Juan de Pareja (ca. 1608-1670)

A portrait painted in Rome in 1650 of a man the painter legally owned. Juan de Pareja was enslaved in Velazquez's household and workshop for at least two decades; within months of sitting for this canvas he was signed towards freedom, and he went on to a documented career as a painter in Madrid with signed altar-scale pictures now in the Prado.

The picture then passed through a Roman cardinal's collection, a British envoy's palazzo in Naples, a Yorkshire antiquary and a hundred and fifty-nine years at Longford Castle before Christie's sold it on 27 November 1970 for 2,200,000 guineas. That is 2,310,000 pounds, the first painting ever to make more than a million at auction, and in 1970 there was no buyer's premium at all, so the bid and the invoice were the same number. The underbidder was Thomas Agnew and Sons, acting for the National Gallery of Art in Washington; the National Gallery in London keeps the firm's account of the loss under the title The One That Got Away.

A portrait of a man the painter owned

Oil on canvas, 81.3 by 69.9 centimetres, painted in Rome in 1650. The sitter stands in three-quarter view against a plain ground, in a worn grey-green doublet with a broad collar of Flemish lace that is the one expensive thing in the picture, and looks out of it directly. The paint is thin and fast. The lace is put in with a few loaded strokes that do not describe a single thread and read at three paces as a whole collar.

The fact at the centre of the picture is not stylistic. Juan de Pareja was of African descent, born in southern Spain, and was enslaved in Velazquez's studio and household for at least two decades, including on the day he sat for this. The Metropolitan Museum states it in those words on its own object page, and states in the same paragraph that enslaved artisanal labour was widespread in the workshops of Spanish painters, sculptors, silversmiths and woodworkers of the period.

Two men worked beside Velazquez for years. One was Juan Bautista Martinez del Mazo, who married the painter's daughter and succeeded him as usher at court. The other was owned. The Met holds a Mazo in the same department.

The sitter was a painter, and the pictures are signed

Pareja is not only a sitter. After his manumission he built a career in Madrid and left works that carry his own name. The Museo Nacional del Prado holds his Calling of Saint Matthew, into which he painted a self-portrait, and his Baptism of Christ of 1667, signed and dated by means of letters rendered as though chiselled into a rock at the lower left. The Met's curator David Pullins describes surviving works by him in Spain that run to over eleven feet wide.

The Calling of Saint Matthew places its self-portrait between a figure identifiable as the Madrid painter Jose Antolinez and an unidentified patron. Carmen Fracchia argued in 2012 that Pareja consciously associated himself with the apostle who evangelised in Ethiopia, a figure whom the Afro-Hispanic poet Juan Latino had earlier turned into a biblical Ethiopian. That is an argued reading with a named scholar behind it, not a settled fact about the painter's intention, and it is recorded here as such.

The Met mounted Juan de Pareja: Afro-Hispanic Painter in 2023 and published a volume of the same name edited by Pullins and Vanessa Valdes. The exhibition is the reason the museum's own catalogue entry on this canvas was revised twice, in 2019 and again in 2023.

Rome, 1650, and what the portrait was for

Velazquez left Spain for Italy a second time in 1649, taking Pareja with him. The immediate purpose of the journey was buying: pictures and antiquities for Philip IV. The portrait was made in Rome in 1650, in the months before Velazquez painted Innocent X, and the two canvases are usually read together, the first as the demonstration piece that secured him the second.

An early biography records that Velazquez showed the portrait in Rome, and that when he sent it round to friends by the hand of Pareja himself they looked from the man to the picture without knowing which to address. The anecdote is repeated in every account of the picture and it is worth being precise about what it is: a story told by a biographer about the effect of a likeness, not a document of a commission.

The manumission

Within months of finishing the portrait Velazquez signed papers that would liberate Pareja by 1654. The Met states the date of the freedom papers as November 1650 in its wider record of the two men. What the documents contain is a delayed manumission: the freedom took effect after a further term of service, which is why the two dates, 1650 and 1654, both appear in the literature and both are correct.

The temptation is to read the picture as the cause of the freedom, and the Rijksmuseum curator Stephanie Archangel has publicly refused that reading on the Met's own audio guide, on the ground that Velazquez had worked beside Pareja for years and would not have needed a canvas to notice him. This record follows her in treating the sequence as a sequence and not as a proof of motive.

A cardinal, an envoy and thirty-nine guineas

The picture surfaces in Rome in the collection of Cardinal Tommaso Ruffo, who had it by 1704 and inventoried it in 1734. It descended through two dukes of Baranello in Naples, and in 1776 Vincenzo Ruffo sold it to Sir William Hamilton, the British envoy to the court of Naples. Hamilton's own inventory of 14 July 1798 says the portrait came from the Baranello collection, and that sentence is what welds the Neapolitan half of the chain to the English half.

Hamilton sold up at Christie's in London on 27 and 28 March 1801. The Velazquez was lot 59 and made 40 pounds 19 shillings. William Buchanan, writing in 1824 about pictures imported into England since the French Revolution, gives the same sale as 39 guineas to a buyer he spells Parkes. The two figures are the same money: 39 guineas is 819 shillings, which is 40 pounds 19 shillings. Anyone reading the two sources side by side without knowing the guinea will record a discrepancy that does not exist.

The buyer was Thomas Lister Parker of Browsholme Hall, a Lancashire antiquary who catalogued his collection in 1808 with this picture as no. 30. By 1 May 1811 the 2nd Earl of Radnor had bought it for 151 pounds 14 shillings and 5 pence, a figure recorded in the Radnor collection catalogue of 1909. A Velazquez that had made 39 guineas at auction in 1801 fetched roughly four times that in a private sale ten years later, and both numbers are documented rather than inferred.

A hundred and fifty-nine years at Longford Castle

From 1811 to 1970 the portrait hung at Longford Castle in Wiltshire. Waagen listed it as picture 147 there in 1857; the house catalogue of 1909 gives it as no. 87; Stirling-Maxwell had already described the Radnor portrait in 1848 as perhaps the exercise Velazquez painted before he undertook Innocent X.

Longford is the same collection that later sold Holbein's Ambassadors and two Holbein portraits to the National Gallery. A great English country-house collection selling its Spanish picture in 1970 was therefore not an isolated event but the late stage of a long unwinding, and the price it made was the reason the unwinding became national news.

The sale: 2,200,000 guineas, and no premium

Christie's sold the picture in King Street on 27 November 1970 as lot 110. The bidding was in guineas and it stopped at 2,200,000 of them, which is 2,310,000 pounds. It was the first painting to make more than a million pounds at auction anywhere, and it was the world record for any picture.

There was no buyer's premium. Christie's and Sotheby's introduced the modern 10 per cent premium in London in September 1975, five years after this sale, and the dealer walkout it provoked belongs to a later chapter of the trade. In November 1970 the number the auctioneer called and the number on the invoice were the same, which is why the hammer and the total are one figure in this record and not two.

The guinea is the only complication, and it is a complication on the seller's side, not the buyer's. Under the old convention a lot was bid in guineas, the vendor was accounted in pounds and the odd shilling in each guinea, five per cent, stayed with the auctioneer as commission. That is exactly why 2,200,000 guineas converts to 2,310,000 pounds. A modern reader who sees a five per cent uplift and assumes a buyer's premium has the mechanism backwards.

A dollar figure of 5,544,000 was widely reported. It is a conversion at the rate of the day, and it is derived. It is not a second price and it should never be set beside the sterling figures as though it were.

The underbidder, and who he was bidding for

Alec Wildenstein made the winning bid on behalf of the Metropolitan Museum. That much is in every account. Who pushed him there is not, and it is the most useful single fact in this record for anyone trying to establish how deep the 1970 price really was.

The National Gallery in London holds, under the reference NGA27/24/17/15, two copies of Geoffrey Agnew's account of his firm's attempt to buy the Velazquez. Agnew was bidding for the National Gallery of Art in Washington. The archive catalogues the item under the title The One That Got Away.

So the room contained two American museums bidding through two dealers, and the price was set by a contest between Washington and New York rather than by a single determined buyer. An underbidder establishes the depth of a price in a way that a hammer alone never can, and this one shows the 1970 record was not an outlier bid but a floor agreed by two institutions.

The Italian journal Antichita viva reported at the time that Wildenstein had bought for the National Gallery in Washington. That is the mirror image of what happened and the Met's own reference list flags it as an error. Contemporary trade reporting of this sale is not reliable on who bought what.

A Velazquez for export

The picture left Britain, and the leaving was contested. The Burlington Magazine ran an editorial in February 1971 headed A Velasquez for Export? and Apollo ran one the same month called The End of an Era. Country Life covered the sale in the issue of 4 February 1971. Whatever the legal position, the trade press treated the departure as a national loss rather than a transaction.

This record does not state what happened inside the British export licensing system, because the sources consulted do not establish it. The Reviewing Committee on the Export of Works of Art existed in 1970 and the Waverley criteria applied to a picture of this standing; whether a licence application was made, whether it was deferred, and whether any British institution was given a matching period are all questions this record leaves open rather than filling with a plausible sentence. The one thing that is documented is that the strongest British-linked bid in the room, Agnew's, was placed on behalf of an American museum.

Condition, and the strip of canvas that was folded under

Hubert von Sonnenburg published the technical account in the Metropolitan Museum of Art Bulletin in June 1971, in the same double issue as Everett Fahy's provenance and Theodore Rousseau's appreciation. Three findings matter.

The canvas has never been relined. For a picture of 1650 that has crossed the Alps, the Channel and the Atlantic and hung for a century and a half in an English country house, that is unusual and it is the main reason the surface reads as freshly as it does.

Part of the canvas had been folded over rather than cut, and the conservation carried out after the purchase returned the picture to its original dimensions by unfolding it. A fold is a reversible loss; a trim is not. This picture was lucky in a way that the Met's Caravaggio, cut on all four sides, was not.

Sonnenburg also observed that the technique is related to sixteenth-century Venetian practice, which is the technical form of the standard art-historical claim that the second Italian journey changed how Velazquez put paint on cloth.

The other version, and why it is not this one

A slightly smaller version of this composition is at the Hispanic Society of America in New York. It was formerly with the Earls of Carlisle at Castle Howard and it has at times been argued to be Velazquez's original. The Met's catalogue entry states plainly that it is a highly accomplished early copy of the Met's portrait.

That is the version trap on this object and it is a live one, because the Castle Howard picture has its own aristocratic English provenance, its own literature and its own photographs in circulation. Any reproduction captioned only Velazquez, Juan de Pareja needs to be checked against dimensions and holder before it is attached to this record.

The eighteenth-century Acquaviva claim is the same problem in an earlier form. Preciado wrote in 1765 that Cardinal Acquaviva owned a Velazquez portrait of Pareja; Fahy concluded in 1971 that Preciado had seen a copy. Two portraits of the same man by the same hand have been in circulation as a claim since at least the eighteenth century, and only one of them is in Gallery 625.