Phase II and Phase III Archaeological Database and Inventory
18AN1534 Belvoir
Diagnostic Artifact Data
Features
Summary
Site 18AN1534, also known as the Belvoir site or Scott’s Plantation, is an eighteenth to twentieth century plantation site near Crownsville, Anne Arundel County. Belvoir Manor is situated on a point of high ground overlooking several cleared fields that now function as places for athletic activities for Rockbridge Academy. The site consists of the house, the lawns and fields surrounding the house, the terraced gardens to the southeast of the house, the ruins of a bank barn, three tenant houses, a possible smokehouse, and the ruins of a possible slave quarters. Soils on site are mostly Collington-Wist complex and Annapolis series, 2 to 15 percent slope.
The land that would become the Belvoir Plantation was originally patented for Nicholas Wyatt (ca. 1620–January 1673/1674) who emigrated from Virginia to Maryland with his wife, Damaris, and a group of nonconformist Quakers in 1649. Among his subsequent Maryland land patents were 450 ac at Wyatt’s Ridge (December 16, 1662); 60 ac at Wyatt's Hills (October 17, 1664); 100 ac at Wayfield (October 21, 1664); and 175 ac at Beare Ridge (August 11, 1664), the future site of Belvoir. Over the next decade Wyatt became one of the wealthiest landowners in Anne Arundel County. At his death in January 1673 his “personal estate alone was appraised at ₤65,788…Capt. Cornelius Howard wrote the will and testified that the testator did not appear to be in condition at that time to remember what he owned”. In January 1664, Wyatt sold half of Wyatt’s Ridge to John Annis and Francis Smith. When Wyatt died, his daughter Sarah (1657–1690), who had married Colonel Edward Dorsey (1645–1705) in October 1671, received the other half of Wyatt’s Ridge as well as Beare Ridge. When Sarah Wyatt died in 1690, her half of Wyatt’s Ridge and Beare Ridge became her husband’s property. The property then descended to Colonel Edward Dorsey’s son John Dorsey (1688–1764 or 1767) before passing to his sons Samuel and Joshua Dorsey. On January 25, 1710, they sold their half of “Bear Ridge” to Amos Garrett.
In the first decades of the eighteenth century, Amos Garrett (1671–1727), a successful merchant and the first mayor of Annapolis, acquired 37 parcels of land along the Severn. Among the properties Garrett acquired, including Bear Ridge, was a 200-ac plantation called Providence. Garrett resurveyed his properties in March 1721, by which time Beare Ridge had been reduced to 100 ac, while Wyatt’s Ridge retained its original 450 ac. Through various land acquisitions, the Providence plantation grew to approximately 6,000 ac by 1725. This landholding included Bear Ridge. Garrett died in 1727, after which his administrators subdivided the estate in 1736/1737 and sold it in blocks of five or six tracts each. John Ross came to purchase a portion of the property in 1736.
John Ross (ca. 1695/96–September 18, 1766) came to Maryland around 1720 as a colonial official for Lord Baltimore. In 1735/36 he began to purchase land along the Severn River, beginning with 62 ac called “Child’s Reserve” acquired from Abraham Childs on November 29, 1735. On September 30, 1736, Ross purchased five tracts totaling 665 ac from Mary Woodward and Elizabeth Ginn, Garrett’s sisters and heirs. The tracts included: the 100-ac Beare Ridge.
When Ross purchased the foregoing tracts, improvements to the built environment were minimal. Other than small quarters for enslaved people and indentured servants, the inventories of 1674, 1704 and 1729 do not report any other buildings. This changed when Ross relocated his family to Beare Ridge where the home later known as Belvoir was constructed. John Ross and Alicia Arnold had two daughters, Anne Arnold Ross (October 9, 1727–January 5, 1811) and Elizabeth Ross (October 24, 1730–September 7, 1819). Anne married Francis Key (1731/2–November 1770) on December 12, 1752, and Elizabeth married Dr. Upton Scott (January 2, 1724–February 23, 1814) on September 5, 1756. While Elizabeth and Upton Scott had no surviving children, Anne and Francis Key had two sons and a daughter: John Ross Key (September 19, 1754–October 13, 1821), Philip Barton Key (April 12, 1757–July 28, 1815), and Elizabeth Key (August 10, 1759–1832) who married Henry Maynadier (March 31, 1759–November 11, 1849) (Ridgely 1908).
Belvoir was transferred to Upton and Elizabeth Scott through an indenture around 1758. On July 23, 1755, Ross transferred “three tracts of land in Anne Arundel Co. totaling 734 ac, including a 626-ac piece called part of Providence” to his daughter Anne. An indenture between Francis and Anne Key and Upton and Elizabeth Scott dated October 27, 1758 shows that the Scotts purchased “Three Tracts or Parcells of Land adjoining together called Childs reserve Containing Sixty Two Acres, Part of Providence Containing Six hundred & Twenty Six Acres Hills & Dales Containing forty six Acres & one half Acre lying & being in Ann Arundel County near Severn River and Containing Seven hundred & thirty four Acres more or less” from the Keys for £500. This is the same land their heirs eventually sold to Brice John Worthington in 1816.
Ross did not surrender his title to the foregoing property in the 1755 or 1758 indentures, however. Ross only promised to “at any Time hereafter at the Reasonable request of the said Upton Scott & Elizabeth or the Right Heirs of the said Elizabeth Execute & perform any assurance or act for the Purpose remising & Releasing all his Right Title Interest”. The indentures thus only granted usage rights “during the Term of their Natural Lives and the life of the surviver & from & after their disease then to the Right Heirs of the said Elizabeth to the use of the said Upton Scott and Elizabeth his wife during the Term of their Natural lives & the life of the Surviver and from & after their Decease to the use of the right Heirs of the said Elizabeth and to no other use intent or purpose whatsoever.” Though Ross identified Belvoir as belonging to Upton Scott in his will of September 1766, no deed of sale executed between October 1758 and September 1766 has been found in the land records.
An indenture made in 1771 after the death of Francis Key named his daughter Elizabeth as beneficiary and transferred Scott’s interest in the 734 acres to Trustees Js. Murray and Js. Brooks”. The transfer of usage rights protected the estate from confiscation once the American Revolutionary War began and Upton went to England. In 1800, Elizabeth Key Maynadier added her husband Henry Maynadier to the beneficiaries of the property usage rights.
While Upton Scott was in England and Ireland, Elizabeth Scott lived at Belvoir. It is possible that her sister, the widowed Anne Arnold Ross Key, also lived at Belvoir (though she may have lived at her son John Ross Key’s estate Terra Rubra) along with her niece Elizabeth Key.
Thousands of French forces commanded by the Comte de Rochambeau camped at Belvoir on the 17th and 18th of September 1781. While some of Scott’s friends, such as Daniel Dulaney, lost everything during the war, Scott emerged with his property intact. His 1783 Tax Assessment shows he retained his 734 ac plantation, two 1-ac lots in Annapolis (one at 4 Shipwright Street and one at the intersection of Corn Hill Street and Market Space), and still held 32 slaves. After the war, it appears Scott remained at Belvoir for a time, while his home at 4 Shipwright Street in Annapolis was used to temporarily house former governor Sir Robert Eden and Henry Harford. The two men returned to Maryland in 1783 in a vain attempt to regain property confiscated during the war and Scott, who had been a beneficiary of Eden’s protection and friendship, agreed to lodge the men in his urban home.
By the mid-1780s, age and health concerns made estate management burdensome for Scott. He was in need of an administrator for Belvoir as well as an heir for his larger real estate. For the daily management of Belvoir, Scott approached Henry Maynadier (1759–1849) during 1787. Maynadier was the husband of Elizabeth Key (1759–18??), the sister of John Ross Key (1754–1821) and aunt of Francis Scott Key (1779–1843). Maynadier took over the daily affairs of Belvoir on January 1, 1788.
Part of the lore surrounding Belvoir concerns the visits of John Francis Key with his grandmother as well as with the Scotts. A closer look shows that while Key probably visited Belvoir as a child, neither his grandmother nor the Scotts lived at Belvoir during the 1790s. On January 1, 1788, Henry Maynadier took over Belvoir, and Ann Arnold and the Scotts moved to Annapolis. In 1789, 10-year-old Francis Scott Key was sent to Annapolis to enter the newly opened St. John's College. Due to the lack of dormitory space for the roughly 80 students, he lived with his grandmother and his great aunt and uncle on Shipwright Street while the school was in session in 1791. He may have accompanied Scott or his grandmother on visits to Belvoir during the 1790s, but no proof of such visits has been found.
Leasing Belvoir to Maynadier only solved part of Scott’s problem of what to do with his estate. Since the Maynadiers had no children, Scott began to look across the Atlantic for an heir. In August 1784, he initiated correspondence with Clotworthy Birnie (July 1, 1765–June 2, 1845), the second son of his sister Frances (who died of breast cancer in 1790) and younger brother of Hugh Birnie. On May 6, 1810, Birnie set forth from Belfast for Maryland with his wife Hester (née Mcnaghten, June 3, 1772 –April 10, 1841) and their eight children. Scott settled Birnie and his family on his property at Runnymede near Taneytown in Frederick County.
Upton Scott died on 23 February 1814. In his will, Scott distributed 3,974 acres and 17 slaves to various family members, bequeathing “all the rest and residue of my estate, both real and personal”, to Clotworthy Birnie, who was also named executor of Belvoir. Clotworthy Birnie lived in Frederick County and could not oversee the daily affairs of the Belvoir estate. Beginning in September 1815, the 55-year-old Maynadier ever more frequently wrote about the difficulty of generating a profit from raising sheep and his desire to move to Annapolis. Birnie decided to sell Belvoir Plantation through public auction on 3 May 1816. Brice John Worthington (1764–1837) purchased four tracts of land totaling 900 ac from Elizabeth Scott, Henry and Elizabeth Maynadier, James Murray, and Samuel Ridout. This property contained numerous flocks of sheep and a variety of buildings, including a large two-story brick dwelling house; a kitchen, dairy and icehouse; a large stone building for servants; a brick barn and stables; a cow house; a sheep house; a corn house; and several other undescribed buildings.
After purchasing 900 ac., including Belvoir, in 1816, Brice John Worthington continued to add to his holdings throughout his life. A resurvey of his lands on April 15, 1828 showed his properties amounted to 2,190 ac. Worthington married Anne Lee Fitzhugh in 1809, and they had seven children.
At his death in 1837, Worthington divided his holdings into seven lots, one for each of his children. His son George Fitzhugh received Lot 1 with 514 ac. in Anne Arundel County, which included the buildings at Belvoir, as well as Lot 4 with 240 ac., 20 shares of the Farmer’s Bank of Maryland, and land previously received. George Fitzhugh also received “the following Negroes now on said farm, and in his possession, that is to say, Harry, James, Alley with her children, Thomas, Moses, Sarah Ann and Harriet, and Lydia with her children, Basil, Cinderilla, John Henry, Lucinda, and Eliza”.
Slavery was part of life in Maryland since the earliest days in 1642. Dr. Upton Scott, the Rosses, Keys, and Maynadiers were all slave owners. In his will dated September 13, 1766, John Ross listed 32 black slaves and two mulattoes, Francis Phillip Key owned about 50 slaves when he died in 1770. At the end of the War of Independence in 1783, Scott owned 32 slaves, which placed him among the wealthier members of the Maryland colonial aristocracy. Upon Scott’s death in 1814, he had 21 slaves at Belvoir and 10 at his Annapolis properties. Henry Maynadier assumed the responsibilities of Belvoir on 1 January 1788 and was listed as the owner in 1798 having a total of 27 slaves, 17 of them between the ages of 12 and 50, who lived in the two slave quarters on the property at Belvoir. There is only one account of voluntary manumission on behalf of Scott in August 1803 to one Jeremiah Tanner, described as a “Mulatto Man of the age of twenty-eight years and son to Negro Peggy a slave belonging to me.” Not all manumission was voluntary, some were actively sought in the courts. Scott was involved in a manumission lawsuit, Scott v Toogood, whereby Eleanor Toogood filed petition on October 8, 1782 petitioning for her freedom from Scott on the grounds that she "is unjustly and illegally deprived of her Liberty and detained in Slavery by Doctor Upton Scott" as a descendent of "a Free white woman and well entitled to her freedom”; historic record does not specify if Toogood was enslaved at Belvoir or one of Scott’s Annapolis properties. Judges sided with Toogood and though Scott appealed the judgement, Toogood received her court-appointed freedom on May 6, 1783. Maynadier manumitted four slaves in 1826, 1829, 1832, and 1833, though this occurred after his residency at Belvoir. None of the future owners of Belvoir is known to have freed a slave; there are no manumissions recorded for John Brice Worthington or any other Worthington living or owning Belvoir in the 1816 to 1844 timespan nor does the name of Thomas Welsh (or Welsch) appear among those who freed slaves before the end of slavery in Maryland in November 1864.
Upon his death on November 14, 1837, John Brice Worthington left a minimum of thirteen slaves to his son George Fitzhugh Worthington. When Thomas Welsh purchased Belvoir from Nicholas B. Worthington, George’s younger brother, and his wife Sophia on 24 January 1850 there were 23 slaves living at Belvoir; in 1860 at the eve of the Civil War Welsh owned 33 slaves living in four quarters. Throughout its existence until the abolition of slavery in Maryland in 1864, Belvoir’s enslaved numbers remained between 30 and 40 men, women, and children.
Thomas Welsh owned Belvoir during the Civil War. In 1861, Maryland remained, or was forced to remain, in the Union, though many of her citizens and virtually all slaveholders strongly sympathized with the Confederate States. Belvoir’s proximity to Washington, DC and location on a strategically important railroad connection to Annapolis quickly pulled the estate into the war. Belvoir/Belvoir Manor was one of 28 stations south of the Annapolis spur with the Baltimore and Ohio Railroad that connected to the Naval Academy. Belvoir’s location on the Annapolis spur brought Union troops to Belvoir Plantation who, like previous occupants of the site, left their archaeological footprint on the property as well. Among the artifacts recovered from Belvoir were Union buttons and a minie ball.
George Fitzhugh Worthington showed little interest in the property at Belvoir and quickly sold Lot 1 of 514 ac including Belvoir to John T. Hodges on February 17, 1838. Hodges too held the property only briefly before selling it to George Fitzhugh’s younger brother Nicholas Brice Worthington on November 9, 1840. On January 24, 1850, Nicholas Brice Worthington and his wife Sophia sold 362 ac, which included Belvoir, to Thomas Welsh. The sale of Belvoir to Welsh did not include sale of the slaves on the property. Rather, it appears that Thomas Welsh moved his slaves to the property. According to the October 1850 census data, 24 slaves were present at Belvoir. Eleven of these slaves were 8 years old or younger and three 55 years old or older. That left 10 slaves of working age, but only three of them, 16, 17 and 32 years old were male; the other seven were female between 28 and 50 years old. In 1860, Richard Welsh took over management of Belvoir from his 70-year-old father Thomas. Upon Thomas Welsh Sr.’s death, his son ThomasWelsh Jr. inherited Belvoir and held it through February 16, 1887 when the property and its 374 ac were sold to Anne C. Hinkley. In 1891, Hinckley sold the Belvoir plantation to James C. McCrum who owned it until his death in 1898 when it was sold to Rev. Casimir Polyanksi.
Rev. Polyanski’s brother and sister-in-law, Michael and Josephine Polyanski, lived at Belvoir; there is no evidence that Rev. Polyanski ever resided at Belvoir. In 1911, Rev. Polyanski sold the 50 ac nearest the railroad tracks to his nephew, Reuben Polyanski. In 1919, William Coleman Rogers acquired the Belvoir house and 307.5 ac from Rev. Polyanski. Michael Polyanski died July 9, 1920, and Josephine went to live with her daughter’s family in Annapolis. Though the nineteenth century saw a long list of rapidly changing ownership, the Belvoir plantation spent most of the twentieth century in the Coleman Rogers family. The Coleman Rogers family retained the Belvoir estate in full until 1985. Upon the passing of William Coleman Rogers in 1965, his son Archibald inherited the estate. Twenty years later in 1985, Archibald sold 228 ac of the planation to the Johannes-Greco Inc. development company but retained the 130 ac that included the house and its associated buildings. Archibald Coleman Rogers died in 2001, and Belvoir entered into an unknown receivership for six years. In 2007, Rockbridge Academy, a classical Christian school, began using a portion of the 144-ac property. Anne Arundel Real Estate Holding Co., Inc. purchased the property in 2019.
In 2012 and 2016, Phase I/II archaeological investigations were undertaken throughout many of the existing property’s field and woodlands which encompasses approximately 30 ac. Shovel testing, surface inspection, and remote sensing resulted in identification of five archaeology loci within the site: Locus 1 is the Button Scatter in Area A; Locus 2 is the Carriage House Ruins southwest of the house; Locus 3 is the Manor House yards; Locus 4 is the Gully remnants in Area B; and Locus 5 is the Quarter in Area B. Additional excavation took place within each loci, including gradiometer survey and the excavation of TUs.
Locus 1 is the Button Scatter in Area A; survey here resulted in the recovery of 319 historic artifacts and faunal remains and the identification of a possible historic buried A horizon in the northwest corner of the locus. Of the datable artifacts, 86 percent dated to the eighteenth to early nineteenth century. A slight concentration was noted in the northern portion of the site, but in general identification survey suggested artifacts were widely dispersed and low density.
Supplemental testing within the Button Scatter locus took place through several field efforts. In 2014, three TUs were placed in the area where a possible historic buried A horizon had been identified in STPs (TUs 17-19). The units were placed approximately 150–200 ft apart as part of a broad spatial sampling procedure in an effort to capture any subplowzone remains of the potential building with which the button scatter could be associated. A geophysical survey including gradiometer survey and GPR took place in 2015, followed by the excavation of an additional six TUs in 2016 (TUs A-F). The remote sensing and 2016 testing were conducted on the ridge top along the Belvoir driveway with the assumption that if ephemeral houses had been present, they would have been located on the ridge top, with artifacts then dispersing across the lower field.
Gradiometric readings in Area A showed several prominent dipole anomalies and weak magnetic highs associated with at least one faint road trace. The dipole anomalies were arranged in a linear or geometric orientation and may be associated with telephone pole relocations. Overhead electrical lines pass directly over this area, and the anomalies may simply represent the most recent infrastructural alignment. Alternatively, these anomalies may represent part of a fence line that existed on the north side of the main drive. GPR survey of the area resulted in identification of a natural features, such as tree roots and stumps, and potential cultural features. Cultural anomalies included potential ditches, roads, utility lines, and surfaces.
Metal detection, shovel testing, and TU excavation within the Button Scatter, Locus 1 of the Belvoir Site 18AN1534, resulted in recovery of a total of 413 historic artifacts ranging in date from the eighteenth through twentieth centuries. Temporally diagnostic artifacts predominately date to the eighteenth to early nineteenth century. Based on the variety of clothing, personal, tobacco, ceramic sherds, and glass fragments, a domestic occupation is suggested. The presence of structural artifacts, such as wrought nails and window glass, suggests a building may have been present. It is possible that the 12-x-12 ft wood house or 30-x-20 ft log quarter noted in the 1798 Federal Direct Tax assessment had been present in the vicinity.
Artifacts were widely dispersed across the landform, resulting in low density with minimal to no horizontal distribution patterning. Identification of multiple plowzones containing a mix of artifacts, with older plowzones buried under likely slope wash, suggests the sandy soils in this area have been modified significantly. This is supported by the remote sensing survey results, which found possible evidence of past modifications. If a quarter was present, it would have had a foundation and hearth as well as associated shaft features, such as privies or wells. However, no features were found, and it is unknown if features remain on the site given past modifications.
Locus 2 is the carriage house ruins; this site was identified on the basis of several above-ground architectural ruins composed of local ironstone and fossiliferous sandstone blocks. The ruins, located approximately 200 ft south-southwest of the main house, consist of two deteriorating uncoursed rubble walls and two partial rubble pillars. Scattered building stone and mortar fragments are visible on the surface within the building’s footprint. The building was partially built into a slope or garden terrace, with its north elevation flush with higher ground facing the house and its southern elevation on grade with the lower ground below the terraced gardens. Investigations of the carriage house ruins included the excavation of six 2.5-x-5-ft TUs oriented in a staggered north-south line to test the ruins’ interior as well as the areas immediately to the north and south. TUs 5, 8, and 9 were placed within the ruins’ visible footprint to characterize interior deposits. TUs 6 and 7 were placed to the south in an effort to determine if remnants of a south foundation wall were present. Lastly, TU 14 was placed to the north of the ruins to test deposits near to where an entrance is believed to have been.
In total, 583 artifacts were recovered from the three TUs within the ruins. Most of the artifacts (n=360) were recovered from Stratum I, the layer attributed to the building demolition and post-demolition soil development. Artifacts were recovered from the carriage house’s potential earthen floor, represented by Stratum II, in TUs 8 and 9 only (n=213). While most of these artifacts represent structural debris (n=182), domestic (n=28), undefined use (n=3), and personal (n=1) artifacts are represented as well. Only seven artifacts were recovered from below the potential earthen floor (Stratum III in TUs 5 and 8), but it is likely that these were pressed into subsoil and are therefore attributable to the overlying deposit. In total, 258 diagnostic artifacts were recovered within the ruins. Cut nails are most common (n=209), followed by wire nails (n=32). Most diagnostics were recovered from Stratum I demolition debris (n=194), though 54 artifacts were recovered from Stratum II in TU 9. Possibly representing the building’s earthen floor, this layer contained cut (n=45) and wire (n=8) nails in addition to a ca. 1870-1910 Blue Bennie marble. Cut (n=2) and wire (n=5) nails were also found pressed into Stratum III below the potential earthen floor in TUs 5 and 8. The presence of wire nails below the demolition debris indicates the carriage house likely stood into the early twentieth century. The stippled glass (n=14) in the upper level of Stratum I could be associated with the use of the building, but it is perhaps more likely that this represents later refuse discarded into the ruins.
In total, 234 artifacts were recovered from outside the ruins. Of these, 69 originated in the A horizon and 138 in the buried A horizon (Stratum II in TU 6 and Strata II and III in TU 14). The rest came from Feature 6, a possible historic drainage ditch, and the B horizon in TU 7. Sixty-one diagnostic artifacts were recovered from TUs 6, 7, and 14, collectively ranging from the mid-eighteenth century to the present. Only four artifacts were definitively produced before the mid-nineteenth century, including the creamware (n=1) and transfer printed refined earthenware (n=3). All of these were recovered from TU 14 and are likely associated with redistributed refuse originating from domestic activities in the main house rather than the carriage house. This is supported by the diminishing counts further from the house and the carriage house’s utilitarian purpose, whereas the house has been the domestic epicenter for substantially longer. Five features were identified during excavation of the carriage house ruins. Features 2, 3, and 5 represent post holes/post molds probably representing the locations of studs/posts for internal partitions. Feature 4 represents a mortar deposit, and Feature 6 may be a drainage trench.
The surviving structural elements, building orientation, artifact content, and relationship to other landscape features suggest that the ruins represent the remnants of a carriage house that may have also served as a garden maintenance building. The building likely originated in the late nineteenth or early twentieth century when Colonial Revival architectural aesthetics were gaining popularity. The Polynaski family owned the property at the end of the nineteenth century (1898-1919), and the Coleman Rogers family (1919-1985) acquired it in the early twentieth century. Based on relative wealth, the Coleman Rogers family likely made many of the modifications to the property in the early twentieth century. As older buildings on the property, such as the stone quarter, were razed, stone may have been repurposed for the construction of a formal carriage house to help impose a neat and historically idealized vision on Belvoir’s cultural landscape. As gardens also were part of the Colonial Revival aesthetic, construction of the building adjacent to the gardens may have facilitated garden maintenance. Diagnostic artifacts, dominated by cut and wire nails, are consistent with this construction period. It is possible that the building sustained some fire damage, as the charcoal deposits in TU 9 may suggest, after which it was razed. There are records that a brick barn on the property burned in the 1950s, and it is possible that the carriage house was also damaged at that time. The building is shown as ruins in a 1970s photograph.
Locus 3 is the manor house yards; excavations were undertaken within the vicinity of the house to test for intact yard deposits and features associated with the building’s construction episodes. Nine TUs were excavated to the east, west, and south of the house. In addition, a remote sensing survey was conducted to the east of the house.
Four TUs were placed in the yard west of the house. One Feature (Feature 1) was found in TUs 2 and 4 to the west of the house. Feature 1 was identified as a possible brick walkway located in the southeast quadrant of TU 2 between 0.25 and 0.6 ft bgs and continuing into the northeast quadrant of TU 4 (Figure 48). The feature consisted of three partially exposed rows of dry-laid red bricks obliquely transect the units along a northeast-southwest orientation. The bricks were laid in what appears to be a standard running bond. Surrounding the few surviving bricks were numerous brick fragments, and heavily compacted dark yellowish brown (10YR 3/4) clay fill. Artifacts included a mix of eighteenth through twentieth century materials in all strata. In total, 777 artifacts and faunal remains were recovered from TUs 1, 2, and 4; no artifacts were recovered from TU 3. Historic artifacts were primarily from the structural group (n=583), which is consistent with testing adjacent to the large dwelling. Other functional groups represented included activities (n=1), domestic (n=11), personal (n=7), transportation (n=1), faunal (n=44), and undefined use (n=130) (Table 52). It is likely that many of the glass and ceramic fragments included in the undefined use category were also used for food and drink storage, serving, or consumption.
Three test units were excavated adjacent to the house foundation to the east and south of the house. Feature 8 was identified as an amorphous dark yellowish brown (10YR 3/4) silty loam stain with heavy concentrations of mortar, brick, and plaster located in the north half of TU 20. Measuring 3.7 ft east-west by 1.7 ft north-south, Feature 8 extended approximately 0.39 to 0.69 ft bgs. The feature may have been part of Stratum III, but it contained more architectural rubble. Feature 9 was identified as a linear dark yellowish brown (10YR 4/4) sandy loam partially overlying Feature 8 and traversing east-west across the center of TU 20. Measuring 1.1 ft wide, Feature 9 was encountered between 0.47 and 0.92 ft bgs at the Stratum III/IV interface. Its orientation and location suggest it paralleled the twentieth century porch and may represent a drip line across its south elevation. Feature 12 was identified as a rectangular post hole with a round post mold found in the southwest quadrant of TU 20. Feature 11 was identified as a brown (10YR 4/3) and yellowish brown (10YR 6/6) builder’s trench abutting the brick foundation wall identified in the southwestern quadrant of TU 24. Measuring up to 1 ft wide and 0.75 ft thick, Feature 11 was identified at the Strata I/II interface. Feature 13 was identified as a brown (10YR 4/3) trench surrounding Feature 11 in TU 24. Feature 13 may represent the original dimension of the builder’s trench before a narrower trench (Feature 11) was cut against the intended location of the brick wall. Feature 14 was identified as trench consisting of dark yellowish brown (10YR 4/4) sandy clay identified in the eastern half of TU 24. This feature may represent a distinct fill episode.
In total, 1,534 artifacts and faunal remains were recovered from the east and south side of the house. Historic artifacts were primarily from the structural group, which made up 77 percent of the assemblage (n=1,175). Other functional groups represented included activities (n=7), domestic (n=32), personal (n=13), faunal (n=35), and undefined use (n=268). It is likely that many of the glass and ceramic fragments included in the undefined use category were also used for food and drink storage, serving, or consumption. Approximate 20 percent of the retained temporally diagnostic artifacts dated to the eighteenth to early nineteenth century. Approximately 42 likely date to the mid- to late nineteenth century, and 40 percent date to the twentieth century. Twentieth century modifications immediately adjacent to the house, including infilling of the foundation, fill deposits to level the land, and construction of a frame porch addition, have affected the integrity of earlier deposits. In addition to the historic artifacts, an Early Woodland projectile point and three Woodland period ceramic sherds (two shell tempered and one sand tempered) were found in TU 21. The prehistoric artifacts were found in association with historic materials. While indicative of a Woodland period occupation, no undisturbed prehistoric component was found.
Investigations in the front (east) yard of the house included STP excavation, remote sensing, and the excavation of two TUs. Subsurface testing was limited in this area so as to not disturb the grounds. Combined results of gradiometer and GPR survey within the front yard of the house indicated the presence of a potential structure with intersecting right angles (Anomaly 1). The E-shaped anomaly appeared situated between several field drainages. Measuring approximately 20 ft north-south by 15 ft east-west, this anomaly was assumed to be representative of a small building footprint. The building appears to have had a continuous foundation on three sides and was segregated into two bays, each of which opened to the east. In addition to the foundation, potential historic surfaces were indicated in the remote sensing results (Anomalies 2, 3, and 7). These surfaces exhibited differential compaction suggesting they were heavily used areas. Two of the surfaces were covered by fill, and additional fill deposits were identified based on the GPR results; fill was likely used to create a level surface.
TU 109 was placed within GPR Anomaly 2, suspected to be a buried surface. TU 110 was placed over GPR Anomaly 1, which was found to be a brick and stone foundation (Feature 109). Feature 109 was a mortared Flemish bond brick foundation resting on a local ironstone footer. Feature 111 was a builder’s trench located on both sides of Feature 109. In total, 327 historic artifacts and faunal remains and one prehistoric flake were recovered from TU 110. Most artifacts (n=200, 61 percent) were found in Stratum I. Structural, foodways, and undefined (likely domestic) artifacts together make up 96 percent of the assemblage.
Despite the documented history of the house extending back to the early to mid-eighteenth century, few artifacts potentially dating to this period were found in the area immediately surrounding the house. The yard area was found to contain disturbed soils and fill deposits resulting from modifications to the house and yard. Features were associated with twentieth century modifications, including a landscaping wall or brick paver path, and the frame porch addition that had been present on the south side of the house. In addition to reflecting disturbance, the lack of early artifact deposits may reflect that excavations took place within the maintained presentation spaces of the house yard, while midden deposits and work spaces may have been located in areas of the property not tested.
The outbuilding in the front yard (Feature 109) appears to date to the late eighteenth to mid-nineteenth century. The building was likely added during the later period of Upton Scott’s ownership and primarily used during Worthington’s ownership. Based on the large amount of faunal remains and variety of ceramics, Feature 109 may have been a detached kitchen. Remote sensing suggests the feature is approximately 20-x-15 ft in size, and the building may be the 20-x-18 ft brick out house noted in the 1798 tax assessment. Based on typical plantation orientation, it is possible that a small outbuilding had mirrored the Feature 109 building to the south, although this may have interfered with the intention of the garden and lawn. Alternatively, the core of the buildings may have been located on the slope to the north of Feature 109.
Locus 4 is located in Area B and is the site of two segments of refuse-filled run-off gullies. One was identified west of the stone quarters, while the other was found uphill in the vicinity of the playhouse. It is unclear if the two drainages were directly connected to one another or if they were independent elements of a broader drainage system that drained the areas around the main house and quarters. Though the full extent of these two gullies was not exposed, it appears they ultimately discharged runoff into the wetlands below, northwest of the plantation’s domestic epicenter.
The quarter gully represents part of a natural drainage that carried runoff and waste from nearby residential loci toward the wetlands to the northwest. The feature may have been used for intentional refuse disposal in addition to accumulating artifacts through runoff. Based on the diagnostic artifacts, it appears that the gully appeared in the second half of the eighteenth century, corresponding with Upton Scott’s period of expanded construction and improvements at Belvoir, which appear to have negatively affected drainage patterns on the property. Lines of stone and brick rubble and pockets of depositional fill may reflect attempts to control erosion and the meander of the drainage. Over time rubble and artifacts slowed water flow, resulting in sedimentation. Two possible post holes (Features 15 and 34) may indicate that a fence was installed along the gully, presumably to prevent people and from entering the unstable and refuse-filled area and to separate waste areas from living areas. The gully was filled by the early twentieth century and no longer carried water, possibly reflecting a stabilization of the landscape and changes to drainage patterns on the property after extensive landscape and building modifications in the early twentieth century.
Artifacts from the gully likely originated from the house and quarter. Isolating socioeconomic signatures among the artifacts is difficult, particularly given that they are secondary deposits. Additionally, most of them could have been used by more than one group of people. For example, many of the imported, decorated ceramics may have originated with Belvoir’s owners, but any number of them may have been passed on for the use of enslaved people. Utilitarian vessels, clothing, and personal artifacts likewise could have belonged to any of the groups of people. The only artifact that potentially originated with Belvoir’s enslaved laborers is a single piece of what may be colonoware; however, it could not be definitively differentiated from prehistoric ceramics and thus remains unattributed to any group of people.
Feature 10 (near the playground) appears to represent a historically filled gully. The walls of the gully were fairly straight, suggesting intentional modification, although the size of the feature narrowed irregularly indicating this was likely a natural gully with minimal modification. The distribution of diagnostic artifacts among Feature 10 and the surrounding strata suggest that it was open for a brief period during the mid- to late nineteenth century. While it contains artifacts definitively produced before this time, the strata into which it incised date from the early to mid-nineteenth century. The fill exhibits no internal stratification, with artifacts from all time periods thoroughly intermixed. The artifact’s functional distribution indicates it was filled with generic domestic refuse, consistent with the kinds of waste expected for a nearby domestic occupation. Artifacts may have washed into the gully from refuse deposits behind earlier outbuildings to the south. By the end of the nineteenth century the gully had been filled, either through sedimentation or intentional filling to control erosion. This may represent one of many short-term drainage gullies that appeared during the historic occupation.
Locus 5 is a stone quarters for enslaved people within the Belvoir Site. The locus is situated on a terrace that in 2014 contained a dilapidated mid-twentieth century tenant house in testing Area B. Shovel testing and surface inspection resulted in recovery of a concentration of eighteenth-century artifacts; in total 500 artifacts were recovered from the area during the shovel testing. Sixty-four percent of the datable artifacts from the STPs dated to the eighteenth to early nineteenth century. Potentially aligned stones were visible on the surface to the west of the extant house. Initial TU excavation confirmed the presence of a brick floor and stone foundation wall. Further excavations revealed the foundations of a 34-x-35 ft masonry building. These dimensions and its location suggest the building was the stone quarter noted in the 1798 Tax Assessment.
In total, 72 contiguous TUs were placed within the quarter area, including 32 in 2014, 37 in 2015, and three in 2018. In addition, four TUs were excavated in the yard surrounding the quarter. Prior to the 2015 excavations, the twentieth century house was removed to allow exposure of the full dimensions of the quarter building. TUs varied in size and shape depending on the exact location and conditions (e.g., structural features in some cases defined unit boundaries resulting in irregular dimensions). Most TUs were 5-x-5 ft in plan. The concrete block house foundation was resting in part on an older poured concrete foundation of a pump house dating to the early twentieth century. The most extensive modern disturbance to the historic quarter resulted from installation of a cellar; other disturbances included a filled well (Feature 86) likely related to use of the building as a pump house, utility trenches, and the poured concrete and concrete block foundations; in many cases, however, portions of the historic features were found intact below the modern foundations. Testing resulted in recovery of 41,344 historic and 22 prehistoric artifacts.
Initially 83 features were recorded during the quarter investigation. Twenty-one of the features were determined to be natural (e.g., rodent burrows), modern, or duplicates. In some cases, it was not clear that different features were connected when encountered in discontiguous TUs, and multiple feature numbers were assigned to what was eventually found to be one feature after the area was exposed. In those cases, generally the first feature number was retained, and later numbers eliminated. Historic features included features associated with the building structure, storage and refuse pits, possible posts, and other features such as ditches, areas of burned soil, and stains of indeterminate function.
The artifact assemblage reflects the long occupation period and domestic nature of the quarter, as well as post-occupation modifications. In total, 22 prehistoric and 40,548 historic artifacts were recovered from the quarter. In addition, 20,295 faunal remains and 14,142 floral items were recovered from excavation and flotation. More than half of the recovered artifacts are structural materials used in the building construction.
In total, 22 prehistoric artifacts were recovered from the quarter area. All prehistoric artifacts were recovered from historic contexts and appear to represent incidental incorporation into the assemblage during historic construction and land use. Artifacts include 12 debitage, nine tools, and one gorget. Tools consist of three quartzite bifaces, two groundstone, and four projectile points. Most artifacts are made from locally available quartzite (n=11) and quartz (n=6). The projectile points include one possible Guilford, one Clagett, one Piscataway, and one unidentifiable tip. The Guilford dates to the Middle Archaic period, while the Piscataway and Clagett points date to the Late Archaic. The slate or rhyolite gorget fragment was 1.7-x-0.8-x-0.2 inches in size with a single drilled hole; the polished artifact appeared to represent the middle section of a gorget. Gorgets were used as adornment primarily during the Late Archaic through Woodland period. The two ground stone artifacts potentially represent pestles.
The activities group includes a variety of artifacts associated with activities that occurred on site. The 163 artifacts in this group comprise less than 1 percent of the historic assemblage. The items were divided into the following categories: animal husbandry (4 horseshoes), commerce (15 coins), entertainment (3 jaw harps), firearms (a bayonet, 10 gun flints, a lead flint wrap, and two flint fragments), fishing (two lead weights), plant husbandry (plant tags), tool (45 tools), toys (glass marbles, frozen Charlotte doll, and iron toys), transportation (horse and wagon related), unidentified, and writing (11 stylus and 4 writing slates).
The domestic category includes items related to activities in the home, such as clothing maintenance, and food and beverage preparation, storage, serving and consumption. Furniture and lighting elements are also included. Food remains, however, are discussed in the faunal and floral sections. During cataloging, general flatware ceramics were defined as food preparation and consumption items, while general hollowwares and small fragments of indeterminate form were cataloged as undetermined function with the assumption that food and beverage storage is more removed from other domestic functions. All ceramic fragments, however, are included in the discussion of domestic artifacts, as are most glass. In total, 16,627 artifacts are included in this discussion: clothing maintenance (n=101), food and drink related (n=16,456), furnishings (n=38), heating/lighting (n=32).
In total, 10,965 ceramic fragments were recovered from the quarter area. Ceramics were ubiquitous across the site. In contrast with the clothing maintenance artifacts, ceramics were primarily recovered from strata on the exterior of the quarter. This is likely indicative of multiple processes. The quarter was cleaned out and salvaged from at abandonment and prior to demolition, which accounts for the presence of mid-nineteenth century artifacts in most contexts. In addition, some yard spaces were likely used as middens (possibly frequently cleared) with some materials dispersed through erosion or other disturbances. Feature contexts yielded 1,180 of the ceramics, while 9,785 were found in TU strata, particularly the demolition layer. Ceramic fragments reflect the long occupation period, with types from the early eighteenth through the twentieth century represented. The mean ceramic date for the quarter site is 1832. Most of the ceramic types were made in England. Based on median manufacture date, only three percent of the ceramics date to the Ross family period of ownership. Forty-four percent of the ceramics date to the Scott family period of ownership. The ceramic evidence supports the conclusion that Upton Scott had the quarter built around the time he built the large brick manor house. The Scotts and Maynadiers (married to Scott family) owned between 21 and 32 enslaved people over their years of property ownership; records for Henry Maynadier from 1798 note that he had 27 enslaved people at Belvoir living in two quarters on the property. Most of the ceramic types with median dates attributed to the Ross ownership have manufacture date ranges that extend into the Scotts family period. The few ceramic types with manufacture periods ending before 1758, such as Astbury, likely represent curated artifacts and items handed down to or acquired by the enslaved population after the wares were no longer fashionable. Approximately 45 percent of the ceramics date to the Welsh family period of ownership. Only 9 percent appear to date to the Worthington ownership period, although it is likely that many of the sherds attributed to Scott or Welsh were in use during the Worthington ownership, but due to long manufacture periods their median date appears later.
In total, 5,299 domestic and indeterminate glass fragments were recovered from the quarter. It is likely that the majority of the indeterminate glass represents domestic vessels. Domestic glass included tableware, bottles, jars, medicine bottles or vials, and unidentifiable non-structural fragments. Most glass fragments were recovered from non-feature contexts (89 percent, n=4,482).
Floral artifacts include remains of food, architectural elements or furniture, and heating fuel. In total, 15,875 macroplant remains (weighing 718.15 g) were recovered. Food remains recovered from flotation include nuts (black walnut), wild or gathered plants (bulrush, cherry, fig, grape, violet), and crops (squash, corn, oat, and wheat).
The personal group includes 1,205 artifacts typically used or carried by one person or associated with individual care and hygiene. The group represents 3 percent of the historic artifacts and was divided into the following categories: accoutrements (17 glass beads, 3 jewelry pieces, 2 parasol parts), clothing (312 buttons, 10 buckles, 34 hoop skirt frames, and other), footwear (10 boots/shoes, 8 eyelets, 1 aglet, 1 shoe buckle), grooming/ health (32 chamber pot pieces, 5 combs, 10 medicine bottle pieces, 4 toothbrushes), and tobacco (723 pipe pieces, 1 snuff bottle).
Structural artifacts represent materials that made up the fabric of the building. Structural artifacts constituted 52 percent of the recovered artifacts from the site (n=21,008). In addition, some analyzed wood samples appear to represent walls, roof or loft structural elements, or posts, as noted in the floral material discussion (e.g., plank from TU 51; possible post from Feature 88, and a plank from TU 71). Materials such as brick and mortar were sampled to record their presence and type, but not quantified due to the volume of these materials encountered. Structural artifacts included building materials such as brick (n=23) and plaster (n=149), window glass (n=5,596), fasteners such as nails and screws (n=15,130), and architectural artifacts with specific functions like hinges (n=53), locks (n=11), and doorknobs (n=2).
The indeterminate artifacts (n=1,523) represent either unidentifiable pieces of unknown items, usually metal, or items that could have served a variety of functions and therefore could not be assigned a specific functional category.
The stone foundation represents a kitchen and quarter for housing enslaved people. The 34-x-35 ft dimensions closely match the stone quarter noted in a 1798 tax assessment (32-x-32 ft). Artifacts suggest the quarter was built soon after Upton Scott acquired the property in 1758. Few artifacts definitively date to the earlier Ross family ownership period. In addition, the quarter was larger than Ross’s stone house. In contrast, Scott added the large brick house to the property and the quarter, made from a less expensive material and downhill from the main house, would have complimented Scott’s landscape design. The quarter housed enslaved people during the Scott/Maynadier, Worthington, and Welsh periods of ownership. Between 25 and 35 enslaved individuals resided on the property prior to emancipation, living in at least two slave quarters, including the stone quarter and log building; others may have resided above the kitchen and in other outbuildings. It appears that the stone building continued to be used for a short time during and after the Civil War. In the late nineteenth century, the quarter appears to have collapsed or was pulled down and stone from the walls, possibly wooden roof framing, and hardware were salvaged for use in construction of a carriage house, garden walls and possibly other buildings and structures on the property.
Features identified within the quarter suggest the building was a 1 ½-story nearly square stone building with brick window and door surrounds. The building had a central chimney. A central entrance was present on the southern facade, and a window and door were present on the east and west sides of the building. Interior brick walls resting on stone footers divided the quarter into three rooms, each with a fireplace that fed into the central flue. The front room is a kitchen with a small cellar; it was 32-x-17 ft in plan with a large central hearth with an 8 ft opening. The back rooms were living quarters; they were each approximately 14-x-15 ft in size. The kitchen and northeast rooms had brick floors, while evidence of the northwest room flooring does not survive. Worn bricks on the east side of the kitchen suggest a doorway between the kitchen and back room. Based on a lack of slate in the assemblage, the building likely had a wood shingle roof.
The stone quarter is an unusual example of a slave quarter in the region. Most documented and surviving quarters consist of rectangular hall-and-parlor dwellings or two separate, single pens with lofts, side to side sharing a flue. Of note, Thomas Jefferson once sketched a building closely resembling the Belvoir quarter, although there is no indication Jefferson built from the sketch.
As noted, the quarter housed enslaved people during the Scott/Maynadier, Worthington, and Welsh periods of ownership from the mid-eighteenth through the 1860s and appears to have collapsed around 1870. With the exception of the enslaved listed in John Brice Worthington’s will, the names of the occupants are not known. Enslaved people noted in Worthington’s will included Harry, James, Alley with her children, Thomas, Moses, Sarah Ann and Harriet, Lydia with her children, Basil, Cinderilla, John Henry, Lucinda, and Eliza. DNA testing of four tobacco pipe fragments resulted in the association of one tobacco pipe positive for female DNA that is most similar to the Mende of Sierra Leone West Africa; a second sample indicated African or African American lineage in general.
The front room served as a kitchen and the back rooms as living areas throughout the occupation. During the Scott/Maynadier and early Worthington ownership period, the enslaved people used a cellar on the west side of the kitchen hearth. As it filled with debris, the pit was cleaned out, leaving a layer at the base dating to the Scott/Maynadier period. The cellar was filled-in completely in the early nineteenth century, possibly around when John Brice Worthington died in 1837. The presence of items such as straight pins and thimbles along with food remains and ceramics reflects the types of activities that took place within the kitchen portion of the quarter.
A cache of minimally fragmented artifacts was found adjacent to the cellar possibly indicating storage adjacent to the cellar; artifacts included a stirrup, knife, smoking pipe, bayonet, kitchen ceramics, tea wares, and a variety of other items. The kitchen appears to have been used for storage, particularly during the later period of occupation and potentially after its domestic use ceased.
Artifacts dating to the Scott, Maynadier, and Worthington periods included a wide variety of materials, including coins, tools, gun flints, locks, tobacco pipes, jewelry, food serving, preparation and storage, and tea wares, and clothing-related items including decorative buttons. The wealth of the landowners is reflected in the material culture they provided to the enslaved population. It is also likely that individuals had some access to markets and could acquire goods on their own. The presence of coins, including in the bottom of the cellar, supports the interpretation the enslaved people had some way to obtain money and purchase items.
Based on the presence of decorative buttons and beads it appears that the enslaved people were provided with or acquired refined clothing. Alternatively, the enslaved may have been tasked with maintenance and repair of the owner’s clothing. Significant quantities of straight pins, thimbles, and a pair of scissors recovered from the quarters suggest sewing took place, and likely laundry as well.
The owners provided the enslaved people with rich material culture or the opportunity to acquire elements of finer material culture, potentially to project the owners’ wealth when the enslaved were in public settings. The people who occupied the quarter may also have sought ways to project their status above other enslaved people and potentially poor white people (Forrett 2004). The food allotment to the enslaved people appears to have been inadequate leading them to supplement with wild caught and gathered food. Faunal and floral remains, particularly from the cellar and hearth area, included significant amounts of fish along with reptile, shellfish, bird, and wild fruits. The presence of bull rush suggests the occupants may have made baskets; squash or gourd may have been used as food or as a container.
By the time Welsh family acquired the property in 1853 it appears to have been in need of repairs. Evidence of repairs includes patched areas of brick flooring, leveling of the floor in the northeast room or possibly removal of an internal structure, and repairs to the hearth in the northwest room. A series of small pits in front of the hearth contained burned materials and nineteenth century artifacts. It is possible that these represent small refuse pits formed during floor or hearth repair. At least one contained artifacts dating to the late nineteenth century and may represent burning of kitchen refuse at the end of the occupation. A series of post features at the northeast corner of the quarter date to the Welsh period. A shed addition or covered work area may have been added during this time. A feature (Feature 55) interpreted as a drainage ditch may instead represent a dripline associated with this work area. A brick patio was also added to the front of the quarter. The variety of artifact types dating to the mid-nineteenth century was much reduced from the variety seen earlier. Welsh period artifacts primarily consist of domestic and architectural materials.
In 1860 at the eve of the Civil War Welsh owned 33 enslaved individuals living in four slave quarters. After the Civil War, use of the quarter dropped off considerably as few artifacts were recovered that date to the late nineteenth century. The presence of a mine ball and Civil War military buttons may indicate that the stone quarter was temporarily occupied by troops possibly working on the farm. Soldiers were stationed at guard posts at regular locations along the railroad to the west and may have taken advantage of services or opportunities available on nearby plantations. Laundry may have included military uniforms.
It appears that when the building deteriorated, the east and west walls were pulled outward to allow for salvage of the stone. The most intact stones appear to have been robbed following the collapse and potentially reused in the carriage house and garden wall constructions. There is an area just north of the quarter with a dense deposit of deteriorating mortar. This may be a location where stone was stripped of old mortar. An iron chisel was found in situ in a rubble pile in this location. Additionally, the recovered bayonet was observed to have mortar adhering to it suggesting reuse as a construction tool.
The quarter may have been disassembled after a period of use as a storage building after occupation ended. The wood roof framing was likely salvaged, and the stones in the walls were used for other buildings and structures. Perhaps the Burned House site was one of the buildings constructed of some of these recyclable materials. It may not be coincidental that Caledonian ceramics were recovered from both sites.
Archaeological investigations of the Belvoir property yielded significant information about the people who lived on the property in the past. Identification survey and testing took place over approximately 63 acres, representing approximately 43 percent of the current 144 ac property boundaries and only a small fraction (3 percent) of the original 2,190-ac farmstead. Additional investigation within the remaining property may identify additional significant resources. In particular, the woods north and northeast of the stone quarter may hold additional resources associated with the enslaved populations on the property. Within areas tested as part of the current investigation, the front (east) yard of the house appears to include significant deposits and historic features that warrant additional investigation (e.g., Feature 109 foundation). Additional resources are likely located within the area between the house and quarter, and potentially in yard areas of the quarter. Overall, Site 18AN1534 is a multi-component site that may still yield further information on land usage and who lived on the property. While some of the site has revealed to be disturbed, there still may be undisturbed portions of the property.
|
