Elizabeth Hemings was the head of a prominent family of enslaved house servants and artisans at the Monticello home farm. The Elizabeth Hemings Site (44AB438) represents Elizabeth Hemings's final residence, where she spent the last decade of her life from about 1795 until her death in 1807. William Boyer of James Madison University initiated archaeological work at the site with a survey in 1981 to locate the structure identified as Elizabeth Hemings's residence on a Jefferson-period plat. Two archaeological fields school groups, one directed by Susan Kern in 1995 and the other by Fraser Neiman in 1996, returned to the area to conduct systematic testing of the domestic site and to explore other landscape features in the vicinity of the architectural remains.

The Elizabeth Hemings site is located on the southern slopes of Monticello Mountain about 350-feet south of Mulberry Row and 30-feet south of the Third Roundabout, one of the roads created by Jefferson to circle the mountain at a constant elevation. Two springs feed the area closest to the site. The South Spring to the east and Bailey’s Spring to the west both are within walking distance. The domestic structure, likely built in 1795 at about the same time as the Third Roundabout, was primarily of log construction (Neiman, McFaden, and Wheeler 2000:8). The architectural remains consist of scattered brick and stone from a chimney and hearth and a single postmold and posthole, perhaps representing the location of a chimney prop. A cobble concentration, located some 65-feet southeast of the site and likely contemporary with it, also marks the landscape.

Elizabeth Hemings, known as Betty to the Jefferson family, was 38 years old and had ten children when she arrived at Monticello. For fifty-one years, from their arrival in 1775 until Jefferson's death in 1826, Elizabeth Hemings's children and her children’s children filled the roles of artisans and household servants in the Jefferson family. Their proximity to the Jefferson household likely afforded them goods and privileges not shared by other enslaved families. Jefferson provided house servants, many of whom were Hemingses, with Irish linen, flannel, and knitted wool or cotton stockings, attire noticeably different from the average field workers’s coarse linens and baggy stockings (Betts 1987:41; Stanton 2000:106). Some male members of the Hemings family, such as Robert, James, and Martin, were allowed to hire themselves out during Jefferson's absences from Monticello and to keep the wages earned during these ventures (Stanton 2000:104). Indeed, the only slaves Jefferson manumitted were members of the Hemings family, another historical fact that betrays their favored status (Stanton 1996:22, 21). Recent multi-disciplinary research confirms the long-held, but often controversial, notion that all of Sally Hemings’s children were fathered by Thomas Jefferson (Foster et al. 1998; Gordon-Reed 1997; Neiman 2000; Neiman, McFaden, and Wheeler 2000:2).

Documentary sources indicate that Elizabeth Hemings was a less active member of the work force by 1790 (Neiman, McFaden, and Wheeler 2000:7). Elizabeth Hemings would have been about sixty years old when she moved into the small log house just south of the Third Roundabout sometime after 1795, a full twenty years after her arrival to Monticello. Here, she likely would have continued to raise poultry and grow vegetables as well as care for young children (Betts and Bear 1986:231). As in life, Betty likely was surrounded by her family in death. Elizabeth Hemings died in 1807 at the age of seventy-two (Betts 1987:60).

At least one reference to Elizabeth Hemings and her family appears on plans for a series of houses along Mulberry Row made by Jefferson in the mid-1770s (Jefferson: N38). “B. Hem’s” also appears on a plat of Tufton, an adjoining quarter farm, and, on the corresponding survey note dated November 10, 1794, Jefferson wrote “Betty H’s house +1” (Neiman, McFaden, and Wheeler 2000:8). The reason for Elizabeth Hemings’s move from Monticello, and likely Mulberry Row, to Tufton remains unclear, though it is clear from records that she was back at Monticello home farm by 1795 (Neiman, McFaden, and Wheeler 2000:8). The Elizabeth Hemings Site appears on two Jefferson plats. The label “B. Hem” appears on a plat dated 1808-1809 and drawn by Jefferson only a few years after Elizabeth Hemings’s death (Jefferson: N215). Another plat, also drawn by Jefferson, identifies a structure in the same location as “Quarter” (Jefferson: N225). A date of 1809 also is attributed to the latter, though one wonders about the difference in identification if both were completed in the same year, two years after Hemings’s death.

The following discussion summarizes Neiman, McFaden, and Wheeler (2000), which is available for download at http://www.monticello.org/archaeology/publications/index.html.

William Boyer, Professor of Anthropology at James Madison University, first identified Elizabeth Hemings’s house site through limited archaeological testing (ER430) during a 1981 Monticello-based summer field school (Boyer 1981). The 1808-1809 Jefferson plat guided this initial survey and allowed Boyer to pinpoint the likely location of the Hemings house, but the short field investigation and few recovered artifacts preclude further discussion of the survey here. Over a decade passed before archaeologists revisited the area to look for substantive confirmation of the structure’s location.

Susan Kern brought an archaeological field school back to the general area of Boyer’s initial survey in 1995. Her goals were to 1) confirm the location the house identified on N215 as B. Hem.; 2) define site boundaries; and 3) locate and verify the location of the Third Roundabout. In all, Kern excavated 60 quadrats of varying sizes, mostly 2-by-2 feet, during the 1995 field season. Quadrats were excavated stratigraphically, and contexts were screened through quarter-inch mesh. The 1995 season also revealed a large sandstone and greenstone concentration, identified by DAACS staff as Feature 4, that measures 12-by-23 feet and is located about 65-feet southeast of the Hemings's site proper. Three 10-by-10 foot test units exposed the bulk of the concentration with four additional test units excavated to reveal its limits (Neiman, McFaden, and Wheeler 2000:11–12). One possible posthole/mold (Feature 5) and two postholes (Features 6 and 7) were excavated within the cobble concentration. The linearity of one edge of the concentration and a corresponding row of three postholes are indicative of a fence line, perhaps one marking the boundary of a field.

A second Monticello/University of Virginia Summer Field School, directed this time by Fraser Neiman, returned to the site in 1996 to conduct additional testing in the area of high artifact concentration around the house, to further investigate the cobble concentration, and to expand the coverage of test units in all directions. The quadrat sizes were standardized to 2.5-by-2.5 feet, and excavations were performed in arbitrary .25-foot increments within strata and quadrats to maximize vertical control (Neiman, McFaden, and Wheeler 2000:11–12). Contexts were screened through quarter-inch mesh. Neiman, McFaden, and Wheeler (2000) also carried out an extensive sampling regime for chemical soil analysis and phytolith analysis, and both provided promising and interpretable results.

Kern:
Kern demonstrated that the archaeological site first identified by Boyer in 1981 is a domestic site that dates between 1770 and 1820. Her work also documented site stratigraphy (Neiman, McFaden, and Wheeler 2000:11). She recovered a small assemblage of approximately 500 artifacts and determined that the area of artifact concentration measures about 100-by-300 feet.

Neiman, McFaden, and Wheeler:
The hearth area, which previously had been identified on the ground surface as a concentrated area of brick and stone, was exposed during the 1996 field season. Much of the hearth debris, identified by DAACS staff as Feature 2, was left in situ and, therefore, does not have excavated contexts associated with it. Neiman, McFaden, and Wheeler (2000:13, 14) determined that the hearth had been completely obliterated by a tulip poplar, which had grown up through the hearth's center, though the minimal amount of masonry remains led them to conclude that the hearth and firebox were made of stone and brick whereas the chimney likely was constructed of wood and mud.

One posthole/mold, Feature 1, located about 10-feet south of the hearth area was interpreted as the location of a chimney prop, or a wooden pole that provides support for the stack and has the advantage of quick removal in the event of a chimney fire (Neiman, McFaden, and Wheeler 2000:15). Based on the spatial relationship of the hearth and chimney prop, Neiman, McFaden, and Wheeler (2000:15) proposed that Elizabeth Hemings’s house faced the Third Roundabout. Other clues to architecture were the lack of “evidence of a wooden floor....” and recovered window glass. Excavations did not reveal any subfloor pits, though contemporary Mulberry Row structures, such as Buildings s and t, were found to have such pits, leading Neiman, McFaden, and Wheeler (2000:17) to infer “that Hemings [likely] had greater control over who had access to the interior of her house than did the residents” of other contemporary quarters. Feature 3 is a possible posthole/mold located just a few feet west of the hearth area. If this feature is a posthole/mold, however, its relationship to the structure remains unclear.

Neiman, McFaden, and Wheeler (2000:18-22) generated a digital elevation model (DEM) using 1-foot interval interpolated elevations based on a 20-foot interval total station survey, hoping to pinpoint the Third Roundabout. A percent slope map, made from the DEM, showed clearly an area of more modest slope running east-west, which, based on the spatial relationship to Hemings’s house, supports its interpretation as the Third Roundabout (Neiman, McFaden, and Wheeler 2000:19). Further testing, with a 12.5-foot trench, across the linear patch identified on the slope map revealed a layer of compact clay between the B horizon and topsoil which had a level surface and a denser matrix of greenstone and sandstone, both indicative of a level and maintained roadbed.

An examination of artifact density distributions across the site indicated that many of the artifacts were deposited behind Hemings's house but with a notable difference in the distribution of creamware compared to pearlware and porcelain (Neiman, McFaden, and Wheeler 2000:33). The area of greatest creamware density was some 40 feet to the southwest of the hearth debris. Neiman, McFaden, and Wheeler (2000:37) suggest that this contrasting depositional pattern results from the fact that creamware vessels at the site are mostly curved, hollow forms, meaning that discard within the core area of outdoor activity, i.e., the yard adjacent to the house, would create a potential hazard. Ceramic analysis revealed twenty-nine vessels that together gave a mean ceramic date (MCD)of 1790.3, which is slightly earlier than the DAACS-generated MCD of 1797 (Neiman, McFaden, and Wheeler 2000:46).

Neiman, McFaden, and Wheeler (2000:47-52) explored the extent to which Monticello residents participated in the 18th-century consumer revolution by comparing discard rates of teaware and plates from the Elizabeth Hemings Site with other Monticello sites, such as the Dry Well, the Mulberry Row quarters, and the Stewart-Watkins Site, occupied by a free blacksmith and his family. They used two abundance indices, one for teaware and one for plates, to measure change through time in the discard rates of refined plates and tea forms. They concluded that Elizabeth Hemings likely participated in the trend toward greater conspicuous consumption seen at Monticello and elsewhere: the “levels of resource access and/or the advertising payoff increased at Monticello over time and...the Hemings assemblage falls at the [later] end of the continuum” (Neiman, McFaden, and Wheeler 2000:49).

DAACS:
DAACS staff cataloged artifacts and digitized field records from the Elizabeth Hemings Site excavations during the summer of 2005. Only stones from the hearth area and cobble concentration were included in the original digital site map. For DAACS, an attempt was made to include all stones drawn on quadrat plans regardless of whether or not they were collected in the field and regardless of the quadrat's proximity to the hearth area or cobble concentration. The goal was to demonstrate stone ubiquity or dearth rather than to assume either is the case, particularly for quadrats further removed from the areas of primary interest. It was clear from the context records, however, that differential mapping of stones by quadrat and field season occurred. In other words, the absence of stones on the digital site map is not evidence for absence of stones in the quadrats.