the Virginia Museum

The Peter Scott Shop

The career of Peter Scott (1694-1775) spanned five decades and is distinguished as the longest of any Williamsburg cabinetmaker. I le is first documented in the town during 1722 and had established his shop on the south side of Duke of Gloucester Street, opposite Bruton Parish Church, by 1733.' Little is known of his early years, and most of the important details regarding his life and business are gathered from six Virginia Gazette notices. It is significant that the earliest of these, the only one placed by Scott, was printed more than three decades after his first known appearance in Williamsburg. It was not intended to promote his business but instead to settle his affairs and announce his intended departure for England:

To Be S C) L I)

BEFORE Mr. Finnic’s Door, on the 23rd Day of October next, Two Lots of Ground, situate on the Back Street, near Col. Custis’s in Williamsburg; on w hich there is a good Dwelling House, containing Six Rooms and Closets, a good dry Cellar, w ith all convenient Out-I louses, and a good Well: Twelve Months Credit will be allowed the Purchases giving Bond and Security. At the same Time and Place w ill be sold, for Bills of Fxchangc or ready Money, Two Negroes, bred to the Business of a Cabinetmaker; likewise will be sold, at the Subscriber’s Shop near the Church, sundry Pieces of Cabinet Work, of Mahogony and Walnut, consisting of Desks, Book­Cases, Tables of various Sorts, Tools, and some Materials. Six Months Credit will be given to those that purchase above the Value of Fifty Shillings, on their giving Bond and Security; and Five per Cent, will be allowed for ready Money.

And as 1 intend to go for Great-Britain the latter Fnd of next Month, therefore, I desire all Persons indebted to me, to make speedy Payment, otherwise they may expect Trouble without further Notice.

Peter Scott2

Although this advertisement notes that Peter Scott ow ned tw o lots on “Back” (Francis) Street with a house and outbuilding, it does not tell us that he rented a dwelling and/or shop on Duke of Glouces­ter Street (colonial lot 354). This was granted in 1717 to John Custis, Martha Washington’s first father-in­law, who w ill re-enter the story at a later date. It w as a convenient business location, and Scott inhabited the house from 1733, until his death on the eve of the Revolution:3

December 2, 1775 Deaths ....

Mr. Peter Scott, in the 81st year of his age, he was upw ards of forty years a Common Councilman of this corporation.4

It is unfortunate that the early court records for James City County were destroyed and that no documentation of Scott’s council activities or his

estate survives. The only known reference to the settlement of his personal property is found in a notice shortly after his death:

January 13, 1776

To be SOLI) before Robert Nicolson’s Store, on Tuesday the 16th instant,

A GREAT variety of cabinet-makers tools, mohogany, walnut, and pine plank, likewise new walnut book cases, desks, tables, &c. belonging to the estate of mr. Peter Scott, deceased. Six months credit will be allowed for all sums above 51. the purchasers giving bond w ith good security.

All persons indebted to the said estate, by bond or open account, are requested to pay off as soon as possible; and those to whom the estate are indebted are desired to call and receive payment, from

EXECUTORS»

Less than tw o months after Scott’s death, and shortly follow ing the sale of his estate, the old house in which he had lived was a casualty of the American Revolution:

January 26, 1776

Mr. Peter Scott’s old house in this city, which he had rented and lived in for fourty-three years, w as burnt down last Sunday night by accident.®

for sale. While he does not advertise custom cabinetwork for sale, the incidental ails in 1755 and 1776 list various items that must have been on hand for direct sale from his shop. Despite the time span between these notices, there is a striking similarity between the items listed, and a consistent pattern seems to emerge. The evidence strongly suggests that in addition to custom order, it was normal practice to keep a large stock on hand, that sales w ere regular, and that demand w as so great as to eliminate the need for advertisement. This also held true in 1766 when W illiamsburg cabinetmaker John Ormcs - ton announced his intention to move to W ilmington, North Carolina. 1 le, too, had completed furniture on hand to sell.9

Although there is no appeal for journeymen or apprentices in Scott’s lone Virginia Gazette notice, it is highly likely that they w ere employed in his shop. The surviving records indicate that he had a fairly large operation, and over a period of five decades he undoubtedly trained apprentices and employed some journeymen. A large distribution of simple chairs, all having Scott construction, corroborates this assumption. Perhaps the answer to this lack of advertisements is his extreme popularity, for he w as established in Williamsburg at an early date and was patronized by many wealthy Virginians by mid­century. I le was so well-know n, in fact, that in Gazette notices about the burning of the house, which actually belonged to the Custis family, it was referred to as “Mr. Peter Scott’s old house in this city.” Scott’s residence is mentioned, matter of factly, as a familiar landmark. Presumably his prominent location and his popularity attracted prospective apprentices and journeymen thereby eliminating any need to advertise. This appears to have been the case regarding custom orders and direct sales of his shop’s products, since no adver­tisements for these goods are known to have ever existed.

Scott’s 1755 Gazette notice of his plan to go to Great Britain mentions his ownership of two black cabinetmakers. If indeed Scott ever left, and if he sold them prior to his departure, then he had purchased or hired another by November 2, 1772, w hen Thomas Jefferson noted in his account book “pd. Peter Scott in Full £16 gave negro man at Peter Scotts 5s.”10 By the time of Scott’s estate sale in 1776, there were no black cabinetmakers listed, indicating that the one mentioned in 1772 had been hired from another owner or was sold in the meantime, or perhaps his fate w as decided by Scott’s w ill, which no longer exists. These incidental refer­ences seem to indicate that black cabinetmakers were an integral part of his shop operation, and that they
played an important role in cabinetmaking in early Williamsburg.

Occasional documentation for Scott’s custom orders survives, and from them vc learn something of his production and activity. One of the earliest is a 1748 entry regarding a desk in Colonel W illiam Bassett’s account book." A desk that descended in the Bassett family of Ilanovcr County may be the same one referred to in this account and will be discussed in depth later in this book (fig. 35). In 1748, 1749, and 1750, Scott received payments from John Mercer, a law yer who lived near Fredericksburg, but the work w as probably for Mercer’s two children who studied at the College of William and Mary during those years. One of the more enlightening entries in Mercer’s daybook is a debit to Scott for “mending of a table for my sons.” This account is listed under the entry “Peter Scott joiner at W il­liamsburg.”12 In another receipt, unfortunately not dated, Scott charged Robert Carter of Nomini Hall for two card tables, a sideboard, and four picture frames.13 It was not until April of 1785, almost 10 years after Scott’s death, that Carter authorized payment of £21.9.3, plus interest, to Scott’s estate for these items.11

Thomas Jefferson also had numerous accounts w ith Peter Scott. From 1771 until the cabinetmaker's death in 1775, Jefferson seems to have patronized him more than any other Williamsburg artisan. Several accounts are listed only as cash sums paid, but a few are more detailed, such as an order recorded in 1772: “wrote to Scott to make table 4f I I. st] and 2 f. 41, high. My tea tables were directed to be 3. f. by 2.f. & 2. f. 3.2 I. high.”15 A ball-and-claw - foot table at Monticello may relate to the first portion of this account (fig. 3 1). A memorandum, also dated 1772, gives measurements that may be preliminary plans for the tea tables in the second portion of the Jefferson account: “Tea table when leaves down 2f by If IVi I, when leaves up 2 f by 2 f9 I I leight 2 f 3Ун I.”16 Another order, dating 1773-74, specified “clothes presses to be made by Scott 4 f 6 I wide.

I ligh as a desk and bookcase.”17 It is fortunate that a clothespress matching this very unusual description (fig. 42) survives among the furnishings owned by Robert Beverley of Blandficld.

No signed or labeled piece of Scott furniture is known at this time, but strong circumstantial evi­dence argues for the attribution of a large group to the W illiamsburg area, and more specifically to Peter Scott’s shop. A summary of this evidence, divided into several groups, is necessary to make the basis for this attribution as clear as possible. The first pro­vides strong reasons for assigning the group to the city of Williamsburg.

—A clothespress in this group descended in the («alt family of W illiamsburg (fig. 36).

—Case pieces in this group share a relationship with other W illiamsburg shops (figs. 50, 77, 85, 86).

—A set of Scott chairs, including a matching settee and corner chair, has a firm history from the Governor’s Palace (figs. 32-34).

—The chairs in this group have strong con­structional ties to other Williamsburg shops (figs. 49, 58-62, 64, 96).

—The application of their knee blocks over the seat rail is the same as that found on the Speaker’s chair in the Capitol (fig. 7).

—Scott chairs are also related in construction to a set that descended in the Benjamin Waller family of Williamsburg (fig. 62) and to a lone example with a history in the Governor’s Palace (fig. 58).

Two furniture fragments excavated at the Anthony I lay site also suggest a W illiamsburg origin for this group. One of these, an casy-chair leg, is related to chair legs in the Scott group (see fig. 65). The second, a fragment of a slip seat, utiliz. es the same upholstery technique as the two Palace sets (figs. 58 and 67).

Since constructional features common to this group are also found in other documented Wil - iiamsburg shops, particularly those from the I lay site, it is necessary to consider the documentary evidence in order to specifically attribute these pieces to Peter Scott:

—The Jefferson ball-and-claw-foot table (fig. 31), possibly conforming to Jefferson’s 1772 order to Scott.

—The Bassett family desk-and-bookcase (fig.

35) , possibly the one recorded in Colonel Bassett's 1748 account due Scott.

—-The Blandfield clothespress (fig. 42) match­ing the description of one ordered by Jeffer­son from Scott.

Finally, the long period of time spanned by the objects within this group corresponds to Scott’s cabinetmaking career. While other cabinetmakers were working in Williamsburg, no other Wil­liamsburg cabinetmaker is known to have been working throughout this period. The geographic distribution of this group, as well, corresponds closely to the documented patronage of Scott and other Williamsburg cabinetmakers.

I laving discussed Scott’s background and hav­ing set out the basic points that substantiate the attribution of this group to him, w e can proceed to a stylistic and chronological study of the pieces. The George II style and the construction of earlier pieces within the group parallel Knglish development in the
1730 period. From that point a chronology is indi­cated as the George II style gives way to the rococo. That change is paralleled by a gradual decline in the quality of carving, which is not surprising consider­ing Scott’s advancing age. Several of these pieces can be associated with documents or historical cir­cumstances that help to establish an approximate date. A list of the details that characterize the chairs in this group helps clarify their relationship:

—Knee brackets. These are applied to the face of the seat rails that pass behind them (lig. 19a). The clear outline of the missing bracket remains on the face of the seat rail.

—Shoe. The shoe and the rear rail are a single piece, with the bottom profile of the rear rail cut out. Occasionally the outer portions of the shaped bottom edge are separate pieces that have been glued on (fig. 16b).

—Stiles. Invariably parallel, they flare outw ard just below the crest rail.

—Splats. The splats are heavily undercut on the back, with a pronounced bevel that is very well finished (fig. 16b).

—Arms. 1'hese are dovetailed into the stile and are reinforced with a screw. This dovetailing is visible from the rear of the stile (fig. 22a).

The first five chairs shown (figs. 16, 18-21) are of the highest quality and are stylistically termed George 11, w ith controlled carving that accents large, plain spaces. The scale of the carving is well proportioned in relation to the chair elements, producing a clarity of overall design that is often not present in later, rococo examples. I lere the design “reads" well and is not cluttered or confused.

The first chair (tig. 16) has several details that indicate it is closest to its London counterparts and, theoretically, the earliest of the group. Its cabriole legs have a more pronounced curve, w hich is most noticeable on the inside at the top behind the knee. More wood has been removed from this area than on later chairs, giving it a graceful, flow ing character. Likewise, the finish work of the undercutting on the bottom profile of the seat rails is more refined than on later examples, and the carved shell with bellflower pendants and plain knee brackets are the earliest type. According to tradition this chair descended in the family of Alexander Spots wood (1676-1740), lieutenant governor of Virginia from 1710 to 1722.

17. Side Chair, London, circa 1130.

Walnut and walnut veneer primary.

Height 42 V*", width 20V*", depth 20'A".

The Colonial Williamsburg Foundation (acc. no. 1936-219).

18. Armchair, attributed to the shop of Peter Scott, Williamsburg, circa 1141.

Cherry primary.

Height 39s/i«", width 243/i", depth 18V*".

Shirley Plantation, Charles City County, Va.

A pair of impressive armchairs (fig. 18) that are original furnishings of Shirley Plantation have de­sign elements related to the Spotswood chair. Their arms are well formed, w ith serpentine supports that terminate in carved dog’s heads. A pair of side chairs in the Virginia Historical Society (fig. 19) was probably a part of this set. These arc also made of cherry and are identical in height and carved detail.

Less elaborate than the preceding example, and fashioned of black walnut, is a side chair recently found in Rhode Island that has an obscure Virginia background (fig. 20).Is Its carving is very well executed and it retains several original voluted knee brackets. The splat is a less expensive variation of the earlier examples, as is the crest rail.

One of the rarest American easy chairs (fig. 21) descended in the Washington family and is thought to have belonged to Mary Ball Washington. The hold legs and feet are typical of other examples in this group, but their visual impact would be much improved by the restoration of the knee brackets and the removal of the added stretchers. The support blocks of the brackets survive and suggest large volutes similar to the preceding chair and to several corner chairs as well (figs. 28, 29). The rear cabriole legs, with ball-and-claw feet, are the only known examples from colonial America, although several Philadelphia chairs of this form have pad or trifid feet on the rear.

Stylistically there are two additional features that tie these pieces to the second quarter of the eighteenth century. The dog’s-head terminals found on several of the armchairs offer an interesting alternate to the more popular eagle’s - and lion’s-head terminals found in quality F. nglish work. Occasion­ally, New York chairs have eagle’s-head terminals, but Williamsburg is the only other center in America that produced sculptural animal forms on the arms of chairs. A second feature, originally found on the

Washington easy chair (fig. 21), arc large volutes flanking the knees. Their design is closely related to the feet and knee blocks on two desk-and-bookcascs (figs. 37, 39).

The first definite indications of the rococo style evident in this group are found on the knee brackets of a side chair (fig. 20). Above the large volute on each bracket is carved a short C-scroll with a slight shell appendage on its top. A small segment of this same design can he seen on the ends of a gadroon molding on the front of an armchair (fig. 22b).

A mahogany armchair (fig. 22) also show s signs of the change to rococo style. The ears of the crest rail have lost their vigorous expanding volutes and are replaced by w eak scrolls that are molded across their fronts. This molding originates beneath a small acanthus carved at the juncture of the crest and stiles and continues until it rolls over the ear and disap­pears around the curve. Unfortunately, the knee blocks are missing, but their carving was probably the same as those on the walnut side chair (fig. 20). The design of the knee carving is identical through­out, although the finish work of the armchair is slightly inferior. This workmanship is particularly noticeable in the background behind the relief carving, which shows many tool marks. T his armchair has an oral history in the Ferneyhough family of Fredericksburg and is virtually identical to another armchair in a private Richmond collection, w hich has a long verbal tradition of ow nership by Jefferson. This example came to the author’s atten­tion too late to be illustrated in this study, but it is extremely important in light of Jefferson’s numerous accounts with Scott.

22. Armchair, attributed to the shop of Peter Scott, Williamsburg, circa П55.

Mahogany primary.

Height W/i", width 28 Vs", depth IS".

The Mary Washington House of the Association for the Preservation of Virginia Antiquities.

Another side ehair with a Fredericksburg his­tory has a simple design without carving and typifies the sober approach of the largest number of chairs surviving in the Scott group (fig. 23). This and a matching armchair from King and Queen County (fig. 24) show many concessions to the rococo style. A change in the handling of the voluted cars on the crest rail of the preceding mahogany armchair (fig. 22) has already been discussed. Here, however, the crest has evolved into a completely different form, and the voluted ears of earlier examples have become rounded into large, swelled lobes. The form is clearly rococo, although the saddle in the center of the crest, the parallel stiles, and the splat design remain from the earlier style. On the armchair the S-shaped arm support is retained, considerably w eaker in form, and a dog’s-head terminal has been dropped in favor of a small, down-curved volute. The construction is the same as carved armchairs in the group, where shoe and rear seat rail are an integral piece with horizontal shaping, and the arms are dovetailed into the stiles.

Quite consistent with these examples in its construction is a child’s armchair (fig. 25). Its serpentine arm support, with a dovetailed arm fitting, is closely related to the preceding chair (fig. 24). The arm terminal is a variation seen on several later chairs in the group. Unfortunately, the crest rail, splat, and front seat rail are replacements. The original crest was probably the squared form, as are the other known examples in this group. The knee brackets, too, are replaced, although the originals were types that overlapped the rail. According to tradition, this chair was an original furnishing of Belmont at Falmouth near Fredericksburg. Also worth noting here is a corner chair w ith the same history, which has similarly shaped arm terminals (fig. 28).

Another child’s chair (fig. 26), found in I)in - w'iddie County near Petersburg, is closely related to the preceding. It, too, has lost its original splat, its crest rail, and the top one-third of its stiles, which appear to have been broken just below the arm joint. This chair has cabriole legs similar to the preceding example and retains its original knee brackets over­lapping the rails. It differs in two respects, however. The arm supports, now cut off even with the rails, were originally an extension of the front legs; the shoe is a separate piece, although the rear rail has horizontal shaping conforming to the group. This separate shoe, however, could be an alteration that accompanied the loss of the original splat and crest rail.

Possessing elements of the previous simple and elaborate examples is a walnut side chair with a tradition of ownership by George W ashington at Mount Vernon (fig. 27). Its splat, though not original, is a very old mahogany replacement, the design of w hich was probably taken from the original, since it is continuous with the design of the crest rail and is heavily undercut in imitation of the Scott group bevel. All other elements that survive

arc original. I Icrc the rounded ear of the crest rail is completely rococo in design, with the top profile formed by three C-scrolls (fig. 27a). The quality of the carving has declined from that of the earliest chairs; this is most apparent on the acanthus leaves of the rounded ears, and on the central leaf element of the crest rail. The cabriole leg and foot also show a loss in quality: the leg is very straight and the foot is somewhat smaller than the earlier examples. There was also carving on the skirt, which originally had a gadroon molding glued along the lower front edge, identical to construction seen in the Ferneyhough chair (fig. 22). Although now lost, the glue mark and undercutting for it are quite obvious. In addition to the rococo back and crest, the symmetry of earlier knee carving has been broken by adding a curled ending to the acanthus leaf.

The follow ing corner chairs (figs. 28, 29) are unusually constructed. Their crests have an arch on each side of the central column and the joining of the skirt to the cabriole leg is very uncommon. One other chair of this form is pictured in the MF. SDA files (#S-6591), although it is much plainer and has straight chamfered legs. These chairs, particularly the two illustrated here, fall solidly w ithin the Scott group. Their carving and major design elements are identical to, or variations of, details seen on the earlier chairs, and they continue the decline in the quality of carving already noted in the preceding examples.

The first of these corner chairs (fig. 28) is conceptually related in design and execution toother chairs in the Scott group. The knee carving, while initially appearing somewhat different, reveals upon closer study that virtually all of the elements come from the preceding examples and are merely rear­ranged. The basic formula was used in designing the earlier chairs (figs. 18b, 19a). It was altered by reducing the shell and moving it upward, away from the knee (fig. 28a). The inward-facing volutes flank­ing the large shell have been moved up until they touch each other and surmount the small reduced shell. Oddly, the bellflower, which is in the center of the knee’s maximum protrusion, has not been moved in the new design but sits awkwardly between the extended acanthus leaves. At the top, branches of the acanthus flare toward the knee blocks, on which they would have continued but now, unfortunately, the blocks are missing. This feature, too, is remarkably similar to the wing-like features that extend beneath the shell and tow ard the knee block (fig. 18b). The knee blocks on figure 18 are replaced, however, and the continuation of these wings cannot be trusted as correct. Below the central bellflower is a scries of grains very similar to those that terminate the knee

carving of figures 2()a and 22b and arc interspersed on figure 16c. These sit upon a large acanthus Hanked by overlapping fronds, virtually identical to those on figures 20a and 22b. Finally, the major acanthus terminates in a folded-back overlap, which is a slight variation of the curved termination of figure 27.

This chair was produced by employing a com­bination of designs that w ere in use two decades earlier. While the foot carving approximates the earlier quality, the knee carving is more crudely executed and is not as successful. The re­combination of these elements produces a slightly clumsy, cluttered design, and the sculptural quality is flatter and somewhat cruder. This example is important, however, as an indication of the extreme effects traditional training could instill in the eighteenth-century tradesman.

The left splat of this corner chair is an old replacement and is coarser in execution than the original on the right. Their design is reminiscent of
the interlaced diamond pattern found on some of the earlier Scott side and arm chairs (figs. 16, 18, 22). This example represents a major departure from traditional corner chair construction by employing four pieces of wood in the crest and arms rather than three. The arms do not meet in the back and the extra piece, centrally glued to the bottom of the crest, is mortised to receive the rear column. An apparent offshoot of this construction exists on several chairs belonging to another W illiamsburg group. Found in the MF. SDA research files, they too have arms that are not joined. These examples, however, lack the arches in the crest and are fashioned from a single piece of wood. This construction replaces the small glucd-on block found on the chairs illustrated here. Another chair in this latter group (fig. 96) does not have this solid construction, although it has a splat nearly identical in outline to that seen on the first Scott corner chair illustrated (fig. 28).

The second corner chair (tig. 29) is closely related to the preceding piece. This example, how­ever, has a ball-and-claw foot with a noticeably different approach (figs. 29a, 29b). It is higher, and probably represents a conscious attempt to depart from the early flattened ball, thus producing a piece more in line with changing fashion. Despite this difference, its affinity with earlier examples is evi­dent in its smooth, slightly knuckled “toes" and its lack of w ebbing. The claws are angular points with little definition to separate them from the toes— carving that is typical of the entire group although several examples have more detailed definition on the claws. The knee carving is a variation of double overlapping fronds seen on several other chairs, although greatly degenerated in quality (fig. 29a). I'he acanthus leaves are an elaboration of those on a side chair, an armchair, and a corner chair (figs. 20, 22, 28) but end in such extended points that they take on the appearance of hooks. The splat of this chair is

a simplification of that design most often used in the group. It has an inverted vase-shaped base with a central heart opening below a series of parallel vertical piercings. The carved volutes that terminate the arms are nearly identical to those on the crest rails of early chairs in the group.

Six side chairs ow ned by Thomas Jefferson are thought to be a part of a set acquired by him from George Wythe of W illiamsburg (fig. 30).19 These chairs, one of them numbered “XX," are the only examples with this splat design and crest rail known to the writer, although they have some similarity to the walnut side chair w ith a Palace history (fig - 58). Two piercings of the splat splav downward and out, and both have seat rails with well-developed moldings. I low ever, it appears that the crest rail on the Jefferson examples was produced for carved decoration, considering the shell outline at the extremes. In fact, their crest rails were cut to allow an extra thickness for the carving of the shell, but as
they were never executed they have a simple scallop cut just above the joint of the stile to disguise the offset (tig. 3()a). I'he Washington chair (tig. 27) is an example in which this allowance was actually carved. There are other pieces with this unusual feature, including a settee and chairs (tigs. 32, 34) from the Governor’s Palace. These are interesting examples and relate not only to the Jefferson chairs in the unused allowance for carving but also to the Washington chair in the profile of their ears. In those front the Palace (tig. 32, 34) the offset is left straight at the joint and lacks the refinement of the slight scallop.

This allow ance for carving, w ithout the follow - through, seems to indicate a large shop production of chairs. Where many sets were being made from patterns designed to accommodate carving, an order for the plainer chairs could have been supplied from the uncarved inventory examples, thus producing the somew hat awkward result seen in these chairs.

In addition to these chairs and several case pieces yet to be analyzed, one other furniture form survives in this group—a ball-and-claw-foot table.

I laving a tradition of ow nership by Jefferson, it is possibly the piece he ordered from Scott in 1772 (fig. З 1; see footnote 15). Once a full-sized dining table, it was cut to its present “breakfast” size in an alteration that is thought to have been done at Monticcllo during Jefferson’s lifetime. The heavy ankles and rounded knuckles are related to other feet in this group—particularly those on the side chair owned by Washington (fig. 27) and those on a corner chair (fig. 29). Its small claws, w hile differing from most in the group, arc like the surviving claw on the Mary Ball Washington easy chair (fig. 21). The rear toe (fig. 31b) protrudes at the top and is concave just above the claw—as are many in this group. It is not carved as a separate element but flow s down from the ankle in a smooth curve, and in this respect it is more like Knglish examples (fig. 17b). The only compar­able toe in this group is found on the rear feet of the easy chair (fig. 21), but since this is the only table in the group the difference may be a stylistic allowance for the taller leg. The front of all the toes have lost some definition through wear, but despite this allow ance they are shallow and the overall result is a cruder foot than others in the group.

The stiff, rather straight cabriole legs of this table are similar to later chairs of the group. The pointed knees of these legs are consistent with those found on case pieces (figs. 37, 39) and suggest a continuation of the corner of the case. Unfortu­nately, the knee brackets are missing, but some evidence remains and it is clear they were not the type that overlaps the rails of chairs in this group.

The knee bracket on the gate was not attached to the leg but was nailed and glued to the table skirt. When the leg swung out, it stayed on the skirt rather than moved in the typieal manner. This feature, like the sharp knee, parallels ease construction with cabriole legs in which knee brackets are glued beneath the case and not to its face. This unusual constructional feature may help in identifying other related tables, and it is reminiscent of the shaped brackets on the skirts of dining tables in the early Williamsburg group (figs. 9, 1 la).

The latest examples in the Scott group comprise a very large set of seating furniture with a history in the Governor’s Palace (fig. 32-34). According to tradition, they were sold in 1776 at the sale of Lord Dunmore’s property, having been left at the Palace when he fled Williamsburg the preceding year. They are said to have been purchased by Thomas Lewis of Augusta County and to have been presented by him to John Stuart and Agatha Lrogg Lewis of Greenbrier County, Virginia (now West Virginia) as a wedding gift. They remained in “Stuart Manor” near Lewis - burg, West Virginia until 1975, when they were purchased by Colonial Williamsburg from Stuart
descendants. It is quite important that the tradition of ow nership is supported by a book which de­scended in the same family. This has the engraved bookplate of John Murray, the Karl of Dunmore, surmounted by those of Thomas Lew is and John Stuart.

Evidence indicates the set was very large. One side chair seat is marked “X111” and apparently there was another settee, since the surviving example has a middle stretcher that was removed from another in an early repair.

These pieces probably date between 1771, when Dunmore arrived, and 1775, w hen he fled the colony. They display the coarsest workmanship of the entire group. As already discussed, they possess offset crest rails with an unused allowance for carving. In addition to this, the lower portion of the splat has the same vase-shape seen on many of the earlier chairs. Near the top of the splats are volutes emphasized by relief carving. These terminate the continuous flowing line of the crest rails low er profile and are similar to their antecedents (figs. 16, 18, 19, 20, 22), although not fully carved. Similar carving of splats occurs again and again in this group and

appears as well on two chairs from the Hay shop (figs. 59, 60). The awkw ard terminals of the arms of the settee are closely related to those on one of the corner chairs (fig. 29). On the settee, however, they have been simplified by the omission of the side volutes. Constructionally, these have all the features of the group outlined at the beginning of this section, w ith the exception of the substitution of a notch for the dovetails on the arms of the settee and the omission of undercutting on the splat. The settee has a rear seat that is integral with both shoes but lacks the horizontal shaping. These, plus the obvious differences in the corner chair, are probably due to the decline in quality seen all along—and to the cheaper aspects of this set in general.

I laving completed an examination of the forego­
ing chairs and table, and having established a chronology for them, the case pieces of this group can now be studied. They relate to the chairs through their identical ball-and-claw feet, common use of beech, and employment of primary woods in secondary usage. Constructionally, these cases exhibit a remarkably consistent approach:

—Case construction. Tops and bottoms are dovetailed into sides w ith hidden dovetails. The drawer blades are one-half dovetailed into the sides and enter from the front. They arc then covered along the vertical front edge by a glucd-on strip. The back is set into a rabbet on the top and sides and is secured by nails.

—Dustboards. These are thinner than the drawer blades and are wedged into grooves in the sides of the case by pine strips that run beneath them from front to back (Hg. 37d). The dustboards are half-lapped onto the drawer blades.

—Base molding. The base molding is not applied to the sides of the case but to a series of rectangular blocks glued to the case bot­tom. The top edge of the base molding overlaps the bottom edge of the case very slightly and sometimes has a very shallow rabbet (fig. 37e).

—Ogee foot construction. The ogee-bracket feet are made of two pieces of wood. The primary wood that forms the exterior is glued to a piece of yellow pine which makes up the inside thickness of the foot (fig. 40b).

—Composite foot. The blocking of the feet is composed of several horizontally grained pieces of pine, with the grain of every other piece perpendicular to the preceding. This stacked blocking is neatly finished on the front feet (fig. 86b) but is left at random on the back feet (figs. 40b; 86b and c). Although figure 86 is from another shop, it conforms to Scott’s in this respect; it is shown because those of the Scott examples have not survived completely intact. Since no terminology has been applied to this construction, that adopted here—“composite”—will be used throughout this work.

—Crow n molding. The crown molding is glued to a series of large rectangular blocks across the front. On the sides it is glued to a continuous block that runs the depth of the case and is mitered at the front corners (fig.

370.

—Draw er construction. The large case drawers have bottoms which arc set into a very deep rabbet on the front and two sides and are then nailed into place (fig. 37g). The nails on the sides are covered from front to back with a continuous strip that also forms the drawer runners. This continuous piece butts against the draw er front, and is cut off at a 45-degree angle on the back. Across the front edge is a series of small blocks with open gaps betw een them. The lowest dovetail on the sides of each drawer is straight along the bottom, conform­ing to the upper edge of the rabbet that receives the drawer bottom (fig. 37h).

—Desk drawers. The small drawers in writing interiors also have a straight edge on the bottom of the low est dovetails. The bottoms

of these smaller drawers are set into a shallow rabbet on all four sides and are Hush w ith the lower edge of the sides. They therefore have no runners and each drawer rides on the entire surface of its base.

—Door construction. The stiles and rails of the paneled doors are very broad, particularly the lower horizontal elements. They have blind tenons and are not usually pegged. The pine desk (fig. 39) is an exception, if, indeed, the pegging is original. The raised panels of these doors are not set into grooved stiles but fit into a rabbet. They are held in place by a small molding, cither quarter-round or flat-bead, nailed around all four sides (fig. 36a). The pine desk is an exception here also.

—Fa 11 board supports. The tailboard supports of. desks have a vertically grained facing that is joined by a tongue-and-groove joint. This joint is visible on the top and bottom.

—Waist joint. In large pieces consisting of upper and low er cases, the upper section sits on a series of shallow blocks that are arranged in identical layout to those of the cornice support blocking. The backboard of the lower section projects equally in height to these applied blocks and forms a support for the upper section across the back.

The walnut desk-and-bookcase (fig. 35) is ap­parently the earliest of these case pieces. It de­scended in the Bassett family and was sold from “Clovcrlea” near 1 lanover, Virginia to the Talley family approximately seventy-five years ago. The present ow ner acquired it from them. The account book of Colonel W illiam Bassett (d. ca.1742) dated 1730-48 contains an intriguing entry that may refer to this piece:

“Cash pd Dec. 9, 1748

To Mr. Peter Scott for a desk <£5.0.0"*°

While it is virtually impossible to research this desk to firmly associate it w ith this account, there are several points in favor of relating the two. The desk descended in the same family and resided in the same house with the account book until the twentieth century. Case pieces in this group are extremely rare and few have survived compared to later work. Documentation on Scott’s furniture is very difficult to find, and the chances of correlating a document and a piece bearing family tradition is very slim. It is important when studying this piece to remember that the Bassett account book records payment for a desk—not a desk-and-bookcase. While the bookcase section of this example came from the same shop as others in the group, it was made at a later date than

the slant-top desk. The desk is finished off on top with a primary wood and is joined by a concealed, or “secret,” dovetail, indicating it w as made w ithout a bookcase. This is reinforced w ithin this group since the other desk-and-bookcase (fig. 37) has a pine top under the bookcase section and proves a different approach to the two types by the same maker. This later bookcase has a w all of Troy molding related to the type on the Blandfield press also attributed to Scott (fig. 42). The bookcase section probably dates to the 1760-70 period and w as made at that time for the earlier desk.

The w riting interior of this desk has drawers arranged on a single plane w ithout stepping or blocking, and there are gracefully cut pigeonhole brackets and a burl walnut prospect door (fig. 35a). Mitered battens are found on the top and bottom of this door and serve to stabilize it from w arpage. The match of wood grains has been made so cleverly that the joints are scarcely noticeable. The bracket feet have been heavily damaged and are partly replaced. It is difficult to determine their original shape.

Л fine elothespress made of mahogany is the only example of this wood w ithin the ease pieces (fig.

36) . It descended in the Galt family of Williamsburg and is of large size with four fixed shelves on the interior. These are made of yellow pine and are faced with a bold, mahogany astragal molding. Although they are now lost, some evidence of wear remains on the shelves from the original sliding clothes trays.

The cornice is distinguished by a w ell-designed wall of Troy molding. Likewise, the doors are well constructed with indented corners on the upper raised panels. The conforming molding that sur­rounds the panel is cut on the stiles, with the exception of the indented corners, where it is cut on two separate pieces mitered together and glued in place (fig. 36b). This piece stands on large, well - proportioned ogee-bracket feet that are double-ply, like all the ogee brackets in this group. Their composite blocking has been replaced.

A second desk-and-bookcase attributed to Peter Scott is equally well executed but has a more imposing stance (tig. 37). Its unusually short cabriole legs are mortised into large, square, inch-thick pieces of beech that are nailed and glued to the bottom of the case (tig. 37e). The ball-and-claw feet on this piece and on the desk that follows are identical to the best quality chairs in this group and leave no doubt that both chairs and case pieces come from one shop.

To balance the taller bookcase section and the outward reaching ball-and-claw feet, the cornice molding of this desk has been made larger and is cut w ith a greater projection than those of the preceding two examples. Likewise, the molding on top of the desk, which receives the bookcase section, is more vigorously shaped. The raised panels in the doors have indented corners like those seen on other case pieces in this group, and the dentil molding of the cornice (fig. 37c) has small loops between the teeth—a design found in Chippendale’s Director and seen on many pieces of English and American furniture. Their addition to the w all of Troy mold­ing, how ever, (also seen on figs. 35 and 42) appears to be an unusual feature seldom encountered in other American furniture. The interior writing cabinet is finished w ith inlayed document drawers flanking a carved prospect door. Made from crotch w alnut, this door is carved in a smooth concave depression terminating in a gothic arch, which is flanked by carved inset spandrels with a punched matt background. Originally, two drawers were behind the door.

37. Desk-and-Bookcase, attributed to the shop of Peter Scott, Williamsburg, circa 1760.

Walnut primary; poplar, yellow pine, and beech secondary. Height 9<M", width 44Zi", depth 23/2".

The Colonial Williamsburg Foundation (acc. no. 1976-95).

29. Desk-and-Bookcase, attributed to the shop of Peter Scott, Williamsburg, circa 1160.

Yellow pine primary; yellow pine and walnut secondary.

Originally painted blue-green.

Height 9m"', width 48 У»", depth 16Va”.

Mount Vernon Ladies’ Association.

The next piece has a history of ownership by the Jett family of Westmoreland and is one of the most unusual American desks (fig. 39). In its lower section are twelve vertical compartments suitable for storing ledgers, above which w ere two large drawers, now missing. The upper section has two narrow draw ers below the fall board. Above the fall board is a bank of twenty-four drawers, and above these, two book­shelves w ith eight slots for height adjustment. These are flanked by a series of five pigeonholes.

The primary wood of this desk, including the five ball-and-claw feet, is yellow pine. It was origi­nally painted a blue-green color. Though now missing, there was once a block foot in the center of the back, opposite the central ball-and-claw. It was square in cross section and, like the five pine ball-and-claw feet, was mortised into a plinth of black walnut. Their addition to the desk was in­tended to support the weight of ledgers, books, and documents.

■HI. Bookcase, attributed to the shop of Peter Scott, Williams­burg, circa 1765.

Walnut primary; yellow pine secondary.

Height 79", width 52 Vt", depth 2 5 Ун"'.

Mrs. William Nash Beverley.

I'he bookcase in figure 40 is one of a matched pair originally made for Blandficld, the home of Robert Beverley. It appears to derive from the same plan as the Galt family clothespress (fig. 36) and while it shares similar features, there are important differences. It is made of walnut and has the same type of dentil cornice, with a small loop betw een the teeth, that is found on the claw-foot desk-and - bookcase (fig. 37c). The most significant differences, however, are to be seen on the interior, where one finds, instead of a series of long shelves, a single vertical partition w ith shelves on either side. A large walnut foot approximately four inches square is double-mortised into a heavy walnut batten that is nailed and glued to the bottom of the case (lig. 40a). The presence of this extra foot, undoubtedly in­tended to bear the added w eight of bound volumes, indicates that it served as a bookcase rather than a clothespress. The survival of this pair of handsome bookcases is remarkable testimony to the eighteenth-century practice of furnishing rooms en suite.

“Clothes presses to be made by Scott 4f 61 wide High as a desk and bookcase,” wrote Thomas Jefferson sometime in 1773 or 1774. The unusual piece (fig. 42) has a case measuring 4 feet V/s inches wide, is fitted with a double-door bookcase section on its top and certainly appears to be the reality of Jefferson’s order. This example, however, was owned by Robert Beverley and still stands at Blandfield with the pair of bookcases just discussed. Jefferson’s order and the surviving press present strong evidence to support the Scott attribution and together they record the only evidence of this form in Virginia known to the writer. Neither the extreme w idth nor the combination of press and bookcase has any known parallel, although both the two-part offset form and the pierced gallery are seen on a similar example (fig. 50). That piece, a dressing chest, differs significantly in construction and bears an attribution to the shop of W illiamsburg cabinet­maker Anthony Hay.

This unusual Scott press embodies all the constructional features outlined in other case pieces of the group, but its unique form adds significantly to our knowledge of design and decoration. To accommodate the w idth of the piece, the feet were made w ider and their ogee curve is considerably stronger than the others. The uppercase has familiar doors with indented corner panels and the cornice has an elaborate wall of Troy molding. The top, however, differs from the others in several respects. It originally had a fretwork gallery and a central plinth made in tw o pieces. As the surviving evidence show's (fig. 42a), the front of the plinth was the same depth as the surviving part of the gallery, and the second piece was twice its thickness. Behind these was glued a support block. On the corners of the case the gallery is mitered, and no evidence of corner plinths can be found there.

42. Clothespress, attributed to the shop of Peter Scott, Williams­burg, circa 1765.

Walnut primary; yellow pine and walnut secondary.

Height 19", width 52'Л", depth 25W.

Mrs. William Nash Beverley.

In addition to gallery and plinth, this press has a double top. It is normal in Scott construction to glue large support blocks to the top of the uppercase and then secure the cornice to these blocks, leaving an open space behind. I lere, however, a second top is made of pine boards running from front to back, nailed above the support blocks and sitting Hush with the top of the cornice.

A border of finished walnut, parallel to the front, extends behind the fret approximately three inches and then continues (still parallel to the front) along the sides, creating cross banding there. This finished appearance in an otherwise invisible area suggests that the press was made for use in a stair hall, where it could be viewed from above.

In the lower section, tucked beneath the top and accessible only after removal of the upper drawer, are two small secret drawers. Made entirely of walnut and less than an inch in height, they are unusual features by any standard.

Some of the evidence presented by these case pieces offers a rare look beyond techniques of construction, revealing shop practices for which evidence has not often survived. One of these, seldom understood today, is the process of manufac­turing ogee bracket feet. A study of the growth ring patterns of the pine lamination backing the w alnut primary wood shows very clearly that the feet were formed from long boards. They were first glued together to form the lamination and were then planed into the ogee shape from end to end. This contoured board was then cut into separate brackets and mitered together, a method that insures that the feet are uniform in their ogee profile. In contrast, rural furniture often has feet that were shaped into ogee curves after attachment to the case and that are easily recognized by their naivete and lack of uni­formity.

Further evidence of shop practices is found in the support blocks for the crown molding on a desk-and-bookcase (fig. 35). Made of yellow pine, they have a large bead molding on the top inside corner. The size of these blocks indicates that they were scrap ends from door or window facings and suggest the possibility that Scott was making, or at least utilizing, interior architectural elements. In further support of this hypothesis are bcaded-edged boards on the backs of several case pieces. These beaded boards are included at random and are obviously not part of a decorative scheme for the piece.

Close examination of this furniture reveals a great amount of information concerning its makers and their practices. Without question, a concen­trated study of the products of other American

centers would shed further light on eighteenth - century cabinetmaking and trade practices. The homogeneous nature of chairs and case pieces attri­buted to Peter Scott is exceptional, and the quantity and scope of circumstantial evidence regarding them is equally unusual. The corpus of evidence is very strong and the attribution of these pieces to Wil­liamsburg is particularly sound. There are a few nagging circumstances, however, that necessitate a note of caution. Such is the case of two corner chairs found near Fredericksburg (figs. 28, 29), as well as a straight-legged corner chair recorded in the MKSDA files (no. S-5964), all of which have unusual arches in their crests. In this feature they differ significantly from the Palace corner chair (fig. 33), but all four share a number of other interrelated features that argue for assignment of the group to Scott’s shop (see discussion with figs. 28, 29, pp. 36-38).

Unfortunately, the extent to which Williamsburg - trained cabinetmakers migrated is unknown, although the wide distribution of related furniture indicates it was substantial. It was not unusual in eighteenth-century Virginia for apprentices to come from great distances for training, and undoubtedly they often returned home to practice their trades. These individuals must have been instrumental in carrying specific shop techniques into distant areas, thus contributing greatly to the dissemination of style. It is possible that some of the numerous chairs related to the Scott group that descended in families from the Fredericksburg area may represent the work of transplanted journeymen. The separation of pieces made in Fredericksburg by transplanted artisans from those purchased and sent to that area in the eighteenth century, however, is a task beyond the scope of this work.21 The lack of related case pieces with Fredericksburg histories is revealing, suggesting that the chairs have an origin outside the area. While there were many cabinet­makers working there, wealthy individuals living in that section of Virginia are known to have patronized Williamsburg artisans.

There are other questions regarding this group that remain to be answered. Lack of written documentation for the use of cherry, occasionally found in Scott pieces, is puzzling, although not of great significance. Its absence from Williamsburg newspaper advertisements may well be due to the prestige of owning mahogany and walnut. Many pieces made of cherry clearly fall within other Williamsburg groups. The lack of a document associating Scott with chair production is also troublesome, although it is not substantial evidence to conclude that he did not make them. Few documents regarding Scott are extant in any form, and although his newspaper ad of 1755 lists nearly every form of furniture except chairs, it ends w ith “etc.” Chairs were certainly more likely to be produced on custom order than other pieces of furniture. The variable number in a set, a tremen­dous variety of designs, and an endless degree of quality are good reasons for their being made to order rather than for direct sale from the shop.

Two of the chairs here assigned to the Scott group were tentatively attributed to the cabinet­maker Mardun V. Eventon by I lelen Comstock in 1952 and 1954.22 Miss Comstock’s attribution was based upon the existence of the chairs at Mount Vernon (fig. 27) and Shirley Plantation (fig. 18) that closely coincided geographically with Eventon’s Virginia Gazette advertisements from Dumfries and Chesterfield. The case pieces in the Scott group, however, are completely different from the signed Eventon desk (fig. 113) and thus remove him as the possible maker of the chairs.

While little is know n about Peter Scott’s early background, the large group of furniture discussed here certainly indicates a strong familiarity with urban British style and technology of the early eighteenth century. The w ording of his 1755 adver­tisement, in which he stated his intention “to go for Great Britain,” may provide a clue to his past. His use of this phrase rather than the more common “England” or “London,” suggests that he could have had a Scottish origin. James Hamilton, a Scottish carver who worked in Williamsburg and w ho used the ambiguous “Great Britain,” was one of many who preferred that designation. Whatever his background, there is no indication of a Virginia ancestry for Peter Scott, and various features of his furniture cause one to conclude that he might have been trained abroad.

The form of the ball-and-claw foot and knee carving on the earliest group of Scott chairs is quite closely related to London work of the 1730 period (see fig. 17). While an affinity with the English chair is obvious, there are significant differences. The Scott carving is a muted version, having lost a great deal of the sharpness and detail, and the carving of the knee and foot, particularly the latter, are watered-down versions of the London style. This same comparison can also be made betw een the Scott desk-and-bookcase and a London example in the Metropolitan Museum of Art (fig. 38). Again, a tremendous likeness is apparent in the squat cabriole legs, flattened ball, and extended claws of the feet. Virtually identical construction was followed in forming the back foot on both pieces, although again the Scott example is a muted version that possesses less sculptural detail. Another feature seen on Lon­

don pieces and consistent in Scott’s work is the ogee bracket foot laminated to a secondary wood. A fine example is seen on a London china cabinet that also has composite or built-up foot blocks which are typical of this and other Williamsburg groups (fig. 41a). The method of securing raised panels in doors, with a small nailed-in molding, is another early Georgian feature seen on English cabinetwork of the period, and is identical to pieces attributed to Giles Grendey of London.

Other English construction can also be seen in chairs of this group, w ith their integral rear seat rail and shoe, knee blocks that overlap the seat rails, and rather stiff cabriole legs—all of w hich are seen on English pieces of the 1720s and 1730s. The use of beech and oak for slip seats and as secondary woods in upholstered furniture is also a continuation of English cabinetmaking tradition that was practiced by Scott and others in W illiamsburg. Little solid evidence of such an extensive use of beech for upholstered furniture can be found elsewhere in America.

I he conclusion that these pieces of furniture w ere made by a tradesman trained in a British urban center, possibly even London, is somewhat startling considering their provincial character. Yet this group unquestionably exhibits an understanding and familiarity with the most advanced urban cabinetmaking technology of its time, while falling short of urban style. No other shop group in America is known to the w riter in which such close affinity w ith English technology is present.

While the presence of urban technology com­bined w ith provincial style may seem contradictory, circumstances did exist in the transition from a large center to a small one that could have produced these results. Urban furniture-making was made up of a series of specialized trades consisting primarily of cabinetmaking, chair-making, carving and gilding, and upholstering. Technical evidence presented by this group suggests that Scott was trained as a cabinetmaker or chairmaker and that he had expo­sure to both. Working through an apprenticeship in a large city shop would have taught him construction and exposed him to high-style carving, although it did not necessarily teach him the techniques of carving and designing. If a person with such a background then migrated to a small city like Williamsburg, w here it is doubtful that he could depend upon finding specialists, he would have been forced to handle these skills himself.

While sueh a tradesman might not have had sufficient skill or business ability to achieve promi­nence in a large city, his success in Williamsburg would have been understandable had he arrived around 1720 w ith a sound know ledge of current fashion. Such an introduction, when styles in America often embodied turn-of-the-century fea­tures, could have met with extreme success. In Virginia, where the leaders of society were continu­ally seeking the latest but not always conveniently available English styles, it is easy to hypothesize why Scott w as so popular.

Twenty-two pieces w ithin the Scott group have been examined in this study. While all the known cases pieces are included, numerous chairs, mostly very plain ones, are omitted. Some of these simpler chairs display different proportions, designs, and quality, indicating that they represent the formation of a regional style. This is easily understood since the group has an early origin and its production spanned five decades. The number of journeymen and ap­prentices working over this time span must have been great, and w hen one considers that these men traveled in all directions and trained others, the development of a strong regional style is understand­able.

As stated earlier, the consistency in this group is remarkable. The quality and design in the early examples are outstanding, although in later pieces the workmanship declined and attempts to change and adapt to the rococo style were generally not as successful. When looking at the Scott group w ith a critical eye, most collectors and furniture historians will attempt to compare their baroque forms with later, more rococo examples. The early Scott chair certainly is nothing like a Philadelphia or London chair of 1770—nor should one expect it to be. In evaluating the group, one must consider that its baroque George II style was founded upon a com­pletely different set of aesthetic values than those of the later rococo taste.

If the dates assigned by this study are correct, the earlier pieces are pioneer examples for their time and are certainly among the purest examples of the George II style produced in America. T his group shows its origin to be British urban cabinetmaking that, transplanted to colonial Virginia, developed into a provincial city style unique in American furniture.

1. Information courtesy of Harold B. Gill, Jr., historian at Colonial Williamsburg. Front his forthcoming book Arts and Crafts in Virginia.

2. Tbt Virginia Gazette, cd. William I luntcr. September 12. 1755. p. }.

3. Ivor Noel Hume, Williamsburg Cabinetmakers: The Archaeological Evidence (Williamsburg: Colonial W illiamsburg Foundation. 1971). pp. 15-19; Virginia Gazette, cd. Alexander Purdie. June 26, 1776, p. 3.

4. The Virginia Gazette, cd. John Dixon and William Hunter. De­cember 2, 1775, p. 3.

5. The Virginia Gazette, cd. Alexander Purdie, January 5, 1776, supplement p. I.

6. Ibid.. January 26, 1776, p. 3.

7. Hume, Williamsburg Cabinetmakers, p. 14.

8. John Custis Papers, 1711-1764, Virginia Historical Society, Richmond. Virginia. Information courtesy of Harold 13. Gill, Jr.

9. The Virginia Gazette, ed. Alexander Purdie and lohn Dixon, Nnril 18, 1776. p. 3.

10. All Jefferson accounts courtesy of Charles Granquist, assistant director, Thomas Jefferson Memorial Foundation, .Montieello, Virginia.

11. Col. William Bassett Manuscripts Account Book (1730-1748), Virginia Historical Society, Richmond, Virginia. Information courtesy of I larold 13. Gill, Jr. (See page 42 for full quote.)

12. John Mercer Manuscripts Fedger (1725-1750) Bucks County Historical Society, Doylcstown, Pennsylvania. Information courtesy of Harold 13. Gill, Jr. '

13. Robert Carter Manuscripts Day Book, Manuscripts Division, Duke University Library, Durham, North Carolina, Vol. 14 (1776­1778), p. 146.

14. Robert Carter Manuscripts Letter Book. Manuscript Division, Duke University Library, Durham, North Carolina, Vol. 6(1784-1785) pp. 134-135.

15. Information courtesy of Charles Granquist.

16. Ibid.

17. Ibid.

18. Letter from John FirktoGraham Hood. April 24, 1972, Accession File No. 1972-230, Department of Collections, Colonial W illiamsburg Foundation, W illiamsburg, Virginia.

19. Report of the Curator 1971 (Montieello, Virginia: Thomas Jefferson Memorial Foundation, 1975) p. II.

20. Bassett Account Hook.

21. Sec Ann W. Dibble, "Fredericksburg-Falmouth Chairs in the Chippendale Style, ' Journal of Early Southern Decorative Arts 5 (May 1 *>78): 1-24. In this article, Ms. Dibble attributes many of the chairs illustrated, as well as others that share similar features, to the Freder­icksburg area. All of the chairs covered by her article are clearly within the Scott group designation used in this work. Ms. Dibble further divides these examples into three sub-groups. Her attribution to the Fredericksburg-Falmouth area is based on the fact that numerous chairs survive with histories there. While it is entirely possible, if not probable, that some production of these chairs occurred in that area, evidence is very strong for the origin of this group in Williamsburg. Several important pieces not covered by Ms. Dibble's article include examples from three sets of chairs with Williamsburg histories (figs. 22, 30, 32, 34, 58, and 62), and all six of the case pieces (figs. 35, 36, 37, 39, 40, and 42).

I he case pieces are firmly linked to the chairs through their use of

identical ball-and-claw feet and the inclusion of beech as a secondary wood—features that are otherwise rare in eastern Virginia. If these provenances are added to Ms. Dibble's map, they create an entirely different picture. They also add important provenances, many of them distant from Fredericksburg. In addition, some have document correla­tions w ith Williamsburg, and specifically w ith the shop of Peter Scott.

A surprising number of chairs in this group have association with the Washington and Custis families. Six are know n: one is shown in Ms. Dibble's article, three are cited in her footnotes, and two others are discussed here (figs. 21, 27). These examples have intriguing connota­tions w hen it is realized that Peter Scott rented both his shop and his home from the Custis family and their heirs for forty-three years. When George Washington married Martha Dandridge Custis in 1759, he took over management of the property, and Scott continued to rent it from him for the follow ing sixteen years. While not conclusive, the longstand­ing association the cabinetmaker had with these two families could explain their ow nership of so many related chairs. This interpretation gains further credence if considered in light of the evidence that associates the group with Scott.

Ms. Dibble uses techniques of pegging as the criteria for dividing these chairs into three groups. This raises some important questions. Inconsistencies of pegging have caused this author to conclude that much, if not all, of the evidence has resulted from repairs that occurred over the years. It was a standard practice to tighten loose joints by clamping and pegging—thus avoiding the necessity of completely dismantling a chair in order to glue it properly. Inconsistencies occur not only from chair to chair but also within a single example, where some joints are pegged and others are not. Occasionally, in more elaborate pieces they are crudely placed in the midst of the knee carving (see tigs. 20a. 28a). Other pegs almost miss the tenons entirely.

Vertically grained, two-part support blocking is also used by Ms. Dibble to segregate one of these groups, gain, there are discrepancies, since it is probable that most chairs originally had blocking, though many have lost it. I low ever, there arc examples w ith this blocking that fall w ithin her other groups, w hich by her designation, should not have them, and thereby prove division by this feature unreliable. These include the Benjamin Waller chair (fig. 62). the \ ashington chair (fig. 27), the Virginia I listorial Society chair(fig. 19), and the Lodge IV chair (fig. 59c). In addition, the horizontally grained poplar blocks of the Shirley chairs (fig. 18) cited by Ms. Dibble, are w ithout question old replacements and are crude in execution when compared with others in the group. I his conclusion is reinforced by the original vertical-grained cherry blocking of the matching Virginia Historical Society chairs. These double vertical blocks are cut w ith a slightly convex surface on the interior, giving them a triangular cross section that conforms to many cited here. Unfortunately, it appears that the support blocking of the entire group has suffered loss at a rate comparable to the knee brackets on the cabriole leg chairs. While it may seem unusual that such a large percentage of the original vertically grained blocking has been lost, evidence for knee bracket blocking used tin cabriole leg chairs is even more fragmentary. An almost intact example can be seen on the Lodge IN' chair (fig. 591"), w here large horizontally grained blocks back up the knee bracket beneath the seat rail. I lere the support oil the right is intact w hile that behind the left knee bracket is missing. One side chair in the Scott group (tig. 20) has a very small splinter from its original yellow pine horizontal block, and w as identical in construction to those of figure 59f. This strongly suggests that all were originally made this way, even though all have been lost except the example cited here. (These horizontally grained blocks form a parallel to the principle of the composite foot found on case pieces attributed to Scott, where horizon­tally grained blocks are also used (fig. 40b). These unusual knee blocks, together w ith the vertically grained support blocks, form another strong tie between the chairs in this group and point to their common origin. Sub-groups w ithin the larger group do exist, as Ms. Dibble has noted, yet for the reasons pointed out in the foregoing discussion, her attempts to separate them are not valid. One of the sub-groups includes the laalgc IV chair. The relationship of this chair to the Scott group is more fully examined under the Edmund Dickinson section of the chapter that deals with the Anthony I lay Shop.

One further argument used by Ms. Dibble to tie these chairs to the Fredericksburg area is found on an armchair that descended in the Green family there (see Dibble footnote no. 17). This chair has construction details found throughout the group: the integral rear rail and shoe w ith horizontal shaping; parallel stiles; serpentine arm supports; and arms dovetailed into the sides (Ms. Dibble describes them as notched). The basis of her argument relies upon the similarity of its splat to those found on earlier style chairs from the northern Tidewater area. I lowever, the splat on the Green chair is a replacement. Removed from a taller piece, its addition to the Green family example necessitated reshaping of the shoe and the underside of the crest rail. The added splat is the only tic between these groups, and consequently there appears to be no valid basis for relating the two.

A F'rcdcricksburg association for one chair of this group is presented by Conover I lunt-Jones in Ihlley and "the great little Madison" (Washing­ton. I). C.: American Institute of Architects Foundation, 1977), p. 96, fig. 96. Ms. I lunt-Jones states that this chair, together with several others surviving from the same set. was purchased in 1773 bv James Madison from an unidentified cabinetmaker in the Fredericksburg area. lTiis cabinetmaker's account book, w hich records the transaction, is in the Joseph Downs Manuscript Library, Henry Francis duPont Winterthur Museum, Winterthur. Delaw are. If the chair illustrated by Ms. I lunt - Jones is in fact part of the set noted in the account Імюк, it w ill mark an important beginning in isolating Fredericksburg area production of Scott types. There is, however, some question in identifying this chair as pan of the set and Ms. Hum-Jones does not discuss how this was determined. She shows another chair having Scott group characteristics with the same family tradition. Still another very different example in the Colonial Williamsburg Collection (acc. no. L1977-307) has an inscription indicating that it, too, belonged to James Madison. Admittedly the rcfcrcnce in the account (шок is to walnut, and it is true that the second Scott example shown by Ms. I lunt-Jones is cherry, while the Wil­liamsburg example is mahogany. Yet the crucial point that remains

unaddrcsscd is the validity of matching a specific example to the account without further support. Traditions often have accuracy problems and the Madison family obviously owned a large quantity of furniture in the eighteenth century. The inexpensive price of the set is also disturbing since the Madison family chair has a splat with piercing. The account book entry to Col. James Madison notes that on Sept. 2, 1773 the cabinetmaker “finsht 12 .Marlborgh chairs walnut £6-6-o,” meaning that each chair cost 10Fi shillings each. This low price is particularly disturbing if compared to the cost of the Windsor chairs at 15 shillings each, that were purchased by Frances Jerdone of Yorktow n from a carpenter in 176V. The price and materials of other forms recorded in the account book confirm a level of modest production. The lack of mahogany also indicates inexpensive production, and makes any close association with the finest carved Scott chairs very unlikely.

In reference to this account hook, Ms. Dibble cites an entry for a "pillar and claw" table as evidence that this shop was producing examples bearing ball-and-claw feet ("Frcdcricksburg-Falnmuth ('hairs in the Chippendale Style," see footnote 3). I low ever, pillar and claw does not refer to ball-and-claw feet—a fact that is clearly shown on page 38 of the laindon Society of Upholsterer's Genteel HousHold Furniture. Three tables having tripod bases are illustrated there in a plate entitled “claw tables," and all of them have scroll feet. This is firm evidence that the term “claw," when applied to this form, referred to its three legs and not to ball-and-claw feet.

22. Helen Comstock, "Discoveries in Southern Antiques," The Magazine Antiques 65 (February 1954): 131-133.