Initially the idea behind reconstructing the barn was to use it as an interpretive space, but not as a working replica. However, as additional research was conducted, it became clear that there was an unusual wealth of information about the barn and the associated outbuildings. Such details could enrich and deepen our understanding of George Washington the farmer, designer, problem solver, and businessman. The decision was made to expand our goal considerably and to reconstruct the treading barn and barnyard as authentically as possible in order to make them a working complex just as Washington had envisioned them and just as the originals were actually utilized.
The original structures were built at Mount Vernon between 1792 and 1795 and are probably among the nation's best documented agricultural buildings of any period. Sources consulted included Washington's correspondence, various plantation records, the building specifications and bill of materials prepared by Washington in 1792, and a photograph from a glass plate negative of the barn that was taken during its last years of existence in the 1870s. Details of craftsmanship not covered by the records were unraveled by modern-day sleuthing. Archaeological work was carried out at the site of the original barn where bricks and nails were uncovered. Our architects, Michael Quinn and Gordon Bingaman, viewed the photograph of the barn on a CAD (computer assisted design) system in order to understand fine points of timber framing and scale. The Mount Vernon crew and consultants could occasionally be found crawling through the rafters of early homes in Alexandria, as well as historic structures at Mount Vernon, examining period materials and joints. Together these methods and sources provided the basis for one of the most meticulously reconstructed 18th-century buildings anywhere.
By the fall of 1792, George Washington had fully developed his plans for the barn complex at Dogue Run, one of four working farms on his 8,000-acre estate. The new barn was to perform the same function as the English-style barns Washington had been studying for many years : grain processing and storage - but in a radically different manner and with several novel features. Although there is no documentation on the process of how Washington actually designed the barn, he was well acquainted with using animals to separate grain from stalk by treading, and he had very specific goals in mind that he wanted to achieve. Beginning with a simple concept - to construct a circular wooden treading floor large enough to accommodate horses and enclosed from the weather - Washington sought to improve the efficiency of this basic treading process and simultaneously reduce the opportunities for theft. The treading floor was to be on the second level of a two-story structure, which the horses could access via an earthen ramp. He conceived the innovation of leaving spaces between the floorboards so that the heads of grain, once separated from the straw, could fall through to a granary below. There they could be temporarily stored in a central octagonal structure, then winnowed and sent to the mill. The result was a building that conceptually is as much a machine as architecture.
Polygonal buildings were unusual but not rare in 18th-century Anglo-America. Therefore, although it is the structure's most distinctive feature, it is not the polygonal shape of the barn that is Washington's innovation. Rather, it is taking an agricultural process and designing a structure to accommodate it. The 16-sided shape and 52-foot diameter probably were arrived at through some system of balancing functional needs with structural concerns. Certainly, the roughly circular footprint was chosen to facilitate the treading of the horses. The reason for the barn being polygonal instead of perfectly circular in shape is undoubtedly due to the greater ease of constructing straight sides instead of curved walls. The selection of 16 sides rather than any other even number is more difficult to explain, although it is clearly related to the octagonal shape of the interior framing system.
The bill of materials George Washington prepared in 1792 and sent to Anthony Whiting, his estate manager, serves as the primary source of information on the wood used. In the list, Washington specifies that certain pieces were to be purchased from Alexandria merchants, while others were to be obtained by Mount Vernon workmen from the plantation. Since the treading floor was to withstand the weight and punishment of trotting horses, durable white oak was his choice. Most of the remainder of the wood for the barn, except the cypress roof and the ground floor interior octagon posts, was of pine. Washington omitted the ground floor posts from the original list, possibly because he was contemplating using brick piers instead. The wood for the original was acquired from Mount Vernon, and because it was procured locally, we felt justified in using oak.
Less is known about the types of wood used in the stables and corn houses. In his instructions to estate manager William Pearce in January 1795, Washington specifies that the posts running along the open side of the sheds were to be of locust. Locust is extremely durable and rot resistant, and since these posts were set into the ground, this was a very good choice. In an earlier letter, Washington specifies cypress for the shingles. Because the stables and corn houses are much more open to the elements than the enclosed body of the barn, more resistant oak was used in the reconstruction, except for siding, shingle lath and doors, which were fashioned from pine, and the cypress shingles. Washington acquired most of the wood for the stables and corn houses from Mount Vernon.
By the 1790s, water-powered sawmills were well established in Virginia. But based on extant structural evidence from Mount Vernon and Alexandria, along with data provided by the plantation accounts, the great majority of wood used at Mount Vernon continued to be prepared by a combination of hand hewing and pit sawing. Since one of our goals behind the reconstruction was to have the buildings look as much as possible like they did in Washington's day, all of the wood surfaces were pit sawn, hand hewn or planed. To speed the process and to cut costs, we cheated wherever we could by cutting logs down to approximate size with a saw mill before making the final cuts with hand tools.
The hand-made brick was made by a company in North Carolina according to our specifications. The color range of the bricks and their size were determined from a combination of the archaeological findings at the original barn site and evidence provided by standing structures at Mount Vernon. Because the Mount Vernon stable has full masonry walls, was constructed only a little more than a decade before the treading barn, and is relatively well preserved, it was selected as the single most important prototype for the appearance of the treading barn masonry. The brick bond used, English bond (alternating rows of stretchers and headers), is clearly visible in the 19th-century photograph of the barn and was used in the replica. Washington expressed concern for neat masonry, and since the photograph indicates the use of other decorative detail, particularly beaded or rounded elements in the molding, we used a hand-ruled joint to provide a more finished appearance to the work. The ruling consists of a line inscribed down the center of the joint.
Although it appears steep, the slope of the barn roof (approximately 43°) is not unusual for 18th-century buildings. Roof pitches between 42° and 48° are typical for this period, and the Mount Vernon stable roof is 44°. The general reason for such a steep pitch is that it encourages faster run-off, resulting in less time for water to penetrate the wooden shingles. Presumably, the flatter slope of the shed roofs was acceptable because of their very narrow width, in combination with the function of the buildings.
Because the barn has an unusual shape, we tend to believe that it was a difficult building for Thomas Green, Washington's carpenter, and the Mount Vernon slave carpenters to construct. Certainly, it was a formidable challenge for our modern-day crew. The most common problem we found was making the calculations for cutting the members and joints at other than 45° angles. In addition, since the angles at which the pieces joined always varied along any given structural unit (a wall section, a floor or roof facet), it was necessary to figure a range of angles so that appropriate framing would match up.
One of the most important decisions Washington made was to design a building comprised of two nested polygons. This refers to the 16-sided outside wall, enclosing the interior octagonal framing system. This design was the most straightforward way to support the roof since the same system for laying out the rafters could be continued from the eaves to the peak of the roof. The transition from the exterior 16 sides to the interior eight means that the structure of the roof had to undergo some subtle adjustments as construction continued upward, including the reduction of the number of rafters from 80 to 48. Since the rafters are the only structural members spanning that space, they serve the crucial function of binding the exterior wall to the frame. By using two sets of rafters, it was possible to reduce the number of the upper set because of the upwardly constricting space.
The pattern of attaching the shingles on the barn also turned out to be an unexpected problem. Since the angles at the 16 corners are so great, the hip that was formed at the junctures was very subtle. It proved impossible to lay the shingles in the normal manner (with alternating superimposed rows), and considerable experimentation was required to determine that butting the shingles was the most viable solution. Once again, constant reference to the circa 1870s photograph was made, and, although it was impossible to determine the exact nature of the actual pattern, our solution appears to conform to that evidence.
Virtually all of these issues had to be deduced during the construction process. While it was conceivable that all aspects of design could be determined at the drafting table, in practice it was impossible to visualize all of the variables involved. In essence, we may have done what George Washington did: made our best effort at puzzling out the problem, confident that some solutions to the troubling details would eventually present themselves, and trusted the skill and craftsmanship of the workmen on site to bail us out.
Finally, how did we do in comparison to Thomas Green and his men? Our crew undoubtedly was better educated than Washington's and had the benefit of both modern tools and measuring devices, not the least important of which were pocket calculators. But, Green and his slave carpenters had in their favor decades of experience in all types of heavy timber construction. It took Green one year to build the barn and probably another six or eight months to complete the other three structures (one of the corn houses was already in existence when the project began in 1793). John O'Rourke, our master restoration carpenter, and his men, together with our masons and other contractors, and with help from our summer interns, reconstructed all five structures in about 18 months - not bad.