Aerial view of Monticello's roof, prior to restoration in 1992, with a painted red roofAerial view of Monticello's roof, prior to restoration in 1992

From the time Thomas Jefferson began planning Monticello in the 1760s, he took a keen interest in the question of how to shelter oneself effectively and economically from the weather. He approached it as an architect and engineer and considered some of the most progressive roof structures and coverings of his time for a variety of buildings. In at least one case he can be credited with a truly innovative roof form. This interest went beyond the theoretical and was based on observation and experience, as might be expected from a man who, according to his overseer, always carried around a little pocket ruler.[1]

An opportunity to greatly advance his knowledge of building technology came during Jefferson's stay in France from 1784 to 1789. It was just such an interest that led him one morning in 1786 to the Halle au Blé – the grain market at Paris – to see the great skylighted dome, designed by architects Jacques Guillaume Legrand and Jacques Molinos and completed three years before. There he met Maria Cosway, and from the friendship that ensued came the famous letter to Maria following her departure from Paris. It is known as the dialogue between the Head and the Heart and is, perhaps, the most self-revealing letter that Jefferson ever wrote. Jefferson recreates the circumstances of their first encounter and reveals the workings of heart and mind. The Head complains to the Heart, "I never ceased whispering to you that we had no occasion for new acquaintance; that the greater their merit and talents, the more dangerous their friendship to our tranquillity, because the regret at parting would be greater."

Heart. Accordingly, Sir, this acquaintance was not the consequence of my doings. It was one of your projects which threw us in the way of it. It was you, remember, and not I, who desired the meeting, at Legrand & Molinos. I never trouble myself with domes nor arches. The Halle aux bleds might have rotted down before I should have gone to see it. But you, forsooth, who are eternally getting us to sleep with your diagrams and crotchets, must go and examine this wonderful piece of architecture. And when you had seen it, oh! it was the most superb thing on earth! What you had seen there was worth all you had yet seen in Paris! I thought so too. But I meant it of the lady and gentleman to whom we had been presented, and not of a parcel of sticks and chips put together in pens. You then, Sir, and not I, have been the cause of the present distress.

Head. It would have been happy for you if my diagrams and crotchets had gotten you to sleep on that day, as you are pleased to say they eternally do. My visit to Legrand & Molinos had publick utility for it’s object. A market is to be built in Richmond. What a commodious plan is that of Legrand & Molinos: especially if we put on it the noble dome of the Halle aux bleds. If such a bridge as they shewed us can be thrown across the Schuylkill at Philadelphia, the floating bridges taken up, and the navigation of that river opened, what a copious resource will be added, of wood and provisions, to warm and feed the poor of that city. While I was occupied with these objects, you were dilating with your new acquaintances, and contriving how to prevent a separation from them.[2]

The object of admiration by the Head was one of the wonders of Paris. The superb dome, he learned, was derived from ideas developed by the sixteenth-century architect Philibert Delorme and published in a book that Jefferson made a point to acquire and bring back to America. He was struck by what he called the "honeycombed" form, with short wooden members joined to create cells small enough to be covered with glass.

Monticello''s roof, post-1992 restoration, with unpainted metal roofMonticello's roof, post-1992 restoration

Jefferson's hope of seeing a similar roof structure used over a market at Richmond was not realized, but by 1800 a dome based on Delorme's principles was under construction at Monticello. Jefferson simplified the complicated joinery of mortises, tenons, and wooden keys by substituting laminated sections fabricated with an abundance of nails readily available from his plantation nail factory. He described his version of the Delorme dome in notes accompanying drawings for a proposed town house dating from about 1800. The fortunate recipient of the design is not identified. He stated that "the ribs (or rafters) are made of oak plank inch thick, 13 1/2 I. wide, and cut into lengths of about 4.f. Four thicknesses are put together, mitered together at their ends according to the curve, and breaking joints so that no two joints come together. They are nailed together with half crown nails, going through and clinched. When set up, they are honeycombed by short pieces put in cross wise from rafter to rafter. The rafter is 60" in length."[3]

Jefferson's enthusiasm for what must have been viewed by the recipient as a radical idea for covering a house is apparent. He reassured the would-be possessor of this design: "Nothing is more simple than this structure, and it is so cheap that it is used for barns in France. A very coarse and uninformed carpenter is making mine, who never heard of a dome before. It is the only way of framing a dome so as to give it's hollow to the height of the room." Jefferson stated that "a model of one of the ribs shall be sent" so that there would be no doubt how to construct it.[4] Surely his remark that similar domes were not uncommon on barns in France is an exaggeration, although he might have had in mind the lavish stable with a Delorme dome built for the Duc de l'Infantado at Paris.

By 1802, Jefferson had conceived a more stunning application of the Delorme system. As president, he proposed building a great dry dock at Washington so that, in the interest of economy and peace, the number of Navy frigates in active service could be greatly reduced. He envisioned a single span structure over an area 175 feet wide and 800 feet long (more than three acres). He asked the English architect Benjamin Latrobe, then working in Philadelphia, to come to Washington and develop the scheme from his preliminary design. Jefferson was "convinced it will be a work of no great cost, that it will save us great annual expence, & be an encouragement to prepare in peace the vessels we shall need in war."[5] Latrobe developed the scheme. The House of Representatives considered its cost and promptly tabled it.

In time, Jefferson had the satisfaction of seeing Delorme principles employed in a variety of buildings. It was used for the dome he designed for the Rotunda at the University of Virginia as well as the roof over the heads of the very representatives who had rejected his dry dock proposal. Although it can be said that Jefferson was responsible for introducing the Delorme method of roofing to this country, his real innovation was the zigzag roof. In January 1803, he wrote to his joiner at Monticello, "I conclude absolutely to cover my terras with sheet iron, and have accordingly written to Philadelphia to see if I can procure sheets 15. I. wide and 9½ f. long."[6] He was referring to the "terras," the low hipped roof concealed behind a Chinese railing on top of the house. The iron was to come from the recently established mill at Philadelphia that rolled wrought iron slabs into sheets. Jefferson's problem had been that the pitch of the hipped roof was too low to prevent water from being driven up under the wood shingles. His solution was to take off the shingles and fix what he called serrated lath on top of the sheathing. The serrations formed hips and valleys of what looked like a series of miniature gabled roofs or "rooflets." Sheathing nailed to the configuration of the lath was then covered with sheet iron – one piece bent and placed in the "gutter," or valley, and another similar sheet placed over the ridge and lapping the valley piece. The object was to make a watertight roof without the need for complicated metal seams. Jefferson's scheme involved considerable carpentry to create the rooflets, but the end result allowed him to maintain the same low pitch of the original hipped roof while insuring that water would run off as effectively as on a roof of regular pitch. He developed several variations including one for the walkways over the dependencies flanking the house. In some versions shingles were substituted for sheets of iron, but they all had in common the series of rooflets.

Monticello''s South Terrace reconstruction showing pillars and serrated roofletsMonticello's South Terrace reconstruction showing pillars and serrated rooflets

Latrobe, who stayed on in Washington after his appointment by Jefferson as Surveyor of the Public Buildings of the United States, credits him with the idea. In September 1803, he wrote to his clerk of the works that he had decided to take off the roof of the north wing of the Capitol and "put on one of the President's zigzag roofs of sheetiron" which, he said, is a "species of roof uniting all the good qualities of the pantile, without its bad ones."[7] Jefferson had, in fact, created an equivalent of the interlocking S-shaped clay tile roofs that he saw in Europe.

Jefferson claimed another advantage to his zigzag roofs: they did not require rafters. More than twenty years after constructing his first serrated roof, he continued to promote the idea. In a letter to Ferdinand Hassler, Jefferson summarized its virtues and went so far as to offer it as an improvement to Hassler's design for an observatory. "Permit me ... to suggest to you a construction of flat roofs different from yours. ... in the house in which I live, and it's offices I have flat roofs of a different construction." He went on to explain in detail how they were made and concluded, "I have had upwards of 20 years experience of these roofs in this house, also in one I built at another residence, and more than half our buildings at the University are flat and so covered. they never have leaked, cost less than a rafter roof, as needing no rafters and admit repairs more easily than any other. I think it the best possible roof for an Observatory."[8]

Over the years, Jefferson tried a variety of roofing materials, including wood shingles of several types, lead, sheet iron, and copper, but it was the lightness and, as he understood it, the durability of tin shingles that was favored in later years. Although tin was known to him by the 1790s, it appears that he did not use it until his work at Central College (later the University of Virginia). Jefferson had the distinction of being not only the architect of the college buildings, but the chairman of the board of visitors as well. In that latter role, he requested in 1818 one of the builders, James Dinsmore (well known to him from his work at Monticello), to report on the recent use of tin at Staunton, Virginia.

The original tin-coated iron shingles from Monticello''s dome, believed to have been installed in the early 1820s.The original tin-coated iron shingles from Monticello's dome. These shingles are believed to have been installed in the early 1820s.

Dinsmore closely examined two roofs, one belonging to a Mr. Smith and another to Mr. Cowan, and he met with Absalom H. Brooks who performed the work. His report concluded that a tin shingle roof "may be made as tight as one of any other metal" and he had it "from good athourity that they have been in use in montreal & Quebeck for forty or fifty years without Painting & are Still Sound." The shingles he was describing were tin-coated iron measuring about 13¼ by 10‚Öõ inches. A box usually contained 225 shingles, which was sufficient to cover about a square and a half (150 square feet). They were imported from Wales. Dinsmore stated that the cost of material sufficient to cover one square (100 square feet) was eight dollars. Brooks's charge for putting on was seven dollars per square, the cost of nails noted as "trifleing."[9]

A few days after the report was submitted, Jefferson received from Dinsmore what he needed to know: a square of heart pine shingling (materials, cutting, and putting on) "Cannot at Present be done for less than ten Dollars".[10] Jefferson and the board elected to pay the premium for longer-lasting material and engaged Absalom Brooks to cover the college buildings.

Jefferson's "diagrams" and "crotchets" gained him a reputation as an amateur architect and occasionally he was called upon for designs. In July 1821, he received a letter from Charles Yancey of Buckingham County, Virginia, who inquired, "believing that you have devoted much of your valuable time & reflection to Subjects of Architecture, I have taken the liberty to trespass upon your time & talents (a Common Stock) which we all seem to have a right to draw upon, growing out of a long & useful life devoted both to publick & private good; to draft for us a plan of our Court house, as much in detail as will Comport with your Convenience."[11] In a matter of eleven days from when Yancey wrote, Jefferson responded, "I duly recieved your favor of the 12th and chearfully undertook a compliance with your request. I now inclose the drawings you desired. every thing proposed in them is in the plainest style, and will be cheap altho’ requiring skill in the workmanship. without this it will be rendered barbarous in the execution. of one truth I have had great experience that ignorant workmen are always dearest." He then explained some of the particulars of the building. His final recommendation: "I would advise you to cover with tin, instead of shingles." It was here that he stated "it is the lightest, & most durable cover in the world. we know that it will last 100. years, & how much more we do not know."[12]

At just about the time he produced the design for the Buckingham County courthouse, Jefferson engaged Absalom Brooks to come to Monticello and cover the small pavilion at the end of the north terrace. Jefferson made a point to observe him closely. About a year later, after his design for the courthouse was finally accepted and the building committee happened to be considering Brooks, he revealed to Yancey how the little job at Monticello had enlightened him on both the subject of tin roofing and workmanship. He was quick to point out "mr Brooks’s price of 6D. 30 the square for laying on the tin of a roof, is exorbitant. it may be done, as well as he can do it for 1. Dollar the square. We went on at the University giving him that price until ¾ of the houses were covered. we were led to it from a belief that it could not be done without the very expensive & complicated machine which he used to bend the tin, which he told us was a patent machine, costing 40. D. and not to be had in the US. at that stage of our business & got him to come and cover a small house for me. seeing his machine at work, and how simple the object was, I saw that the same effect could be produced by two boards hinged together. I had this done accordingly, and it did the work as neatly & something quicker than his 40. D. machine, while this could be made for 50. cents. we then ceased employing mr Brokes, and set a common negro man to work with our board machine, & he has covered all the remaining houses as neatly and securely as those done before. any person will learn to do it in a day as well as in a year."[13]

There were others who wrote seeking Jefferson's wisdom on the use of tin. To James Paxton in 1824, he cautioned, "I would advise against painting it. it certainly does no good, and possibly may corrode the tin." He said some workmen put the tin on in whole sheets (13¼ by 10‚Öõ inches) but he claimed that "the half sheet is what 100 years of experience elsewhere has approved, in larger pieces it may contract and dilate too sensibly."[14]

When Nicholas B. Pryor from Nashville, Tennessee, wrote him about problems with his tin roof Jefferson speculated that, "if your tin covering has failed, it must have been from unskilfulness. perhaps it has been put on in whole sheets, or plain like shingles, which will not do. altho the oper[atio]n is so simple that any person of common sense may learn it in 3. hours as well as 3. years, it would take sheets of writing to give all it’s details."[15]

Jefferson continued to promote tin covering, but it should not come as a surprise that we find him early in 1825 intrigued with a new material. William Coffee had been producing cast ornaments for the University of Virginia and Poplar Forest and mentioned his idea for lightweight tile roofs. Jefferson reasoned, "I should suppose your manufactory of flat tiles, as light as slate would probably succeed. costing but 5. D 70 c the square they will come cheaper than any other covering known, and the sufficiency of tile is well enough established by experience."[16]

Jefferson's more than half century of experience with roofs might well have led him to reflect philosophically on the merits of roofing systems: but what would he have written, what instruction given, in response to Francis Eppes's news that the roof at Poplar Forest was leaking not in one but in a hundred places? One suspects that in a dialogue between the Head and the Heart, the Heart would have been troubled but the Head, with reason firmly fixed, apt to repeat the admonition to John Hemmings from 1819: "miss no opportunity of every possible further search whenever rain falls, because if the cause can be unquestionably ascertained, we can remedy i[t]."[17]

- William L. Beiswanger, 1992.  Originally published as "The Art of Roofing, and The Restoration," in Fall Dinner at Monticello, November 5, 1992, in Memory of Thomas Jefferson (Charlottesville: Thomas Jefferson Foundation, 1992).


  1. ^ Hamilton W. Pierson, Jefferson at Monticello: The Private Life of Thomas Jefferson (New York: Charles Scribner, 1862), 20. See Bear, Jefferson at Monticello, 32-33.
  2. ^ Jefferson to Cosway, October 12, 1786, in PTJ, 10:444-45. Transcription available at Founders Online.
  3. ^ The drawing, misidentified as an elevation drawing of Barboursville, is designated N4/K205 and is in the Coolidge Collection of Thomas Jefferson Manuscripts, Massachusetts Historical Society.  The accompanying notes are designated N4a1 and are in the Thomas Jefferson Papers, University of Virginia (quoted section is on verso).
  4. ^ Ibid.
  5. ^ Jefferson to Latrobe, November 2, 1802, in PTJ, 38:619-20. Transcription available at Founders Online.
  6. ^ Jefferson to James Dinsmore, January 3, 1803, in PTJ, 39:256. Transcription available at Founders Online.
  7. ^ Benjamin Henry Latrobe, The Correspondence and Miscellaneous Papers of Benjamin Henry Latrobe, eds. John C. Van Horne and Lee W. Formwalt (New Haven: Published for the Maryland Historical Society by Yale University Press, 1984), 1:325.
  8. ^ Jefferson to Hassler, December 3, 1825, Thomas Jefferson Papers, Library of Congress. Transcription available at Founders Online.
  9. ^ James Dinsmore’s Report on Tin to Central College Board of Visitors, November 10, 1818, in PTJ:RS, 13:381. Transcription available at Founders Online.
  10. ^ Dinsmore to Jefferson, November 18, 1818, in PTJ:RS, 13:399. Transcription available at Founders Online.
  11. ^ Yancey to Jefferson, July 12, 1821, Thomas Jefferson Papers, Library of Congress. Transcription available at Founders Online.
  12. ^ Jefferson to Yancey, July 23, 1821, Thomas Jefferson Papers, Library of Congress. Transcription available at Founders Online.
  13. ^ Jefferson to Yancey, July 12, 1822, Thomas Jefferson Papers, Library of Congress. Transcription available at Founders Online.
  14. ^ Jefferson to Paxton, September 10, 1824, Coolidge Collection of Thomas Jefferson Manuscripts, Massachusetts Historical Society. Transcription available at Founders Online.
  15. ^ Jefferson to Pryor, October 16, 1824, Coolidge Collection of Thomas Jefferson Manuscripts, Massachusetts Historical Society. Transcription available at Founders Online.
  16. ^ Jefferson to Coffee, January 5, 1825, Coolidge Collection of Thomas Jefferson Manuscripts, Massachusetts Historical Society. Transcription available at Founders Online.
  17. ^ Jefferson to Hemmings, November 27, 1819, in PTJ:RS, 15:247. Transcription available at Founders Online.