The most likely explanation seems to be that Structure F was the foundation of a smokehouse. A recently excavated foundation in what was known as Brunswick Town, North Carolina, is almost identical (except for the use of ballast stone in the fire chamber and the building foundation). This also is believed to be a smokehouse foundation, since similar structures are still remembered from the days of their use.[159]

XIII

Pits and Other Structures

STRUCTURE D

An exploratory trench was dug northward several yards from a point on Wall D, on axis with Structure B. An irregularly shaped remnant of unmortared-brick structure, varying between two and three bricks wide and one course high was discovered at the undisturbed level. This measured 8.5 feet by 6 feet. Adjacent to it, extending 5.8 feet and having a width varying from 6.5 to 7 feet, was a pit 2 feet 8 inches deep, dug 2 feet below the undisturbed clay level, and filled with a heavy deposit of artifacts, oystershells, and animal bones. The artifact remains were the richest in the entire site. Some of the most significant of these are the following:

59.1656 Key (fig. 88)

59.1942 Iron bolt (ill. 69)

59.1663
59.2029
59.1939

}Two-tined forks (ill. 55-57)
 

59.1664 Jeweler’s hammer (ill. 78)

59.1665 Fragments of a penknife (fig. 85c)

59.1668 Knife blade and Sheffield handle (fig. 86b)

59.1669
59.1670

}Pewter trifid-handle spoons (fig. 86f and g, ill. 58)
 

59.1672 Pewter “wavy-end” spoon (fig. 86e, ill. 59)

59.1675 Fragments of reeded-edge pewter plate (fig. 86a)

59.1676 Pewter teapot lid (fig. 86c, ill. 60)

59.1678 Brass rings (fig. 83i)

59.1680 Steel scissors (ill. 61)

59.1681 Large fishhook (ill. 88)

59.1682 Chalk bullet mold (fig. 84b, ill. 51)

59.1685 Slate pencil (fig. 85d, ill. 54)

59.1687 Octagonal spirits bottle (fig. 80)

59.1688 Wine bottle: seal “I ^C.M 1737” (fig. 78, ill. 37)

59.1679 Handle sherd of North Devon gravel-tempered earthenware (ill. 15)

59.1698 Buckley high-fired, black-glazed earthenware (fig. 65)

59.1699 Buckley high-fired, amber-glazed earthenware pan sherds (fig. 65, ills. 17 and 18)

59.1700 Brown-decorated yellowware cup or posset-pot sherds (fig. 64c, ill. 16)

59.1701 Nottingham-type brown-glazed fine stoneware sherds (fig. 67a)

59.1762 Sherd of Westerwald blue-and-gray stoneware, with part of “GR” medallion showing (fig. 66d)

59.1704 Large sherds of brown-glazed Tidewater-type earthenware pan (fig. 63a, ill. 11)

59.1706 Blue-and-white delft plate, Lambeth, ca. 1720 (fig. 69)

59.1707 Blue-and-white delft plate, [?]Bristol, ca. 1750 (fig. 70)

59.1714 Kaolin tobacco-pipe bowls, and one wholly reconstructed pipe (fig. 84f, ill. 53)

59.1715 Steel springtrap for small animals (ill. 86)

(Also numerous sherds of Staffordshire white salt-glazed ware and creamware. A single disparate sherd of pink, transfer-printed Staffordshire ware, dating from about 1835, is the only intrusive artifact in the deposit.)

The bones were virtually all pork refuse, except for a few rabbit bones. The oystershells, found in every refuse deposit, reflect the universal taste for the then-abundant oyster.

The significance of the structure is not clear. It was probably the site of a privy, the remaining bricks having been part of a brick floor in front of the pit.

STRUCTURE G

A few feet southeast of Structure D, another much smaller pit was found, surrounded on two sides by a partial-U-shaped single row and single course of bricks. This brickwork measured 5 feet in length, with a 4-foot appendage at one end and a 7-foot appendage at the other. The pit was small and shallow. Typical ceramic artifacts were found, as well as fragments of black basaltes ware (ill. 32) and some early 19th-century whiteware. The function of this pit is unknown.

PIT AT JUNCTION OF WALLS A-II AND D

Just north of the northeast corner of the wall system a small trash pit was uncovered. It contained a scattering of wine- and gin-bottle sherds, a few miscellaneous, small, ceramic-tableware fragments, and about one-third of a blue-and-white Chinese porcelain plate (figs. 55 and 77).

UNIDENTIFIED FOUNDATION NEAR POTOMAC CREEK (STRUCTURE H)

About 60 feet from the shore of Potomac Creek, at the southeast corner of the old road that runs from the highway to the creek, bordered by Wall A, were indications of a brick foundation. This structure was explored to the extent of its width (about 15 feet) for a distance northward of 17 feet, then the east wall was traced 22 feet farther north until it disappeared into the bankside and a thicket. The excavated area disclosed quantities of brickbats, a layer of soil, a number of burnt bricks, a layer of black charcoal ash, and a 6-inch deposit of clay. The brick walls were 1.5 feet thick. The structure had been built into the hillside, so that the north end was presumably a deep basement.

Artifacts were few. A complete scythe (fig. 90) was found embedded in the clay above the brickwork on the east side of the structure, and next to it a large body sherd of black-glazed Buckley ware. A few small ceramic sherds occurred—pieces of redware with trailed slip (fig. 64), and small bits of delft, salt glaze, and Chinese porcelain.

The location and implied shape of the building suggest that it had a utilitarian purpose. Near the waterfront, it would conveniently have served as a warehouse, or possibly as either the brewhouse or malthouse, each described by Mercer as having been 100 feet long, of brick and stone. Whether one was of brick and the other of stone, or both were brick and stone in combination, is not clear. There was no evidence of stonework in Structure H. On the other hand, the 100-foot-long rectangular stone enclosure, of which Wall A formed a part, shows no evidence of brickwork. The purposes of both these structures must, for now, remain unexplained, but association with the brewery seems plausible.

XIV

Stafford Courthouse South of Potomac Creek

INTRODUCTION

The chief archeological problem of Marlborough at the time of excavation was whether or not Structure B had served as the foundation for both the courthouse and for John Mercer’s mansion. Although the possibility still remains that the sites of the two buildings overlapped, preceding chapters have demonstrated that the foundation was constructed by Mercer for his house, and that it did not stand beneath the courthouse.

However, in 1957 it was thought that exploration of the late-18th-century courthouse site, located upstream on the south side of Potomac Creek, might reveal a structure of similar dimensions which would help to confirm the possibility that Structure B had originated with the Marlborough courthouse. Furthermore, the Potomac Creek site was of interest by itself and was closely related to John Mercer’s legal and judicial career.

The location of the site is depicted in surveys included with suit papers of 1743 and 1805.[160] These papers were brought to our attention by George H. S. King of Fredericksburg, and were mentioned in Happel’s carefully documented history of the Stafford and King George courthouses.[161] Previously, we had been led to the site by a former sheriff of Stafford County, who recalled listening as a boy to descriptions of the old courthouse building by an ancient whose memory went back to the early years of the 19th century. The old man’s recollections, in turn, were reinforced by similar recountings of elders in his own youth. Unscientific though the value of such information may be, it emerges from folk memories that often remain sharp and clear in rural areas, spanning in the minds of two or three individuals the periods of several conventional generations. As clues, at least, they are never to be ignored. In this case we were taken to a rubble-strewn site on an eminence that overlooks Potomac Creek. At the foot of a declivity below, on the old Belle Plains road, we were shown another obvious evidence of structure, which we were told had been the jail. Just to the east of this where a road leads away to the site of Cave’s tobacco warehouse (now the “Stone Landing”), we were informed that the stocks had once stood.

Of the latter two sites we have no confirming evidence, although both claims are plausible enough. No archeological effort was made to investigate them, since funds were limited. The surveys of 1743 and 1805 are sufficient to confirm with accuracy the courthouse site. Accordingly, an archeological exploration was made between August 19 and August 23, 1957, revealing unmistakably the footings of a courthouse. As will be shown, these footings in no way bore a resemblance to the Structure B foundation.

HISTORICAL BACKGROUND

The history of the Potomac Creek courthouse site has been presented thoroughly by Happel, but a brief review is in order here. Happel shows that a courthouse was ordered built in 1665, a year after the establishment of Stafford as a county. He quotes a court reference in 1667 to the road along the south shore of Potomac Creek, running from the “said Ferry,” near the head of the Creek, “to the Court house to the horse Bridge,” which he identifies as having spanned Passapatanzy Gut. In his opinion, this courthouse was near the mouth of the Creek, but he fails to show that it equally well may have been near the site of the later 18th-century structures.

We have seen that in 1690 court was first held in Thomas Elzey’s house, seemingly located near the 18th-century courthouse site, and that orders were given that it continue to meet there until the new courthouse was ready. The history of the new courthouse at Marlborough has already been recounted, its final demise occurring about 1718. The court’s official removal from Marlborough was agreed upon July 20, 1720, and, as already noted, “the head of Ocqua Creek” was designated for the new site, although obviously by error, since Potomac Creek plainly was intended.

Happel tells us that the Potomac Creek building burned in 1730 or early 1731 and that the justices were ordered on April 27, 1731, to rebuild at the same place. It is this next building that was depicted on the 1743 survey plat (see fig. 58). In 1744 a bill was presented in the Assembly to relieve persons who had suffered or “may suffer” from the loss of Stafford County records “lately consumed by Fire”;[162] apparently the courthouse had again burned. There seems to have been a delay of about five years in rebuilding it this time. Pressures to relocate it were exerted in the meanwhile and hearings were held by the Governor’s Council on a petition to “remove the Court House lower down.”[163] The Council listened, then "Ordered, that the new Court House be built where the old one stood."[164]

This settled, Nathaniel Harrison and Hugh Adie contracted in 1749 with the justices of Stafford court to build a “Brick Courthouse, for the Consideration of 44500 lb. of Tobacco, to be furnished by the last of October, 1750.”[165] Harrison was a distinguished member of the colony who, as a widower, had moved to Stafford County the previous year and had married Lucy, the daughter of Robert (“King”) Carter of “Corotoman” and widow of Henry Fitzhugh of “Eagle’s Nest.”[166] Harrison, who later built "Brandon" for himself in King George County, probably provided the capital and the materials, and perhaps the design, of the courthouse. Adie, of whom nothing is known, was doubtless the carpenter or bricklayer who actually did the work.

The construction was delayed by “many Disappointments, and the Badness of the Weather.” Finally, in the spring of 1751, it was about to be brought to completion, “when it was feloniously burnt to the Ground.”[167] In April 1752 a special act was passed in order to permit a levy to be made which would allow the Stafford court to reimburse Harrison and Adie for the amount of work which they had accomplished on the courthouse and the value of the materials they had provided.[168]

No record exists of the contract for the next—and last—courthouse building on the Potomac Creek site. Quite possibly Harrison and Adie again did the work. This building was used until removal of the court to a new building completed between 1780 and 1783 on a site near the present Stafford courthouse. It remained standing throughout most of the 19th century, according to local memory. In surveys of 1804 and 1805 the structure was identified as the “old court house.”

DESCRIPTION OF EXCAVATIONS

Excavations were conducted in the simplest manner possible, in order to arrive at the objective of determining the dimensions of the courthouse without exceeding available funds. An exploratory trench soon exposed a line of rubble and disturbed soil. This line was followed until the entire outline of the building was revealed. At several points bricks in mortar still remained in situ, especially at the south end. Two brick piers extended 4 feet 5 inches into the structure, midway along the south wall at a distance of 5 feet 9 inches apart.

Portions of wine bottles (of the same dimensions as the Mercer “1737” bottle from Marlborough) were found (ill. 5), along with small fragments of late 18th-century types. A section of the rim of a large, octagonal, white, salt-glazed-ware platter with a wreath and lattice design was recovered from the north-wall footings (ill. 86), and fragments of a salt-glazed-ware dinner plate occurred in the south trench. An oystershell found nearby suggests how the platter may have been used. Two pieces of a white salt-glazed-ware posset pot round out a picture of elegant eating and drinking in the 1760’s, as do the fragments of polished, agate octagonal-handled knives and forks. The latter were badly damaged by fire.

Pieces of blue-and-white delft punch bowls were found, as well as a sherd of polychrome delft which dated apparently from 1740 to 1760. Two sherds of creamware plates with wavy edges in the “Catherine” shape reflect the last years of official use of the courthouse. A tantalizing find is a small fragment of cobalt-blue glass, blown in a mold to make panels or oval indentations. This piece may have come from a large bowl or sweetmeat dish.

Three sherds of black-glazed red earthenware are the only evidence of utilitarian equipment. Pipe-stems belong to the mid- and late-18th-century category. A George II copper penny is dated 1746. A large mass of pewter, melted beyond recognition, was found near the south end of the structure. Bits of charcoal are held within it. The pewter originally may have been in the form of mugs or tankards.

Evidence of the structure is found in a large number of hand-forged nails, in quantities of window glass melted and distorted, and in pieces of plaster. The last is the typical hard, coarse oystershell plaster of the area, having a smooth surface coat, except for fine lines left by the trowel. There is no evidence of paint. A small slide bolt of wrought iron probably fitted on a cupboard door, or possibly the gate in the bar (ill. 87). Another iron fixture is not identified.

Two kinds of window glass occurred. One, the earliest type, is a thin, yellowish glass which is coated with irridescent scale caused by the breakdown of the glass surface. None of this glass shows signs of fire or, at least, of melting. The remainder is a grayish-blue aquamarine, much of it melted and distorted, and some of it accumulated in thick masses where tremendous heat caused the panes literally to fold up. A fragment of yellowish-green glass pane, related to the early type and again coated with scale, varies in thickness and was apparently from a bullseye. No evidence exists of diamond-shaped panes, but, as should be expected, there is indication of square-cornered panes in both types of glass.

ARCHITECTURAL ANALYSIS

The plan of the footings (fig. 60) shows a T-shaped foundation. This was an immediate clue to the nature of the structure, for the T-shaped courthouse was virtually a standard 18th-century form in Virginia. This foundation, in fact, is almost a replica of the plans of both King William and Hanover County courthouses, each built about 1734[169] (figs. 5, 61, and 62).

The King William courthouse measures 50 feet 4¼ inches long and 26 feet 4 inches wide in the main structure. Its T section extends 14 feet 9 inches to the original end (to which an extension has been added) and has a width of 23 feet 10¼ inches. The Stafford foundation is 52 feet long and 26 feet wide in the main structure. The T-section is 14 feet 5 inches long and 25 feet wide. A closer comparison could scarcely be expected.

Hanover’s length is 52 feet 4½ inches, the width of the main section 27 feet 10 inches, while the T-section is 15 feet 2½ inches long (in its original part) and 26 feet 7 inches wide.

A third example, completed in 1736, is the Charles City County courthouse.[170] The measurements of this building are not available to us, but close examination of photographs discloses a building of about the same size.

The earliest of these T-shaped buildings thus far recorded was the York County courthouse, completed in 1733. Destroyed in 1814, its site has been excavated by the National Park Service. Its foundation, measuring 59 feet 10 inches in length and 52 feet in full depth, including the T, was somewhat larger than the others known to us. The records show that it was rather elaborate, with imported-stone floors and compass-head windows.[171]

All these buildings had arcaded verandas. Marcus Whiffen raises the question as to which of them, if any, was the prototype, then concludes by speculating that none was, and that all four may have derived from the 1715 courthouse at Williamsburg, the dimensions of which, however, remain unknown. The introduction of the loggia first at the College of William and Mary and then at the capitol led him to postulate that its use in a courthouse also would have originated in Williamsburg.[172] The Stafford foundation showed no trace of stone paving where an arcade might have been, but, since virtually all the bricks had been taken away, it is likely that such a valuable commodity as flagstones also would have been removed as soon as the building was destroyed or dismantled. Two brick piers at the west end of the structure (fig. 36) remain a mystery. They are equidistant from the longitudinal walls, and may have been the foundations for a chimney. However, their positions do not relate to the floor or chimney plans at Hanover or King William courthouses, the other features of which are so nearly comparable. One would suppose every basic characteristic of the Stafford building would have been the same as in these buildings. The piers were perhaps late additions or modifications.

The roof was apparently of wood; there were no evidences of slate shingles. The bricks were approximately 8½ inches by 4 inches by 2¾ inches, and were probably laid in a patterned Flemish bond, as at Hanover or King William, since some of the bricks were glazed. No lead or other signs of “calmes” used in leaded sash were found, so we must assume that the 1665 courthouse was built elsewhere.

CONCLUSION

It may be assumed that the Potomac Creek courthouse, which was built of brick, resembled the courthouses of Hanover, King William, and Charles City, and that its architecture, symbolizing the authority of Virginia’s government, reflected the official style expressed in the government buildings at Williamsburg. All the successive Stafford courthouses from 1722 on probably were built on the old foundations; if so, the Stafford building was the earliest T-form courthouse yet known in Virginia. Its similarity to the three structures built in the 1730’s shows that an accepted form had developed, possibly, as Whiffen suggests, deriving from a prototype in Williamsburg.

The courthouse bears no resemblance, either in its shape or the absence of a basement, to the Structure B foundation at Marlborough. The site, reached more easily than Marlborough from any direction, dictated the removal to it of the courthouse in 1722, thus contributing to the demise of Marlborough as a town. The last structure, especially, was historically important because of the meetings of the Ohio Company held in it. It is of particular interest to the story of Marlborough because John Mercer was, for most of its existence, the senior justice of the Stafford court.

ARTIFACTS

XV

Ceramics

Most of the ceramic artifacts found at Marlborough can be dated within John Mercer’s period of occupancy (1726-1768). A meager scattering of late 18th- and early 19th-century whitewares and stonewares reflects the John Francis Mercer and Cooke ownerships (1768-1819).

COARSE EARTHENWARE

Tidewater type.—Mercer’s purchase in 1725 of £12 3s. 6d. worth of earthenware from William Rogers (p. 16, footnote 54) probably was made for trading purposes, judging from the sizable cost. Rogers operated a stoneware and earthenware pottery in Yorktown, which evidently was continued for a considerable time after his death in 1739.[173] An abundance of waster sherds (unglazed, underfired, overfired, or misshapen fragments cast aside by the potter), supposedly from Rogers’ output, has been found as street ballast and fill in Yorktown and its environs. Microscopic and stylistic comparison with these sherds relates numerous Marlborough sherds to them in varying degrees. For purposes of tentative identification, the ware will be designated “Tidewater type.” Some of the ware may have been produced in Rogers’ shop, while other articles resembling the Yorktown products may have been made of similar clay and fired under conditions comparable to those at Yorktown.

A Marlborough milk pan (USNM 59.1961, ill. 11, and USNM 59.1580) has a salmon-colored body and a lustrous mahogany glaze with fine manganese streaking. Another milk pan (USNM 59.2039, ill. 2, fig. 63a) has a buff body and a glaze of uneven thickness that ranges in color from thin brown with black flecking to a glutinous dark brown approaching black. The most typical glaze color, influenced by the underlying predominant pinkish-buff body, is a light mahogany with black specks or blotches. It occurs at Marlborough on a small sherd (USNM 60.201). A variant glaze occurring on pottery found in Yorktown appears here in a yellowish-buff sherd flecked with black (USNM 60.154). The flecking is only in part applied with manganese; it is also the effect of ocherous and ferruginous particles which protrude through the surface of the body, assuming a dark color. Occasionally the manganese is spread liberally, so that the natural body color shows through only as flecks in a reverse effect (USNM 59.1855); now and then the vessel is uniformly black (USNM 60.141).

Tidewater-type forms found at Marlborough include milk pans 15 inches in diameter and about 4¼ inches deep (in 1729 Mercer bought “2 milk pans” for 5d. and 5 “gallon basons” for 4s. 7d.), a black-glazed jar cover with indicated diameter of 6½ inches (USNM 59.2013), and fragments of other pans and bowls of indeterminate sizes. A portion of an ale mug has a tooled base and black glaze (USNM 59.2043, fig. 63d, ill. 12). Its diameter is 3⅚ inches.

Molded-rim type.—This is a type of redware with a light-red body and transparent, ginger-brown lead glaze. It is characterized by a rolled rim and a tooled platform or channel above the junction of rim and side. A small number of pan and bowl rims was found at Marlborough. The ware is usually associated with early 18th-century materials from such sites as Jamestown, Kecoughtan, Williamsburg, and Rosewell. It may have originated in England.

North Devon gravel-tempered ware.—The coarse kitchenware made in Bideford and Barnstaple and in the surrounding English villages of North Devon is represented by only two sherds. This ware is characterized by a dull, reddish-pink body, usually dark-gray at the core, and by a gross waterworn gravel temper. It occurs in contexts as early as 1650 at Jamestown and as late as 1740-1760 at Williamsburg. One of the Marlborough sherds is part of a large pan. It is glazed with a characteristic amber lead glaze (USNM 60.202). The other sherd is a portion of an unglazed handle, probably from a potlid (USNM 59.1679, ill. 15).[174]

Slip-lined redware.—Numerous 18th-century sites from Philadelphia to Williamsburg have yielded a series of bowls and porringers characterized by interior linings of slip that is streaked and mottled with manganese. These are glazed on both surfaces, the outer surface and a border above the slip on the inner surface usually ginger-brown in color. Comparative examples are a bowl from the Russell site at Lewes, Delaware, dating from the first half of the 18th century, and several pieces from pre-Revolutionary contexts at Williamsburg. A deposit excavated by H. Geiger Omwake near the south end of the Lewes and Rehoboth Canal in Delaware included sherds from a context dated late 17th- to mid-18th centuries.[175] Several fragments of bowls occur in the Marlborough material (USNM 59.1613, 59.1856, fig. 64g).

English yellowware.—The few sherds of so-called combed ware occurring at Marlborough, although only the base fragments connect, all seem to have come from a single cup or posset pot having a buff body and characteristically decorated with spiraled bands of dark-brown slip that were created by combing through an outer coating of white slip, revealing an underlayer of red slip. The vessel was glazed with a clear lead glaze (USNM 59.1700, fig. 64c, ill. 16). Comparative dated examples of this ware include a posset pot dated 1735.[176] A chamber pot bearing the same kind of striping was excavated by the National Park Service at Fort Frederica, Georgia (1736-ca. 1750). A piece similar to that from Marlborough was found in the Rosewell deposit, and another in the Lewis Morris house site, Morrisania, New York.[177] Although this type of ware was introduced in England about 1680, its principal use in America seems to have occurred largely between 1725 and 1775. Archeological evidence is corroborated by newspaper advertisements. In 1733 the Boston Gazette advertised “yellow ware Hollow and Flat by the Crate” and again in 1737 “yellow and Brown Earthenware.” In 1763 the Gazette mentioned “Crates of Yellow Liverpool Ware,” Liverpool being the chief place of export for pottery made in Staffordshire, the principal source for the combed wares.[178]

Buckley ware.—I. Noël Hume has identified a class of high-fired, black-glazed earthenware found in many 18th-century sites in Virginia. He has done so by reference to The Buckley Potteries, by K. J. Barton,[179] and to waster sherds in his possession from the Buckley kiln sites in Flintshire, North Wales. The ware probably was made in other potteries of the region also. This durable pottery, more like stoneware than earthenware, is represented by a large number of jar and pan fragments. Two body types occur, each characterized by a mixture of red and buff clay. In the more usual type the red clay dominates, with laminations and striations of buff clay running through it in the manner of a coarse sort of agateware. The other is usually grayish buff with red streaks, although sometimes the body is almost entirely buff, still showing signs of lamination. The glaze is treacly black, often applied unevenly and sometimes pitted with air bubbles. The body surfaces have conspicuous turning ridges. Rims are usually heavy and flat, sometimes as wide as 1½ inches. A variant of the ware is represented in a milk pan with a dominantly red body which has a clear-amber, rather than black, glaze. (USNM 59.1887, ills. 17, 18, and 19 and fig. 65).