to a depth of only 16 cm (0.52 feet). The trench was completely filled with large chunks of building plaster, suggesting that not only exterior, but interior work was being undertaken. 58 Figure 36. Photograph of the plaster-filled repair trench. The excavation of the extant portions of the trench yielded over three hundred artifacts, ninety percent of which was architectural debris consisting of nail fragments and window glass. The remaining ten percent included as small quantity of domestic refuse, including ceramics, vessel glass, animal bone, unidentifiable hardware, a straight pin, and a bone handle. The TPQ for the trench fill was 1820, based on a fragment of porcellaneous ceramic. Similar to the marl path, the repair trench appears to coincide Burwell Bassett's ownership and occupation of the house.

Sheet Refuse Layers

Although heavily dissected by utility lines, several layers of nineteenth-century sheet refuse were excavated along the rear of the original house, on either side of the dining room. On the west side, of dining room, two distinct layers of sheet refuse were evident. The lower level (M9) measured approximately 8 cm (0.26 feet) thick and appears to have been contemporary with the marl path during the first half of the nineteenth century. A TPQ of 1820 was attributed to the layer based on fragments of whiteware. The soil matrix of the layer consisted of a dark gray (2.5Y4/1) silt loam with substantial quantities of charcoal and brick inclusions, suggesting that a fire may have occurred during the layer's accumulation. A total of one hundred seventy-one artifacts were recovered from the layer. Architectural debris consisting of window glass and nail fragments composed sixty-five percent of the artifacts from the layer. The domestic artifacts accounted for thirty-five percent and included primarily ceramics and glass, as well as a small quantity of animal bone, a single tobacco pipe fragment, and some unidentifiable iron hardware fragments.

The second layer of sheet refuse (M10) accumulated directly on top of M9. The soil matrix consisted of a 10 cm (0.33 feet) thick layer of dark yellowish brown 59 (10YR4/4) silt loam, and with many fewer inclusions than the earlier layer. The percentage of architectural to domestic refuse from M10 was nearly identical to the earlier (M9) layer. Once again, sixty-eight percent of the artifacts found were window glass and nail fragments, and domestic refuse accounted for thirty-two percent. Although the percentage of architectural debris versus domestic refuse was very similar between the two layers, the upper layer of sheet refuse consisted of a substantially larger quantity of artifacts. The upper layer yielded 1799 artifacts, as compared to only 171 from the lower layer. Among the artifacts was a relatively large number of late eighteenth and nineteenth-century ceramic ware types (see Table 4). A TPQ of 1864 was designated for the upper layer, based on the recovery of Leighton's patent

More nineteenth-century sheet refuse (M11) was found along the west gable end chimney. Although only a small area of sheet refuse remained intact, a relatively substantial quantity of artifacts was recovered. Architectural debris accounted for fifty-seven percent of the artifacts, while domestic refuse accounted for the remaining forty-three percent. The soil matrix for the layer was a dark brown (10YR3/3) silt loam. The soil color and consistency closely resembled that of upper layer of sheet refuse west of the dining room, as did the artifacts. Late eighteenth- and nineteenth-century ceramics (see Table 4), window and bottle glass, nails, and animal bone were all recovered from the layer. Similar to M10, a TPQ of 1864, based on Leighton's patent glass, was established for the layer. M11 is likely the same layer as M10, only located further to the north and west.

On the east side of the dining room, additional nineteenth-century sheet refuse was unearthed. The sheet refuse consisted of several overlapping lenses of brown silt loam and brown sandy loam. In the field, each lens was excavated and recorded separately. Artifact analysis, however, revealed no temporal or spatial distinctions between the lenses. In order to facilitate comparison with the sheet refuse from the west side of the dining room wing, the individual lenses were combined into a single master context (M12). The combined depth of the lenses was 10 cm (0.33 feet), and contained artifacts very similar to the west side. The array of artifacts consisted of a large quantity of late eighteenth and nineteenth-century ceramics, window and bottle glass, nails, and animal bone (see Table 4). Architectural debris accounted for fifty-four percent of the artifacts, and domestic refuse accounted for the remaining forty-six percent. Similar to M10 and M11, a TPQ of 1864, based on Leighton's patent glass, was attributed to the sheet refuse on the east side. The artifacts and stratigraphic position of the layers suggest that the accumulation was likely forming throughout the first three quarters of the nineteenth century, and possibly even the fourth quarter.

All three nineteenth-century west side sheet refuse layers were grouped together, and considered together as a single accumulation spanning for most, if not all, of the nineteenth century. The artifacts from the east side sheet refuse layers were then compared to those from the west side. The comparison of the artifacts revealed some interesting differences in the nature of the artifacts from the east and west sides of the building. Although sheet refuse on either side of the house date to approximately the same time period (ca. 1800-1875), distinct differences in the distribution of ceramic ware types were evident. In particular, among the sixteen different ware types present within the east side sheet refuse, creamware accounted for twenty-seven percent, followed by English delftware at fifteen percent, and pearlware at ten percent. The remaining thirteen ware types composed the rest of the ceramic assemblage, but none accounted for more than eight percent of the total. The relatively high concentration of creamware, English delftware, and pearlware suggests that the majority of the layer's ceramics probably accumulated closer to the turn of the nineteenth century. In contrast, the distribution of ceramic ware types on the west side of the house which were more evenly distributed between different ware types representing successive periods 61 of production in the late eighteenth and nineteenth centuries. Of the variety of different ware types, creamware and pearlware each accounted for twenty percent of the total number of ceramics from west side, and whiteware and Chinese porcelain accounted for seventeen percent and thirteen percent respectively. None of the twelve different remaining ware-types accounted for more than six percent (Figure 37). This pattern is suggestive of a gradual accumulation of ceramics in the layer over time, at a relatively even rate throughout the layer's buildup. Additional comparison of six general artifact categories indicates only slight variation in the types of artifacts deposited on either side of the building (Figure 38). In particular, only domestic glass (bottles and tableware) measured any significant variation (>5%) from one side of the dining room wing to the other.

Figure 37. Distribution of ceramic ware types from the nineteenth-century sheet refuse layers.

The analysis of the sheet refuse artifact patterning suggests that changes occurred in the land use around the rear of the house in nineteenth century. At the beginning of the nineteenth century, domestic refuse was being deposited in the rear yard, on either side of the dining room. However, as the years passed, less domestic refuse was thrown out on the east side, although it continued to be regularly deposited in the rear yard to the west of the dining room. The reason for this change may be linked to the changing architecture of Bassett Hall during this period. In the nineteenth century, a porch was added to the rear of the original house on the east side of the dining room (Figure 39), although no archaeological features relating to the original porch were 62 Figure 38. Distribution of artifacts categories from the nineteenth-century sheet refuse layers. Figure 39. Photograph of the nineteenth-century porch addition along the east rear. 63 found in 2001. The construction of the porch may indicate a more formal use and conception of the east rear yard. Accordingly, trash disposal would likely have decreased in this new formal space. The west rear yard did not share the same degree of formality, as indicated by the continued disposal of domestic refuse in the yard for much of the nineteenth century. The architecture of the building also suggests that the west side was becoming more of a utilitarian space. In particular, the rear west yard contained two service entrances into the house in the nineteenth-century.

Based on the preceding analysis, it is possible to link the refuse disposal practices to specific owners and occupants of Bassett Hall. During Burwell Bassett's ownership, household refuse disposal does not seem to have been limited to either side of the rear yard. Significant quantities of household refuse, dating to Bassett's occupation, were recovered in both portions of the rear yard. However, during the Durfey and Smith families' ownership periods later in the century household refuse disposal was definitely concentrated in the west rear yard.

Etched Window Glass

Finally, among the most unique artifacts recovered from the entire excavation was a single fragment of etched window pane glass that was found in a shallow charcoal deposit adjacent to the north porch. Inscribed into the glass was, "2 Friday." Most likely, the etching was made to commemorate the engagement of Confederate Army Captain John W. Lea and Margaret Durfey, the daughter of Bassett Hall's owner Goodrich Durfey. Captain Lea was among the Confederate wounded captured by the advancing Union Army after the Battle of Williamsburg in May 1862. Captain Lea was taken to Bassett Hall after the battle to recuperate from his injuries. While recuperating at Bassett Hall, Captain Lea fell in love with Margaret Durfey, and shortly thereafter, the two were engaged to be married on Friday, May 23, 1862. To commemorate the event, the date was etched into several window panes of Bassett Hall. Several of the etched panes have remained intact in the ground floor windows of Bassett Hall to the present day (Williamsburg Garden Club 1950:47-52; Blackford et al. 1984:19-20; Rouse 1989:97-99; Taylor 1981).

The wedding took place at Bassett Hall later that summer. Attending the wedding, and serving best man, was Lea's former West Point classmate, and now a Captain in McClellan's Union Army, George A. Custer. The wedding reportedly turned out to be gala affair, and the contrast in Lea's and Custer's uniforms standing beside one another was undoubtedly an extraordinary sight to behold (Bullock 1950:47-52; Blackford et al. 1984:19-20; Rouse 1989:97-99). Custer later described the wedding in a letter to sister:

I was at the residence of the bride long before the appointed hour. Both (the bride and her cousin, Maggie, the bridesmaid) were dressed in white with a simple wreath of flowers upon their heads. I never saw two prettier girls. (John) Lea was dressed in a bright new rebel uniform trimmed with gold lace; I wore my full uniform of blue.(cited in Williamsburg Garden Club 1950:48)

After the wedding, Custer remained as a guest at Bassett Hall for two weeks, while visiting with is former classmate and his new bride. In describing his visit, Custer once again wrote to his sister: 64

I remained with Lea, or rather, at his father-in-law's house, for two weeks, and never had so pleasant a visit among strangers. Cousin Maggie would regale me by singing and playing on the piano, "My Maryland," "Dixie," "For Southern Rights, Hurrah,", etc. We were all fond of cards and would play for the Southern Confederacy. When doing so Lea and I were the only players, while the ladies were spectators. He won, every time, when playing for the Confederacy, he represented the South, I the North. Lea has since been exchanged and is now fighting for what he supposes are his rights.(cited in Williamsburg Garden Club 1950:51)

As Custer described, Captain Lea returned to fight with the Confederate Army after the wedding, and served until the end of the war. After the war, Lea became an Episcopal clergyman in West Virginia. Custer, on the other hand, was quickly promoted through the ranks, becoming a brigadier general with the Union Army at age twenty-three (Williamsburg Garden Club 1950:47-52; Blackford et al. 1984:19-20; Rouse 1989:97-99).