Portrait of a Young Man Holding a Roundel

A poplar panel of about 1480, 58.4 by 39.4 centimetres, showing a young Florentine man holding, on a parapet, a painted roundel of a blessing saint. The roundel is not Botticelli's. It is a separate mid-fourteenth-century Sienese panel, dated by Richard Offner and attributed to Bartolommeo Bulgarini, which was set into the wood before Botticelli painted around it. The picture is one object, two paintings and two hands separated by roughly a hundred and thirty years.

It came out of Tuscany in the 1780s with a Welsh baron, descended through his family for a century and a half, was sold by trustees at Christie's in London in December 1982 for 810,000 pounds, and spent almost the whole of the next forty years on the walls of the National Gallery in London, the Metropolitan Museum and the National Gallery of Art in Washington. On 28 January 2021 it made 80 million dollars at the hammer, published at 92.2 million with fees, a record for the artist and for any Old Master sold at Sotheby's.

The record for Botticelli before that afternoon was 10,442,500 dollars. Nothing about the panel had changed. What had changed was that it was for sale, and that half a century of museum labels had done the work an attribution argument does elsewhere.

One panel, two paintings, a hundred and thirty years apart

The support is poplar, 58.4 by 39.4 centimetres, painted in tempera. A young man of perhaps twenty, in a red cap and a plain dark tunic over a white shirt, is set against a warm grey ground behind a parapet. His right hand is held to his chest. In his left he holds a circular gilded panel showing a bearded saint in the act of blessing.

That roundel is a real object. It is not a painted representation of a picture; it is a separate painting, made in Siena in the middle of the fourteenth century, cut down to a disc and physically inserted into the poplar panel before Botticelli painted the portrait around it. Richard Offner dated it to the trecento. It is attributed to Bartolommeo Bulgarini, a Sienese painter of the generation after Duccio who died in 1378, more than sixty years before Botticelli was born.

The junction is disguised with some care. The sitter's fingers rest on the parapet and support the roundel from below, and the shadow they cast makes the disc read as set back into the picture space rather than lying on its surface. The illusion is of a young man holding an old and precious thing towards the viewer. The physical fact is that a fourteenth-century devotional fragment has been let into a fifteenth-century portrait.

For an archive that exists to say what an object actually is, this is the hard case. Catalogue it as a Botticelli and you are right, and you have also just described about a tenth of the painted surface incorrectly. Catalogue the roundel separately and you destroy the object. The correct entry is the one that holds both facts at once, which is why this record's medium line reads as it does.

Why a Florentine would hold a Sienese saint

The reading most often offered is a vanitas: living youth in the hands of the sitter, aged sanctity in his hands as well, the new face and the old gold in one frame. That is a modern reading and it belongs to the modern readers who make it, not to whoever commissioned the picture, and this record marks it as such.

What can be said with more confidence is what the gesture was for. A Florentine of the 1480s holding a devotional panel of the previous century towards the viewer is holding an heirloom, and heirlooms in that society were evidence: of a family old enough to have inherited things, of piety old enough to have been practised by grandparents, of a lineage. The saint blesses out of the picture at the person looking at it. Whatever else the arrangement does, it puts the sitter in a line of descent and lets a fourteenth-century hand do the blessing.

The choice of a Sienese panel rather than a Florentine one is not explained in any source consulted here, and this record does not manufacture an explanation for it.

Botticelli had done it before, with metal

The device is not unique in his work, which is the strongest argument that it was his idea rather than a later addition. The Uffizi holds a Botticelli of about 1474, inventory 1890:1488, in which a young man holds a medal of Cosimo de' Medici the Elder, and that medal is also a real object set into the panel: a cast in gilded gesso, not a painted image of a cast.

The two pictures are close in size, 58 by 45 centimetres against 58.4 by 39.4, and close in conception. In both a young man holds out a portable object of a previous generation. In one the object is a medal of the dead ruler of Florence; in this one it is a devotional panel by a dead painter of Siena. In both the object is materially present in the picture rather than depicted in it, which is a much more literal-minded solution than the illusionism the period is famous for.

What the Uffizi picture establishes for this one is that inserting a physical object into a portrait panel was something Botticelli's shop knew how to do and had done at least once before. The roundel is not an anomaly to be explained away. It is a repeat.

The chain: a Welsh baron in Tuscany, and a hundred and fifty years in one family

The provenance published by Sotheby's in 2021 is unusually continuous for a Florentine portrait, and every link before 1935 is inside one family.

Sir Thomas Wynn, third baronet and first Baron Newborough, lived in Tuscany from 1782 to 1791 and is the likely purchaser. From him the panel descended to his son Thomas John Wynn, second Lord Newborough; to his brother Spencer Bulkeley Wynn, third Lord; to Frederick George Wynn, the third Lord's youngest son, who died in 1932; and then to Thomas John Wynn, fifth Lord Newborough. The family seat was in Caernarvon, in north Wales, which is where the picture was when it entered the literature in the 1930s.

It left the family in about 1935, bought by the London dealer Frank Sabin. By May 1941 it belonged to Sir Thomas Ralph Merton, and it stayed with the Mertons for forty years: to Violet Marjory Merton, who died in 1976, and then to trustees, who sold it at Christie's in London on 10 December 1982 as lot 92 for 810,000 pounds, about 1.3 million dollars. The buyer was Sheldon Solow, the New York property developer.

The commission for this record specified a Newhouse provenance. There is no Newhouse in the chain. The name is Newborough, and the distinction matters, because the whole of the pre-1935 history hangs on a Welsh peerage rather than on a twentieth-century American collecting family.

Sir Thomas Merton, who owned it and also x-rayed pictures

The owner from 1941 is worth naming properly. Sir Thomas Ralph Merton was a spectroscopist and a Fellow of the Royal Society, and he collected Old Masters. A picture that spent four decades in the hands of a working physicist who took an interest in the scientific examination of paintings is not an ordinary provenance link, and any technical file on the panel that predates the 1980s would most plausibly begin there.

Whether Merton examined this panel, and whether any record of it survives, is not established in the sources consulted here, and this record does not assume it. It notes the coincidence and marks the question open, because that is what an archive is for. If a Merton file exists, it is the earliest technical documentation of an object whose central physical peculiarity is an insert that would show instantly in a radiograph.

Fifty years of wall time, and what it was worth

The exhibition record is the part of this object's history that the price databases do not carry and that explains the 2021 result better than any connoisseurship.

The panel was at the Royal Academy in London in 1960 in Italian Art and Britain, as number 345. It was on loan to the National Gallery in London from 24 November 1969 to 17 February 1978, a continuous eight and a quarter years. It was at the Metropolitan Museum of Art in New York from 26 February 1987 to 24 January 1989. It was at the National Gallery of Art in Washington from 19 April 1990 to 15 May 2013, a continuous twenty-three years, and appeared there in Virtue and Beauty in 2001 and 2002 as number 26. It was at the Stadel Museum in Frankfurt in Botticelli: Likeness. Myth. Devotion in 2009 and 2010 as number 19. It went back to the Metropolitan on 15 May 2013 and stayed until 25 February 2020.

Add those up. Between November 1969 and February 2020, a span of fifty years and three months, the picture was hanging in a major public museum for roughly forty-two of them. Sheldon Solow bought it in 1982 and lent it out for essentially his entire ownership. When it came to market in 2021 it arrived with an unbroken half-century of institutional endorsement, no condition surprises, no attribution to argue about, and a literature running from 1950.

That is what is being bought at 80 million dollars, and it is the opposite of a sleeper. A sleeper is an object nobody has looked at. This is an object that several hundred curators and several million visitors had looked at, in three national museums, for five decades. The premium is not for a discovery. It is for the absence of any question left to ask.

Hammer against premium, and two totals that are the same number

The sale was lot 15 in Master Paintings and Sculpture Part I at Sotheby's New York on 28 January 2021, with an estimate published as in excess of 80 million dollars. The lot carried no guarantee and no irrevocable bid, which is unusual at that level and means Sotheby's took the risk of a failure in the room.

The hammer was 80 million dollars. Bidding was reported as thin and slow from 78 to 80 million, and closed between Lilija Sitnika, a Sotheby's director and senior client liaison on the Russia desk in London, and Alexander Bell of the Old Master paintings department in London. Sitnika's client won. The underbidder was bidding for an Asian collector. The buyer has never been named.

The published total was 92.2 million dollars in Sotheby's own press release and 92.1 million in The Art Newspaper. Those are not two prices and neither is an error. Both are consistent with a single figure of 92,184,000 dollars, which rounds to 92.2 and truncates to 92.1, and Iconotheca has not located that figure printed in full.

The fee is therefore about 12.2 million dollars, which is 15.2 per cent of hammer. That does not correspond to any single rate. It reconciles as follows: a Sotheby's New York buyer's premium of 25 per cent on the first 400,000 dollars, 20 per cent from 400,000 to 4,000,000, and 13.9 per cent above 4,000,000 produces 100,000 plus 720,000 plus 10,564,000, a premium of 11,384,000. That leaves 800,000 dollars unaccounted for, and 800,000 is exactly one per cent of the hammer, which matches the separate overhead premium Sotheby's charged buyers at the time. Both components are DERIVED here from the published hammer and total; neither schedule has been checked against a printed Sotheby's chart for January 2021.

The point of setting it out is not pedantry. Anyone comparing this sale with the 1982 purchase is comparing a fee-inclusive 2021 total against a 1982 hammer of 810,000 pounds, and the two figures are not the same kind of measurement. The like-for-like comparison is 80 million dollars in 2021 against 810,000 pounds in 1982.

Sheldon Solow, and the ten weeks

Sheldon Solow built and owned office towers in Manhattan, most famously 9 West 57th Street, and was known for holding property rather than trading it and for litigating at length. His art collection was substantial and rarely discussed, which is consistent with a man who lent this picture to museums for four decades rather than hanging it at home.

The consignment was announced in September 2020, with the 80 million dollar expectation attached to it from the start, so the decision to sell was made while Solow was alive. He died in November 2020. The picture sold on 28 January 2021, about ten weeks later, and the sale is recorded as being by his estate.

That sequence matters for anyone reading the result as an estate liquidation. It was not. It was a sale the owner had already set in motion, completed by his executors on the timetable he had chosen.

What the price says about the artist's market

The Botticelli auction record before this sale was 10,442,500 dollars, set by the Rockefeller Madonna at Christie's in New York in 2013. The hammer here was almost eight times that and the total nearly nine times it. Sotheby's called it a record for the artist and for any Old Master painting the house had sold, and the second most expensive Old Master ever sold at auction.

Almost exactly a year later, on 27 January 2022, Sotheby's sold Botticelli's Man of Sorrows in the same series of sales for 45,419,700 dollars. That is the most useful comparison available. Half the money, a year apart, same artist, same house, same week of the calendar. The difference is that the Man of Sorrows is a devotional picture of Christ from around 1500 and this is a secular portrait of a named type from Botticelli's best decade, with fifty years of museum wall behind it.

Sotheby's stated in 2021 that this was one of only three Botticelli portraits remaining in private hands. That claim is the substance of the price and it is also unverifiable from outside the trade; this record repeats it as the house's assertion and does not adopt it.

What is not settled

The sitter is unidentified. The picture has been catalogued in the past as Portrait of Giovanni di Pierfrancesco de' Medici, whose brother Lorenzo di Pierfrancesco was one of Botticelli's most important patrons, and the identification has been proposed rather than demonstrated. Sotheby's own catalogue treats it as uncertain.

The date is unsettled within a twenty-year band. Proposals in the literature run from about 1470 to about 1490, which straddles the whole of Botticelli's central period.

The division of labour between Botticelli and his workshop has been argued over in the same literature, even while the picture is sold and shown as autograph. Technical examination by radiography and infrared reflectography has recorded extensive underdrawing and continuous revision beneath the paint, which is the ordinary evidence for a picture worked out on the panel rather than copied onto it.

And the roundel is attributed, not documented. Bartolommeo Bulgarini is a name proposed for a mid-fourteenth-century Sienese panel of which only a disc survives, embedded in another painting, and unavailable for examination from the reverse. It is a reasonable attribution and it is not a signature.