The following article describes one the most exciting periods in the stewardship of a historic home, the unraveling of the mysteries that time and change bind up in its fabric. Montpelier has been fortunate to have the consummate professionalism of Mark Wenger as chief sleuth for this job, seconded by Myron Stachiw. Mark and Myron’s unyielding drive for excellence, their meticulous approach to research and physical investigation, and their experience and inspired speculation have led our research team to conduct an investigation that stands as a model for stewards of all historic homes, whether public or private. The Montpelier Foundation is grateful that this project was so generously supported by the estate of Mr. Paul Mellon. And we are very pleased that the excellence of the Montpelier investigation has been recognized by the Vernacular Architecture Forum with its prestigious Paul E. Buchanan award.
This type of investigation is the most critical step in planning for the future of a historic home. No proper decision about a historic building can be made in the absence of a fact-based understanding of its construction history.
Montpelier is already benefiting from the investigation. The knowledge it provides has reshaped our interpretation for visitors and altered some of the routine jobs of maintenance and repair. Most importantly, the investigation has given confidence to our long-range planning, as the fact that the Madison home survives and can be authentically restored enables the foundation and the National Trust to tackle the task of envisioning Montpelier’s future.
Michael C. Quinn, president, The Montpelier Foundation
On the morning of September 30, 2002, a weary band of sleuths clambered into the back of a pickup truck, heading for the office of Montpelier Foundation president Michael Quinn. They had just concluded an 11-month survey of James and Dolley Madison’s celebrated house in Virginia, funded through a grant from the estate of Paul Mellon. The mood in the truck was festive, for now the fruits of that study -- more than 2,500 pages in 15 bound volumes -- were to be deposited on Mike Quinn’s sofa.
The declared goal of this report fully warranted its bulk and cost. Our charge was to characterize James Madison’s house and to assess whether the information was sufficient to support its full restoration. To conduct the necessary investigations, the Montpelier Foundation enlisted the assistance of architectural historians at Colonial Williamsburg. In November of 2001 staff members from both organizations began the enormous task of detailing Montpelier’s physical history. On the Colonial Williamsburg side, consultant Myron Stachiw and CW historian Mark R. Wenger co-directed the probe, while Montpelier staffers Maya Berrera, Felicity Blundon, Jeff Carey, Misty Eppard, Todd Gordon, John Jeanes, Alfredo Maul, and Maggie Wilson completed the survey team.
The investigation undertaken by the Colonial Williamsburg Foundation was the fourth major effort to trace the story of Montpelier’s evolution and continued the Foundation’s episodic investigations conducted between 1997 and 2001. The findings of these earlier scholars greatly informed our investigations and were incorporated into the final interpretation. These included research and an exhibition by Conover Hunt-Jones in 1977 that identified the three major phases of construction and developed plans of those phases by Frederick D. Nichols and Nicholas A. Pappas; a survey of the building’s fabric by Winston-Salem architects Charles Phillips and Joe Opperman, together with Colonial Williamsburg’s retired director of architectural research, Paul Buchanan; and a historic structure report (1990) by historian Ann L. Miller that compiled documentary evidence regarding Montpelier, followed by physical investigations of the building between 1990 and 1993 by Miller and Larry Dermody, an archeologist with the National Trust for Historic Preservation.
The most recent investigations involved methods that have come to be known as “building archeology.” While traditional archeological excavations occur primarily in the ground, building archeology treats the entire building -- floor, walls, and ceilings -- as an archeological site. Principles and techniques of investigation and documentation of in-ground archeology are thus used in building archeology, but are adapted to treat layers of plaster, paint, and wallpaper as archeological layers. Traces of the removal or addition of building fabric are treated as archeological events to be documented and interpreted within the chronology of the structure’s history of use and occupation. The chronology of changing building technology was revealed in such things as nails and the marks left by wood-processing tools, which provided important dating markers for the researchers. To process and integrate this very large body of information, the research team created an innovative system of computer-based recording, which included drawings, digital images, and text.
By May 2002 the research team had completed 228 subsurface “excavations,” documenting each with photographs and narratives. Another 67 probes opened during previous studies were “adopted” and documented as well. Between May and September the results of these excavations were collated and the final report was written, edited, printed, and bound.
Because of the limited goals of the project and the challenging schedule for its completion, our study focused on a 27-year period between 1809 and 1836. Post-Madison alterations (including extensive ones by William duPont and later his daughter Marion duPont Scott, the owners from 1901 to 1983) were considered only to the degree that was necessary to disentangle them from the other, earlier periods of construction. Thus, the resulting report cannot purport to be a complete history. Yet it has answered many important questions about the appearance and character of Montpelier during the years before and after James Madison’s ownership.
Although the report focuses on James Madison’s era, a brief summary of the history of the house will help put the research efforts in context. The earliest part of Montpelier was a simple eight-room, brick house built around 1760. Over the years it has gone through many changes. In 1797 a four-room addition was added. Between 1809 and 1812 (the time period that was the main focus of research), James Madison renovated the house and added one-story wings at each end. Dolley Madison sold the house in 1844 and subsequent owners made numerous additions and renovations to it. The property was purchased in 1901 by William duPont, Sr., and over the next couple of years, the home was nearly doubled in size. Thus researchers had the challenge of untangling 250 years of additions and alterations.
Summary Findings on the Exterior
Like the studies that preceded it, this investigation produced numerous exciting discoveries. Many aspects of Montpelier’s external appearance now come into sharper focus. The roof coverings employed on all portions of the house are now known -- that of the main house and portico consisted of round-butt wooden shingles, judging from examples recovered in the attic. Most gratifying, however, was new information concerning the “flat” roofs that originally covered the north and south wings. Previously known through graphic (paintings and drawings) and documentary evidence, the exact character of these roofs had long been a matter of speculation. Now physical evidence has revealed their precise height and slope, the thickness of their sheetmetal coverings, the height and slope of the decking applied over them, and the profile of the cornice that dressed their edges -- despite the fact that the original roofs have long since been replaced. Equally important, it was determined that these roofs remained in place throughout Madison’s lifetime.
The character of Montpelier’s early windows and doorways has also been clarified. Virtually all Madison-era doors and windows have been located precisely and their place in Montpelier’s architectural chronology established, including a number of long-vanished openings.
Heretofore unknown was James Madison’s reconstruction reconstruction of virtually all windows in the main house between 1809 and 1812 -- a step that affected the mansion’s appearance substantially. The intention behind this costly undertaking was to give uniformity to windows and doors from three different periods. Under the front portico, for example, Madison diminished the widths of two doorways flanking the main entry so that they aligned with newly proportioned and relocated windows above. Graphic and documentary evidence shows that new doors and frames were completed for these openings in the 1809-12 renovation and that they remained for many years after Madison’s death (the present, triple-hung windows date to c.1860).
As for Montpelier’s windows and the opulent doorway below it, all doubts concerning their date have been removed. The collation of physical and documentary evidence shows that they were part of the 1809-12 remodeling, the 1802 Thornton watercolor notwithstanding. (Thornton, an amateur architect involved with the design of the U.S. Capitol, is believed to have prepared the sketch during a visit to Montpelier. The image was not so much a portrait of the house as a proposal for its alteration, containing several elements that are known to post-date Thornton’s stay at Montpelier.) In addition, the Period I front windows that preceded these openings have been located and their positions precisely recorded.
By 1811 most of the windows were protected by louvered or “Venetian” shutters. Analysis of the 1809-12 building accounts revealed that builders James Dinsmore and John Neilson built shutters for nearly all first- and second-floor windows except those shaded by the front portico. Of the doors, only the “Venetian doorway” above the colonnade was afforded this protection. Physical and graphic evidence reveals that these shutters were not hung in double tiers as at Monticello. Rather, a single, louvered blind hung on each side of every window.
Judging from an unaltered example under the portico, the original windows were much smaller than those we see today. In the cellar of the main house these windows had no provision for a sash -- they were open grills. Because the cellars of the wings were heated, habitable spaces, their lunette windows were fitted with an operable glazed sash.
Still on the front of the house, investigations in the attic have shown that the lunette window on the front of the portico is not original. It seems to have been added between 1818 and 1836 when the earliest dated view to depict the window was published. Investigations of the exterior may further narrow the period in which it was installed.
At the base of the portico, excavations located remains of a base for giant Tuscan columns. (Originally, these bases stood just above the deck of the portico.) Fixing the height of this point has allowed the architects to locate the bottom of the original column base, thus fixing the original height of the porch deck. Viewed in concert with the building accounts, this dimension yields the number of risers from the deck to the ground and so allows us to approximate the elevation for exterior grade during Madison’s lifetime.
On the rear facade, the Madison-era colonnade remains essentially intact, retaining most of its original framing on both levels -- a breath-taking survival given the highly vulnerable location of this appendage. The doorways and windows opening onto the lower level and a “Venetian doorway” opening onto the upper deck were all confirmed as early features (though the present exterior trim of the upper doorway is modern).
Above and to the north of the Venetian doorway two dormers -- previously unassigned in the Montpelier chronology -- may date as early as c.1860, judging from their construction and from the evidence offered by an early photograph. Perhaps the most dramatic discovery on the exterior was that concerning the stucco finish of the dwelling -- it postdates James Madison’s lifetime. Physical evidence ties it to a mid-19th-century building campaign that included the closing of many windows and doorways, revision of the portico, rehabilitation of the cellar, construction of new hipped roofs on the wings, and covering of the entire house with a standing-seam metal roof. To a degree we had not fully appreciated, Thomas and Frank Carson of Baltimore transformed Montpelier’s outward aspect around 1860.
Archeology has been a nearly continuous source of revelation about the larger setting in which James Madison’s house stood -- lawns, gardens, outbuildings, work areas, fencing, paths, ravines -- even a moat! (actually an areaway that ran across the rear of the house). This part of the investigation was planned and conducted by Montpelier’s director of archeology, Dr. Matthew Reeves. Based on an intensive survey in the vicinity of the dwelling, coupled with carefully targeted excavations and the discovery of an early insurance plat, the archeology team has begun to explain where things were, how the site evolved, and how these changes affected the dwelling. Inside the house, excavations in the cellar uncovered important new evidence concerning that neglected space and its relationship to the rooms above.
Investigations were equally fruitful in other areas of the house. The disposition of all rooms and the circulation between them has been determined -- virtually every doorway, window, wall, chimney, and hearth has been located and its original size worked out. The need for precise dimensional information in advance of potential restoration drove our investigation and set it apart from earlier studies. Where early features had disappeared, it was important to establish locations, widths, and heights as exactly as the evidence allowed.
Where physical evidence failed us, James Dinsmore’s remarkable building accounts, based on his own careful measurement of the work, often provided the answers we sought. Researchers analyzed these records, comparing the quantities of moldings and other components called for in the accounts with those actually present. In James Madison’s chamber (M-104 now used as a stair hall), this comparison revealed that the embrasure of a vanished rear window had extended to the floor, rather than stopping at a window seat. In the same space, Dinsmore’s measurements showed that the window beside the chimney stopped at the sill, and that it was only two lights wide -- an arrangement that correlated with astonishing accuracy to the available space between the chimney and the corner of the room. In the room directly above, it became clear that the corresponding second- floor window was three lights wide -- a consequence of the narrower chimney mass on the upper floor.
Sometimes graphic evidence allowed us to interpret the documents correctly. Near the end of the project, an early photo of Madison’s chamber, which was heretofore unknown to us, showed a doorway that we had missed, and thus accounted for surplus door trim in the Dinsmore account -- a problem that had tormented us for months.
Two major stairways, formerly black holes in our understanding of the house, can now be reconstructed spatially, and some aspects of their finish determined. The tread and riser of the Period I stair (M-110) have been fixed within a narrow margin, and the character of its finishes (exclusive of the newel and rails at the second floor) have been worked out. Evidence for wainscoting at the bottom of this stair and for a skirt around the second floor aperture has been exposed and recorded. It now seems that the stair ascended from a raised landing between the south passage (M- 109) and the chamber (M- 111) and that this landing was accessible from both rooms. In addition, what we believe to be the Period I frames for those doorways have been identified -- both having been reused in the cellar.
On James and Dolley’s end of the house, evidence for the north stair is equally remarkable (M-106). From the cellar one can see a remnant of the lower newel post, showing the width of the lower flight and fixing the point where its ascent began. Over the rear doorway is clear evidence for the stair landing, which spanned the entire width of the north passage. On the upper floor, the truncated remnant of an upper newel established the size of the stairwell and the alignment of the railings, while a plugged opening in the south wall revealed their height. The removal of duPont-era flooring in this area supplied the last pieces of the puzzle -- the configuration of the second-floor landing and the method of supporting it.
Undetected in earlier studies was a bank of closets contained between the passage and the large front bedchamber. On the surface of the flooring was evidence for the extent of the closets, for a door to the passage, and for the direction of its swing. The removal of plaster from the masonry front wall revealed that the closet had been whitewashed rather than plastered. Reused studs bearing much whitewash are still trapped inside a late-19th-century partition nearby. When liberated, these will provide additional information. Among the most surprising and pleasing discoveries was the extent to which James Madison’s Montpelier is still present -- more of its early fabric survives than any of us had supposed. At least 37 of 52 Madison-era doors survive, more than half of which found their way to the duPont additions in 1901-2. Through the study of measurements, moldings, and hardware evidence, it has been possible to put most of the errant doors “back in their holes.”
As in the case of the doors, more of the 1809-11 window frames survive than anyone suspected. Some had been relocated in the duPont wings during the 1901-2 alterations. Others were cut down and reused in a mid-19th-century rehabilitation of the cellar.
Upstairs, researchers spotted reused soffits and keystones from the arched doorways that once adorned the first-floor vestibule by James Madison’s front door (M-107). Downstairs in the drawing room (M- 108), researchers found still more components from these arched doorways, including the original imposts, paneled jambs, and plinths. These exciting discoveries tie the Madison doorways to other examples of James Dinsmore’s work and thus provide a basis for detailing the few unrecoverable elements.
Elsewhere, Madison-era finishes have been more elusive, yet persistent pursuit of these items has yielded some important successes. Physical evidence for surbases or chair rails has now been located in virtually all of the dwelling’s major rooms. At the same time, the early date of the existing drawing room (M- 108) cornice was verified, while the existence of such trim was ruled out in Dolley’s chamber, and in most other spaces as well.
Just as James Dinsmore’s accounts were useful in divining the size and configuration of windows, they offered an important guide to the character of other finishes. In the wings we learned that some doors were adorned with full architraves while others received only a backband molding. In the drawing room, James Dinsmore’s mention of “1 Pediment over Door” led to the spectacular discovery of an overdoor treatment for the Madison opening that once served the south passage (M- 109). The picture of this vanished doorway treatment was completed when we realized that the original door and the paneled jambs for this opening now stand in Marion duPont’s Art Deco “Red Room” (M- 118).
In the south passage (M- 109) the existing wainscot was definitively associated with that space and certain sections of this paneling were linked to specific locations. This wainscot was also shown to have existed in the rear end of the passage and its relationship to the rear door embrasure was proven. Inarguable evidence for similar wainscoting was uncovered in the dining room (M-112) and in the Period I stair enclosure (M-110) as well.
The survival of original flooring is extensive, especially upstairs, where it provided important evidence of change. The survival of numerous Madison hearths and fireplaces was equally surprising. Early chimneypieces are now associated with several of these fireplaces and some appear to stand in their original locations. Where chimneypieces have disappeared, remaining finishes or remnants provided remarkable evidence for the vanished features, including indications of width, height, and even molding profiles. Such evidence is available for Dolley’s chamber (M-100), the dining room (M- 105), and the drawing room (M-108).
Paint data revealed that the woodwork, inside and out, was generally painted an off-white color, with gray baseboards inside. Contrary to expectation, none of the doors appears to have been grained in this period.
A remnant of red, flocked wallpaper from atop the middle drawing room window provided important evidence for the treatment of James Madison’s walls. This appears to be a Madison-era paper, judging from the red wool flocking, the red-lead pigment of the painted ground, and the flax fibers in the paper substrate. In concert with this discovery, it was noted that the drawing room paper had been applied directly to the brown-coat of plaster -- a common 18th-century practice.
Thus, to what was already the most complete Madison room in the house has been added a pedimented doorway and flocked wallpaper. Indications in the plaster for an 1809- 12 surbase and chimneypiece complete the architectural setting. In the realm of furnishings, nail holes in the original plaster will help to locate the larger works of art, known to have hung in this room.
The results of the investigation have been plentiful, but that is not to say the work is done -- many questions remain. In the event of a restoration of Madison’s home, it will be possible to pursue these issues. Meanwhile, the Montpelier staff continues to record, collate, and interpret the physical evidence bearing on Montpelier’s development.
Publication Date: Fall 2003