The Mocking of Christ

A panel the size of a sheet of writing paper, painted in egg tempera and gold on poplar in Florence or Pisa around 1280, found in June 2019 hanging beside the hotplate of a kitchen in Compiegne and sold four months later for the highest price ever paid at public auction for a work made before 1500. It is one of three surviving scenes from an object that was cut up for the trade, probably in the nineteenth century: radiography at the Louvre's laboratory has since proved that this panel and the National Gallery's Virgin and Child were sawn from the same plank of poplar, with this scene sitting directly beneath the Virgin.

The French state refused it an export certificate on 20 December 2019 and classed it a national treasure, freezing it in France for thirty months. The Louvre did not complete the purchase until 25 October 2023, nearly four years after the sale and more than a year after the freeze had run out. The museum paid with income from the Louvre Abu Dhabi brand licence, a final million euros from the Societe des Amis du Louvre and a gift from two American patrons; the price has never been officially published.

Conservation between 2022 and 2024 then changed the picture itself. Under a nineteenth-century repaint, Christ's open eyes turned out not to be Cimabue's: beneath them lay a white cloth band across the eyes. The scene is therefore not the mocking by Pilate's soldiers that every account of the sale described, but the earlier mocking before the high priest. The object was first shown to the public in January 2025. As of the Louvre's own online record it is not on display.

It is a fragment, and everything about it says so

The panel measures 25.8 by 20.3 centimetres overall and is 1.2 centimetres thick in its present state. The painted surface is smaller than the board: 24.6 by 19.6 centimetres. That gap between panel and picture is the first piece of evidence, because it means the paint was laid down inside something, against edges that are no longer there.

The support is a thin poplar plank with a vertical grain, prepared with a gesso ground, gilded over a bole in the background and painted in egg tempera. The gold is not plain. Its upper border carries punch marks the size of a pinhead, worked into a frieze of volutes, and the same tooling runs through the ornament of Christ's halo. The Louvre's laboratory notes that the identical degree of refinement appears in the monumental Maesta at the other end of the museum's Cimabue holding.

Along the bottom and the left side the gesso ends in a barbe, the small ridge of ground that forms where preparation meets an applied frame. Under the microscope, gilding from that lost frame survives beneath the paint layer on exactly those two sides. On the right edge, X-ray fluorescence found vermilion under a black repaint: the remains of a painted red fillet. The top and right sides carry no barbe at all. So the panel was gilded and painted inside an engaged frame that ran along its bottom and left, and was divided from a neighbouring scene on its right by a painted red line.

When it reached the Paris trade it did not look like a fragment. Strips of wood had been nailed against the grain around all four sides to give it the profile of a small independent icon, which is precisely how the family in Compiegne understood it. Those strips were taken off in the recent restoration. The panel has also been thinned; so have the two panels in London and New York, and all three are now the same thickness, which is the argument that they were cut apart and planed down in a single operation, perhaps in the eighteenth or nineteenth century.

The plank, not the style, is what proves the group

Attribution by eye is arguable for ever. Carpentry is not. In 2019 the auction house established that the National Gallery's Virgin and Child had sat directly above this scene by matching the grain and the worm tunnels of the new panel against a photograph of the back of the London picture. Eric Turquin, whose firm authenticated the painting, put it flatly before the sale: it is the same poplar panel, and that is objective proof.

The photographic match has since been replaced by something harder. At the Centre de recherche et de restauration des musees de France, Elisabeth Ravaud radiographed the panel and set the plate beside the radiograph of the London Virgin and Child. The two compositions came from the same plank of poplar, with the Mocking positioned below the Virgin. The vermilion fillet found under repaint on this panel's right edge points the same way towards the Frick Flagellation, which carries a painted border on its left. The laboratory's conclusion was that the material characteristics of all three small pictures agree and support their belonging to a single object.

What that object was is less settled than the press coverage suggests. The reconstruction that has held since 2003 comes from Miklos Boskovits, published by Dillian Gordon in Apollo: a series of eight scenes arranged as two wings of four, on the model of a set of eight panels by the anonymous Maestro di San Martino alla Palma of about 1320 which Boskovits had reconstructed as a diptych. Gordon went further and proposed that the later painter had used Cimabue's object itself as his model. All three known Cimabue scenes belong to the left wing, and the barbes place them precisely: the Virgin and Child upper left, the Mocking below it, the Flagellation to the right of the Mocking. The missing fourth scene of that wing would be a Betrayal, the Kiss of Judas. The four scenes of the right wing, extrapolated from the later series, would be the Way to Calvary, the Crucifixion, the Entombment and the Last Judgement.

The Louvre's own conservation department declines to close the question. Its published note allows that the original may have been a diptych, following the auctioneer's 2019 publication, or a horizontal dossal, or an altar frontal, or an altarpiece. The Frick describes the parent object as probably eight scenes mounted as a diptych. The National Gallery's catalogue entry is the most cautious of the three, offering a single devotional panel or a dossal before it reaches the word diptych. A reader who wants one answer should note that the museums that own the pieces do not give one.

The restoration changed what the picture is of

Every account written between 2019 and 2024 describes a Christ with his eyes uncovered, and treats that as the point of the picture. Donal Cooper, writing in Apollo a fortnight before the sale, made the observation carefully: Cimabue's standing Christ follows Byzantine precedent, whereas some of his Tuscan contemporaries had already adopted the seated, blindfolded Christ familiar from Northern European stained glass and manuscript illumination, a development traced by Anne Derbes in her analysis of thirteenth-century Passion imagery.

Scientific imaging and microscopy during the 2022 to 2024 restoration established that Christ's eyes are not autograph. They are a repaint laid over a band of white cloth. The blindfold is Cimabue's; the open eyes are somebody else's.

That single finding moves the subject. The Louvre's laboratory records that the presence of the white band indicates the scene is not the derision by Pilate's soldiery but the earlier episode before the religious assembly led by the high priest, and cites Mark 14, verse 65, where those around Jesus cover his face, strike him and demand that he prophesy. The blindfold is not decoration in that passage. It is the mechanism of the taunt: the man who claims to see is made to guess who hit him. Duccio painted the same moment a generation later on the back of the Siena Maesta, with Christ blindfolded and bound before the seated high priest and the cock of Peter's denial on the column behind. The title under which this panel was sold, catalogued and reported, and under which it still stands here, describes the wrong hour of the Passion.

The rest of the reading survives the change. Cooper described a deliberately confusing chain of limbs, two sheathed swords set symmetrically to left and right, a hand that seems to pull the halo away from Christ's head and the hard edge of a sword striking his temple; Stephane Pinta of the Cabinet Turquin described the two compact groups pressing in like the pans of a balance, faces reduced to grimacing masks, and against them a Christ who lets his arms hang. The Louvre's own summary adds what makes the panel unusual for its date: Cimabue rooted the scene in the daily life of his time and dressed his figures in the clothes of his own century.

Who it was made for, and what it was for

No contract, inventory or church record for this object survives. What exists is an argument built on the later series, and it is a good one.

In the Crucifixion from the Maestro di San Martino alla Palma set, now in the Gemaldegalerie in Berlin, Saint Francis and Saint Clare kneel at the foot of the cross. Gordon and Boskovits read that as a commission for a house of Clarissan nuns; Richard Offner had already suggested in 1947 that the Berlin Crucifixion and its companions were painted for one of the two Clarissan convents of Florence, Monticelli, founded in 1260, or Montedomini, founded outside the Porta San Gallo in 1311. The Maestro di San Martino worked only in Florence and its immediate surroundings, which shortens the list. Cooper's reconstruction of the sequence runs: a Clarissan house commissioned an updated version of an older object it knew, and that older object was Cimabue's.

The alternative is Pisan. Cimabue painted the Maesta now in the Louvre for the Franciscan church of Pisa, Vasari records a small Crucifixion painted for the same friars, and Cimabue died in Pisa in 1302. Gordon cautiously identified the London Virgin and Child with a small Madonna di Cimabue recorded in 1829 among the pictures of Carlo Lasinio, keeper of the Camposanto in Pisa, and Pisa had an early and important Clarissan community at Ognissanti. Lasinio, however, gathered pictures into Pisa from across Tuscany, so the Pisan record does not settle a Pisan origin.

What the object was for is clearer than where it hung. Eight scenes of equal weight with no central image give the viewer no natural focus and no hierarchy; the arrangement obliges an eye to work through every compartment in turn. Holly Flora, who curated the 2006 Frick exhibition that first reunited two of the panels and wrote the most recent monograph on the artist, reads the sequence as the visual parallel to the Meditations on the Life of Christ, the immersive Passion handbook composed at almost the same moment for a Clarissan convent at San Gimignano and the most widely copied devotional text of the later Middle Ages. The Louvre put the most richly illustrated surviving copy of the Meditations in its 2025 exhibition to make the point, and programmed a public lecture by Flora on Cimabue and the Franciscans alongside it. This is a Franciscan instrument for a practice of imagining the Passion scene by scene, at arm's length, in a cell.

The material choices belong to the same argument. On a board twenty centimetres wide the laboratory found an abundance of lapis lazuli and red lake, alongside vermilion, ochres, green earth, orpiment and lead white. Ultramarine ground from lapis was the most expensive pigment available and had to be bought before it could be used. Cimabue's contemporary Duccio was made to promise, in the contract of 15 April 1285 for the Rucellai Madonna, that he would pay for the ultramarine and the gold himself. Whoever ordered this object was not economising, and did not intend a private, portable devotion to be a cheap one.

What such an object cost when it was made

Nothing is documented about the price of this panel, or the object it came from, and that is the honest starting point. Two figures from the same decades give the order of magnitude.

The Compagnia dei Laudesi, a lay confraternity attached to the Dominican church of Santa Maria Novella in Florence, commissioned Duccio in 1285 to paint the picture now called the Rucellai Madonna, at 450 by 290 centimetres the largest panel painting known from the thirteenth century. The contract, dated 15 April 1285 and generally described as the earliest surviving Italian document of its kind, set the fee at 150 lire and required Duccio to paint the whole thing himself, to take no other work until it was finished, and to supply the ultramarine and gold at his own cost.

The only payments documented to Cimabue are from the end of his life. The records of Pisa cathedral show him paid ten soldi a day between 2 September 1301 and 19 February 1302 for executing the figure of Saint John the Evangelist in the apse mosaic. A Florentine lira ran to twenty soldi, so 150 lire is 3,000 soldi, and Cimabue's documented daily rate at Pisa would have taken three hundred working days to earn it. That arithmetic is ours, from the two cited figures, and it compares a lump-sum commission for materials and labour with a day rate paid to a mosaicist twenty years later in another city. It is a scale, not an equivalence.

What it establishes is the distance to be crossed. The 19.5 million euros paid in a saleroom in Senlis in 2019 has no relation to what the object cost the people who ordered it, and any modern equivalent quoted for a 1285 sum will differ by an order of magnitude depending on whether it is calculated from prices, wages or share of output. The useful comparison is not to 1280 but to the market the panel was actually sold into, and that is a different section.

Ten works, or twelve, or fifteen, or forty

Cimabue is not a documented artist with an attributed fringe. He is an attributed artist with one documented work.

He is first recorded in Rome in 1272, described as a Florentine painter. The only surviving work documented as his in his lifetime is the mosaic figure of Saint John the Evangelist in the apse of Pisa cathedral, paid for between September 1301 and February 1302; he died in February or March 1302, and the mosaic was finished by Vincino da Pistoia in 1320 and restored at least six times afterwards. Nothing he painted is signed. Every panel given to him, including this one, is an attribution.

The size of the corpus is therefore a matter of where a given scholar draws the line, and the published counts move. At the sale in 2019 the Cabinet Turquin's specialist Stephane Pinta said ten works were known worldwide, all of them in public collections; Anna Brady reported the same figure in The Art Newspaper. The Louvre's own press material for the 2025 exhibition says some fifteen works have come down to us. A reviewer of that exhibition put the number of securely attributed works at twelve. Looser counts that include workshop and disputed pictures run to about forty. This record uses the range and does not pick a number.

That is the frame in which the 2019 discovery has to be read. A new work by Cimabue is not one more picture on a long list. It is a ten to fifteen per cent addition to everything that is known to survive, and the only devotional object of private scale to be added to it.

The attribution, and who has doubted it

The attribution of this panel rests on an attribution to its siblings, and the siblings took fifty years to sort out.

The Frick Collection bought the Flagellation in 1950 through Knoedler in Paris. Scholars agreed at once that it was an important thirteenth-century picture and disagreed about who made it: the choice was between Cimabue and the Sienese Duccio di Buoninsegna. Millard Meiss published it in 1951 as Duccio. The Frick's own account says the doubt persisted for half a century and was ended by the reappearance of the London Virgin and Child in 2000, which was attributed to Cimabue on comparison with the Louvre Maesta and then tied to the Frick panel by carpentry: the grain of the wood, the thickness of the two thinned panels, the pattern of craquelure, the worm holes, and a punched border of two rows of pinprick punches around a scroll of tiny punch marks common to both.

By the time the Compiegne panel surfaced, the framework existed to test it against. Cooper, writing before the sale, called it a clear-cut case and noted that if it were somehow an intelligent fake, technical investigation would quickly say so. He also recorded what he called predictable grumblings in the press over the authenticity, without a named scholar behind them. The specialists he canvassed by email were on the other side: Holly Flora, who expected the discovery to deepen understanding of Cimabue's part in the development of narrative devotional painting; Andrea De Marchi, who observed that this was the first time Cimabue's crowded, choral narration, so evident in the Assisi frescoes, could be seen at small scale on panel; Joanna Cannon of the Courtauld, who described the interweaving gestures and exchanged looks as masterly.

The French state then put the question to a formal body. The Commission consultative des tresors nationaux considered the export application on 20 November 2019 and recorded a remarkable unpublished work by Cimabue, with the attribution supported by unanimous agreement among specialists and by consistent architectural elements, gold decoration, types of halo and identical pigment composition on matching poplar. No named scholar has published a dissent in any source consulted for this record. What follows from that is narrower than it sounds: the panel belongs beyond argument to the same physical object as the Frick and National Gallery pictures, and the name attached to that object is a connoisseur's reconstruction that has held since 2000.

How it was found

In June 2019 a woman in her nineties near Compiegne, in the Oise, was leaving the house she had lived in and called in a local auctioneer to say whether anything in it was worth selling. What was not worth selling was going to be thrown out. The auctioneer, Philomene Wolf, then thirty-two and in partnership with Dominique Le Coent-de Beaulieu at Acteon, had a week to go through the contents and clear the building. The panel was hanging on the wall of the kitchen beside the hotplate of an open-plan kitchen unit, among ordinary objects of no particular interest. The trade press was blunt about how close it came: it escaped the tip by very little.

The family knew the picture was old and considered it an icon. Accounts of what kind of icon differ; the owner is variously reported as thinking it Russian or Greek, and as not knowing where it had come from. Wolf's own recollection, given to The Art Newspaper before the picture was shown to the press, was that the family thought it special but thought it an icon.

The panel went to the Cabinet Turquin in Paris, the firm most closely associated in France with the technical examination of disputed old master pictures, where the specialist in charge was Stephane Pinta. Infrared examination produced an underdrawing consistent with the small body of work given to Cimabue, and the carpentry and punchwork did the rest. It was, Pinta said, the oldest painting the firm had ever examined. The picture was presented to the press in Paris on Tuesday 24 September 2019; the discovery had been announced the day before.

What the auctioneer thought it was worth before that examination is the one figure in this story that no primary source consulted here carries. Tertiary accounts state that Wolf's first working view was 300,000 to 400,000 euros, and the figure is widely repeated. It is recorded here as unverified. What is documented is the published estimate once the attribution was in hand, 4 to 6 million euros, and the opening bid on the day, 3 million.

The sale, bidder by bidder

Acteon offered the panel on Sunday 27 October 2019 at the marche Ordener in Senlis, in a room that seats up to nine hundred people and was full. Dominique Le Coent-de Beaulieu took the bidding. No work by Cimabue had ever been offered at public auction, which meant there was no price history for the artist at all; Le Coent said as much before he started, calling it the first opportunity to give the artist a market value.

The lot opened at 3 million euros against an estimate of 4 to 6 million. The trade press counted eight bidders, three of them in the room; The Art Newspaper counted seven registered bidders, three in the room, and identified them as Italian dealers from Paris and London and a contemporary art gallery from Brussels. The Paris dealer Giovanni Sarti bid for a client to around 15 million and stopped, leaving the Brussels contemporary gallery against Francesco Ortenzi of the London gallery Moretti, who was bidding for an anonymous private collector. Ortenzi won.

The hammer fell at 19.5 million euros. The total with buyer's fees was 24.18 million euros, which The Art Newspaper and most English-language coverage rounded to 24.1 million. Dollar equivalents of 26, 26.6 and 26.8 million circulated depending on the day of conversion and are not additional prices. The two euro figures are the real ones and there are two of them.

Turquin's reading of the bidding was that interest came mainly from museums or people close to museums, with new interest from buyers out of the contemporary art world. Sarti, who had dropped out at 15 million, thought it could have gone higher, and set it against the Madonna and Child by Duccio bought by the Metropolitan Museum of Art for 45 million dollars in 2004 in a private transaction: comparable in his view, and by the father rather than the son.

One bidder present did not bid. The Louvre attended the sale and, for want of sufficient funds, could not exercise its right of pre-emption, the mechanism under which a French public museum may step into a saleroom result at the hammer price. Everything that followed for the next four years was the consequence of that.

What 19.5 million euros meant in that market

The result made the panel the most expensive work made before 1500 ever sold at public auction, and by the auctioneer's own reckoning on the day the eighth most expensive Old Master of any date. The seven above it were Leonardo's Salvator Mundi in 2017, Rubens's Massacre of the Innocents in 2002, Rubens's Lot and his Daughters in 2016, Pontormo's Portrait of Cosimo de' Medici in 1989, Rembrandt's Portrait of a Woman in 2000, Raphael's Portrait of Lorenzo de' Medici in 2007 and Canaletto's Grand Canal in 2005.

That list explains the size of the number better than the estimate does. Pictures of this date effectively do not come to auction, so there is no cohort to price against; the closest comparanda are not sales at all. The Metropolitan's Duccio Madonna, bought privately in 2004 for a reported 45 million dollars, is the benchmark for the period and never went under a hammer. The National Gallery's Virgin and Child, the panel that sat immediately above this one on the same plank, never reached a saleroom either: it was destined for Sotheby's and was instead taken by the state, and its credit line records it as accepted by the British government in lieu of inheritance tax and allocated to the National Gallery in 2000. The Art Newspaper reported the transaction differently, as a private treaty sale for around 7.2 million pounds; both accounts involve tax relief and neither is a market price. Whichever is right, the London panel cost the British public purse a fraction of what its neighbour made nineteen years later.

The other way to read the number is as the price of certainty. Between June and late September 2019 the object did not change; the evidence about it did. The published estimate that resulted, 4 to 6 million, was itself an attribution discount: it priced a Cimabue that scholars might yet reject. The room paid more than three times the top estimate because the carpentry argument removed the risk that the estimate was pricing. The lesson a collector can take from this record is not that attic finds are worth sixty times an initial guess. It is that in a market with no comparables, technical evidence, not connoisseurship, is what the money responds to.

The export machinery, in the order it actually ran

The application for an export certificate was filed on 14 October 2019, thirteen days before the sale. Under the Code du patrimoine, a cultural object above certain thresholds cannot leave France permanently without a certificate, and a certificate can be refused only on the ground that the object is a national treasure.

The Commission consultative des tresors nationaux, chaired by Edmond Honorat, examined the application on 20 November 2019 and advised refusal, finding a remarkable unpublished work whose rediscovery was of major significance to the national heritage. The minister of culture, Franck Riester, signed the refusal on 20 December 2019; it was published in the Journal officiel on 22 December, and the ministry announced it on 23 December. The refusal carried national treasure status for thirty months from the date the decision was notified to the owner, expressly to allow public collections time to raise the money and to let the panel join the Cimabue Maesta already in the Louvre.

Those dates matter because the Louvre's own catalogue entry gets them wrong. Its object history states that the panel was classed a national treasure on 20 November 2019, which is the date the commission gave its opinion, not the date of the arrete. The INHA's Repertoire des tableaux italiens repeats the Louvre's wording. The decision is dated 20 December 2019 and the official text says so.

What the thirty months actually oblige the state to do is less than the phrase saved for the nation implies. Under article L121-1 the administrative authority may present an offer to buy in the interest of public collections; it is not required to. If it does, the price must reflect prices practised on the international market, and if the owner rejects it, each side appoints an expert at its own cost, the experts report jointly within three months, and a third expert settles any disagreement. Once the valuation is in, the state has two months to make an offer at that figure. If no offer is made, the refusal of the certificate cannot be renewed and the object is free to go. The freeze is a right of first refusal with a clock on it, not an expropriation.

The clock here ran out. Thirty months from a decision notified at the end of December 2019 expires in the middle of 2022. The purchase was registered on 25 October 2023, forty-six months after the refusal. No published source consulted for this record records a formal purchase offer under L121-1, and no notice appealing for corporate patronage under article 238 bis-0 A of the tax code, the mechanism that gives a company a 90 per cent tax reduction for funding a national treasure acquisition, has been found for this object. The panel was bought, in the end, in a negotiation with an owner who by then was legally entitled to take it away.

Who owned it in the meantime, and how the Louvre paid

The buyer at Senlis was anonymous on the day and remained officially anonymous through the export dispute. The Louvre's own catalogue entry now names it: acquired for 19.5 million euros excluding fees by the Alana Collection. The Alana Collection is the collection of Italian painting formed by the Chilean financier Alvaro Saieh and Ana Guzman, based in Delaware, and at the time of the sale a large part of it was on public view in Paris at the Musee Jacquemart-Andre. Ortenzi of the Moretti gallery bid for it. Several accounts of the sale, including this record in its previous state, named Fabrizio Moretti as the bidder; the contemporaneous report names Francesco Ortenzi of that gallery, and that is what is recorded here.

Early reporting on the buyer was worse than anonymous. The BBC told its readers the panel had been sold to an anonymous buyer from northern France, and the Associated Press reported a buyer near Chantilly. Both were wrong, and both remain online.

The seller did not live to see the consequences. The elderly owner of the picture died two days after the sale, leaving three heirs. Nothing has been published about the settlement of her estate, the tax paid on the proceeds, or what the family received.

The Louvre's internal process ran through 2023: the acquisitions committee on 12 April, the artistic council of the national museums on 10 May, and the decision, assignment and inscription in the inventory all on 25 October. The work is owned by the French State and held by the Louvre's Department of Paintings under the number RFML.PE.2023.33.1. The minister of culture, Rima Abdul Malak, and the museum's president-director, Laurence des Cars, announced it on 3 November 2023 together with a second national treasure, a Victor Hugo drawing.

The money came from three places and none of them was the public purse in the ordinary sense. A Louvre statement quoted at the time says the museum used part of the income from the brand licence for the Louvre Abu Dhabi; the Societe des Amis du Louvre contributed the final million euros that made the acquisition possible; and a couple who are members of the American Friends of the Louvre also contributed. Des Cars named Harry and Linda Fath, American property developers, and the Societe des Amis du Louvre as the patrons behind the two acquisitions; the society's own president attached its gift to the Hugo drawing, funded from a bequest, which by elimination puts the Fath contribution against the Cimabue, though no statement says so in terms. The Art Newspaper understood that the Louvre matched the 24.1 million euro sale price. The museum has never published the figure it paid.

The contrast with the museum's other national treasure campaign of the same period is instructive. Chardin's Basket of Wild Strawberries sold at Artcurial in March 2022 for 24.3 million euros with fees to the Kimbell Art Museum through the dealer Adam Williams; the export was blocked; and the Louvre completed the purchase in March 2024 with about 15 million from LVMH, about 7.8 million from the Societe des Amis du Louvre and more than 1.6 million from nearly ten thousand individual donors giving an average of 165 euros each. Two objects of almost identical price under the same law in the same year, one bought with a national subscription and a luxury group's money, the other with Gulf brand-licence income and a private American gift, and only one of them ever asked the public.

Four years in a laboratory

The panel that made the headlines was not the panel that now hangs. Photographs put out in 2019 were, as the Louvre's own conservation service records with some asperity, often virtually brightened to increase their impact; the real condition of the painting was not so flattering.

It came to the C2RMF for laboratory study in 2019 and again in 2022, and was restored there between 2022 and 2024 in the painting studios of the Flore site. Clarisse Delmas led for the C2RMF; the paint layer was restored by Audrey Bourriot and the support by Jonathan Graindorge-Lamour; Sebastien Allard, head of the Department of Paintings, and Thomas Bohl, curator of Italian paintings, were the responsible curators. Elisabeth Ravaud led the technical study.

The paint film is exceptionally thin and worn, and had been repainted in several places. A surface clean showed that the dull brown appearance was not simply dirt, and solubility tests found four superimposed varnishes above the thirteenth-century paint. A committee approved a progressive cleaning, treating each layer selectively. The nailed strips came off. The false eyes came off. As of the published account of November 2024 the cleaning of the gold ground was still under way and the final varnishing and retouching had not been settled.

The work has been published where a student can reach it. The C2RMF put out two illustrated accounts, of the study and of the restoration; the technical results appear in TechnE number 58, in an article by Audrey Bourriot and Elisabeth Ravaud running to sixteen pages, in an issue devoted to Cimabue and Tuscany at the end of the thirteenth century and edited by Bohl, the late Marco Ciatti and Ravaud. The Louvre's catalogue entry lists nine bibliographic references and its collection database serves a machine-readable record. The museum's own photography of the object, by Gabriel de Carvalho, is not free of copyright, which is why the plate at the head of this record comes from a different reproduction.

Where it has been shown

The exhibition history of this object is four years old and can be given in full.

Before the sale it was shown to the press in Paris on 24 September 2019 and, according to French accounts, to the public at the town hall of Compiegne on Wednesday 23 October 2019, where somewhat over four hundred people came to see it in the days before the auction. That public showing is recorded here from a tertiary source and has not been confirmed against a primary one.

Between the sale and the acquisition it was not exhibited at all. It was in the laboratory in 2019 and from 2022, and under restoration until 2024. The Louvre presented it in a lunchtime lecture in the Michel Laclotte auditorium on 13 November 2024, given by Thomas Bohl in the museum's Oeuvre en scene series, two months before the exhibition opened.

Its first public presentation as a museum object was A New Look at Cimabue: At the Origins of Italian Painting, at the Louvre from 22 January to 12 May 2025, in room 717, the Salle Rosa, on the first floor of the Denon wing. The show was curated by Thomas Bohl, brought together some forty works, and reunited the three surviving panels of the group for the first time; the catalogue, published with Silvana Editoriale, gives the panel as catalogue number 38. The exhibition was sponsored by Lusis. Its internal Louvre title differs from the one on the posters: the collection database records it as Cimabue. Aux origines de la peinture moderne en Occident.

It then travelled, once. From 6 June to 19 October 2025 it was lent to the Musee du Petit Palais in Avignon, the museum that has held a deposit of 320 Italian primitive paintings from the Louvre since 1976, for the exhibition La jeunesse de la peinture. Cimabue ouvre la voie. The loan was announced as the first result of a strengthened partnership between the Louvre and the city of Avignon.

The siblings have their own loan histories, which are short. The National Gallery panel went to New York for the Frick's exhibition Cimabue and Early Italian Devotional Painting, 3 October to 31 December 2006, coordinated by Holly Flora with Denise Allen, which put it beside the Flagellation for the first time in America; it was at the Ferens Art Gallery in Hull in 2017; and it went to Paris for the 2025 exhibition. The Frick has never lent its panel out of New York in the period covered by the sources consulted here, and it lent it to Paris in 2025.

As of the Louvre's own online record, consulted on 3 August 2026, the panel is not on display.

Three near misses

Objects of this age survive by accident, and this one has a documented record of accidents.

The first is the house clearance. The auctioneer had a week; the contents that were not worth selling were to be thrown away; the panel was hanging next to a hotplate, which is to say in the worst environment in the building for a thin poplar board with a gesso ground. Nothing about the picture announced itself. The auction trade's own account of the sale describes it as having narrowly escaped the tip.

The second is a hailstorm. On Saturday 3 May 2025, nine days before the Cimabue exhibition closed, a violent storm over central Paris drove water through the glazing above the Salle Rosa. Drops fell on the base of the Three Acolytes by the studio of Nicola Pisano, on loan from the Bargello, and about a metre from Duccio's Madonna of the Franciscans, lent by the Pinacoteca Nazionale in Siena and fortunately in a case. The Louvre closed the room within half an hour, firefighters inspected the roof, a damaged glass seal was replaced and the show reopened the next morning. No work was damaged. The Maesta, unglazed and painted on wood, was a short distance from where the water came in.

The third is the one that already happened. The eight-scene series by the Maestro di San Martino alla Palma that made the reconstruction of this object possible has itself been broken up, and the panel from it that corresponds to this Mocking of Christ entered the Barlow Collection in Lancashire and was destroyed in the bombing of Manchester during the Second World War. It survives as a black and white photograph. Cimabue's own Crucifix in Santa Croce spent the same war propped against a wine cask in the cellar of the Castello di Oliveto outside Florence, came back in 1948, and was submerged by the Arno in November 1966.