Part I

Robert B. Simon

Chapter 1

The Discovery of a Masterpiece

First Sighting

In early 2005 an intriguing painting depicting Christ emerged from the estate of an American businessman. It did not come from an historic house, neither was it found in a culturally celebrated region of the country, nor was it a part of an illustrious collection. In fact, it appears to have been the only painting of significance owned by its recently deceased owner—a work obtained through inheritance and evidently retained more for its devotional qualities than for its artistic merit. The progression of this painting from anonymity to its recognition and acceptance as a lost painting by Leonardo da Vinci is here related in narrative form.

My first sighting of the Salvator Mundi came in the form of a small black-and-white illustration in the catalogue of a minor regional American auction house (Fig. 1.1). The St. Charles Gallery was the subsidiary venue of the New Orleans Auction Gallery, one of a few auction houses active in the lively antique trade of the Louisiana city. Important works of art, furniture, and objects would pass under the hammer at their principal location on Magazine Street. Lesser material was relegated to the St. Charles Avenue location.¹ Their sale of 9 and 10 April 2005 promised 1,196 lots of furniture, decorative objects, antiques, paintings, and works of art drawn from several estates. Lot 664 was designated as ‘After Leonardo da Vinci,’ an auction-house descriptor signifying a copy painted at a considerably later date than its known source, and titled ‘Christ Salvador Mundi [sic].’ The only further description noted that the picture was ‘presented in a fine antique gilt and gesso exhibition frame.’ Indeed, the catalogue illustration of the painting unusually included its frame, perhaps to enhance its somewhat dubious decorative value.

Fig. 1.1St. Charles Gallery catalogue of April 9–10, 2005, p. 110 including the Salvator Mundi.

Considering the limited research resources available to the auction house, as well as the garbled title of the painting, one must assume that the association with Leonardo and the subject of the Salvator Mundi did not arise from any extensive investigation by the St. Charles Gallery, but was transcribed either from a label, later lost, or from an old inventory or appraisal of the former owner’s property. In any case, I immediately recognized the composition of the painting from my own previous studies of Leonardo’s life and works. The Christ as Salvator Mundi was one of three lost paintings by Leonardo known from copies—the others being Leda and the Swan, believed to have been in the French royal collections until the end of the eighteenth century, and the Battle of Anghiari, the grand mural painted in Florence’s Palazzo Vecchio, covered and likely destroyed in the later sixteenth century.

The painting and the problem it presented were immediately intriguing. Copies of works by the great masters are legion and Leonardo’s works were no exception. Hundreds of Mona Lisas populate living rooms, antique shops, and museum storerooms worldwide—most of which were made by dedicated copyists working from the original in the Louvre or from other copies. These, predominantly from the nineteenth or twentieth centuries, are painted on canvas and are often proudly signed by the copyist on the reverse. Earlier copies, some painted during or shortly after Leonardo’s lifetime, are known as well and are the frequent object of speculative theories seeking a second ‘original.’ However, none has found acceptance except from its promoter.

But this image of Christ was more of a curiosity, since relatively few versions of the Salvator Mundi composition were known and, unlike copies of the Mona Lisa, there was little consistency in the details among them; each seemed almost an interpretation of the subject rather than a faithful reproduction of a common original. However, what further set this image apart from a Mona Lisa copy was that the original painting on which it was based was lost.

Strangely, my response to coming across a version of a lost painting by Leonardo did not include speculation that it might be the missing original. Rather it was almost antiquarian in nature. What, if anything, might this painting tell us about Leonardo’s original? And, if one could determine that it was ‘period’—that is, of Leonardo’s time or soon afterward—whether it might have been painted by one of Leonardo’s able pupils or followers—artists overshadowed by the master, but of significant talent and, in the marketplace today, value.

At this point the matter was only at the level of idle speculation for me, the sale being two or three weeks distant, when I received a telephone call from Alexander Parish. Alex and I had been good friends for thirty years and shared an enthusiasm for the problems of ‘applied connoisseurship’ that working as a dealer in old master paintings entails. Our backgrounds, personal and professional, were quite diverse. While I had a rich academic background, with a doctorate in art history from Columbia University, Alex’s deep knowledge in the field came from years of hands-on experience examining paintings, first on behalf of auction houses and dealers, and later on his own account. As colleagues often pursuing the same goals, we would frequently discuss paintings on the market and compare our responses to them.

Alex’s call concerned the same painting that had caught my attention. He had seen it on the website of the auction house and asked whether I had as well. The online illustration (Plate 2) was somewhat better than the printed one, and he was able to note passages of considerable quality, although the painting as a whole seemed compromised by damage or crude repaint, or both. It was impossible to judge the age of the painting from either illustration, but here the basic cataloguing was intriguing. The four words ‘oil on cradled panel’ suggested both that the work was painted no later than the sixteenth century, when wood panel supports gave way to canvas, and that, because the panel had been cradled (an expensive process used to stabilize paintings on wood), the picture had at some point in its history been considered worthy of preservation.

Alex and I agreed that this painting merited further attention. He would contact the auction house for further information and better photographs, while I would undertake some preliminary research. That would necessarily be limited to my personal library, as I was shortly to leave for England to take part in a scholarly conference in Cambridge.

Acquisition

When we next spoke, I was in London and Alex had obtained additional photographs of the painting (Fig. 1.2). In the auction house’s zeal for transparency, these were taken at extreme angles to accentuate the evident faults of the painting, particularly the irregular bowed shape of the panel and the broad areas of repaint on it. They were, however, rather blurry, and gave no further indication of the quality of the painting.

Fig. 1.2Auction House photographs of the Salvator Mundi.

For my part, I had been able to confirm that the painting essentially corresponded to the composition associated with Leonardo’s lost Salvator Mundi, known most reliably from an etching dated 1650 by the Bohemian artist Wenceslaus Hollar (Fig. 1.3): a half-length frontal portrayal of Christ, slightly bearded and with his hair parted in the centre, raising his right hand in blessing as his left held a sphere before him. But how did this painting compare with any previously known painted versions of the subject, and might the painting coming up for sale have been previously recorded? With only a few hours left before I had to leave for Cambridge, I rushed to the Witt Library of the Courtauld Institute of Art, the great repository of photographs of paintings, for a swift review of images associated with Leonardo and his many students and followers. Comparing these with the catalogue illustration was inconclusive. None could be considered superior in quality or design to the painting at auction. Yet, one’s enthusiasm for that painting was immediately quelled by what one could see of Christ’s face, particularly by its crudely rendered beard and vacuous, disproportionate eyes.

Fig. 1.3Wenceslaus Hollar after Leonardo, Salvator Mundi, etching, London, British Museum.

Alex and I conferred once again by phone. Our opinions were similar. The faults and weaknesses evident in the head of Christ could be accounted for in one of two ways. Either the artist who painted it was mediocre, or what we were looking at was the result of one or, more likely, several restoration attempts that had gone awry. Yet, the blessing hand, as much as could be determined from the poor photographs we were studying, seemed exquisitely rendered and in relatively good condition. That itself seemed worthy of one of Leonardo’s better students, and with that conviction, we decided to bid on the painting. Of course, in retrospect it seems almost irresponsible that one of us did not travel to New Orleans to study the picture and attend the auction. But we were both committed to being elsewhere and so Alex arranged to bid on the painting, sight unseen, by telephone.

The factors that determine how high one is willing to bid at an auction are many, but the two most significant for a dealer are the amount one believes the work can ultimately be sold for, and what costs will be involved in achieving that number. These are tempered by one’s confidence in one’s judgment, the time it will take to sell the painting, and the attendant financial risk. It is fair to say that in our informal calculations we were comically inaccurate on all accounts. Yet our belief that we were looking at a fine, if perhaps compromised, painting by an artist associated with Leonardo compelled us to agree to bid at a level many times above the modest auction house estimate of $1,200 to $1,800.

The sale took place early on a Sunday morning, and when I later called Alex from England, he gave me the good news that we were successful in buying the painting. I braced myself for the price we had paid and received the surprising information that there had been only one other bidder and that he had dropped out after one bid. (I suspect he is not happy right now.) The hammer price of the painting was but $1,000; with the buyer’s premium charged by the auction house, the total purchase price was $1,175.

I returned from England the following week, during which time Alex had made arrangements for the painting to be shipped to New York. It was handled exigently, but in a manner far removed from the sophisticated transportation logistics that the painting would require within a few years. The frame had sustained slight damage in transport, but the picture itself appeared to have survived the journey intact. Our first viewing of it out of its frame was revelatory as we both slowly examined our mysterious acquisition in silence, oscillating from studying details to contemplating the effect of the painting overall (Plate 3).

Our initial response was one of immediate relief that we had indeed purchased an old painting and that, whatever it was, we had not been misled by the photographs and the assumptions we made on their basis. The second concerned the areas of Christ’s head that we had deemed problematic. There was no doubt that they were painted by an inferior hand, but it was clear that these were later additions, either broadly painted swaths of repaint or discoloured retouches distinct from the original paint surface (Fig. 1.4). Our ultimate response, one of cautious satisfaction, arose not only from appreciating the remarkable passages that we had suspected were of high quality—in particular, the blessing hand (Fig. 1.5)—but discovering others that we had been unable to discern from the photographs: the intricate knot-work of the crossed bands, the delicate pattern of folds of his tunic, and the extraordinary rivulets of his falling hair.

Fig. 1.4Salvator Mundi, pre-treatment. Detail of the head.

Fig. 1.5Salvator Mundi, pre-treatment. Detail of the blessing hand.

We knew that we had something exceptional on our hands, but how exceptional we little suspected. A decision about how to proceed was warranted, and fortunately we chose a path that would prove the most significant of the saga. And that was to treat the painting at the highest standards possible; not to view it simply as another work that might pass through our inventory, but as an object worthy of the most serious study. As a record of a lost composition by Leonardo, the painting deserved such attention. That the painting proved to be more than that can be attributed to the decision to focus on it the best scholarly and conservation resources at our disposal. Had we not done so, the painting would likely have continued to live in the obscurity that it had for several centuries.

Visiting A Dear Friend

A providential event helped determine our course of action. An art historian friend, Janet Cox-Rearick, mentioned that she was planning to have drinks with Mario Modestini and asked whether I might wish to join them. Mario was a celebrated, or better legendary, figure in the art world—a painting restorer of unsurpassed talent, an advisor to collectors and museums (most notably, Samuel Kress), and a wise and exceptional connoisseur of Italian painting. He had been a generous mentor to me, often welcoming me into his apartment-cum-studio where a startling masterpiece—from a collector, a dealer, or a museum—might not so casually be on his easel. At this time Mario was an extraordinary ninety-eight years of age, which age had generally confined him to his home. I had not seen him for a few years and welcomed the opportunity to visit a valued friend who had been so influential in my career.

I spent a wonderful hour catching up with Mario and his wife Dianne, whom I had known since my time as a fellow at the Metropolitan Museum in the 1980s.² Mario, though physically weak, seemed hardly diminished intellectually. He sat, attired in a stylish silk robe, his warm voice and quiet conspiratorial laugh little changed from the way I had remembered it. But it was clear that his enforced isolation at home had not diminished his enthusiasm for the art that had been his lifeblood. We spoke about recent exhibitions, museum acquisitions, and the art figures that he respected—or not. As I was leaving, I asked Dianne whether she thought Mario might enjoy seeing a fascinating painting I had just acquired—one that I could bring to show him. She heartily accepted my offer and we made plans for my return the next day.

While my proposal was prompted by what I thought might be a kind of entertainment for a friend, I realized that this might be a special opportunity for some exceptional counsel on the painting. Dianne and Mario, independently and together, were recognized as among the most knowledgeable and sensitive conservators in the field of Italian Renaissance painting. Dianne directed the Kress Program in Painting Conservation at the Institute of Fine Arts in New York, where she supervised both the care of Kress paintings in American museums and the training of a generation of conservators to treat them. Following her tenure at the Metropolitan Museum of Art and before establishing this program she had been in private practice in New York, where she had carried out a brilliant restoration of a painting by Parmigianino with which I had been involved.³

Mario had been long retired, but over the years had given me sage advice about problematic paintings that had come my way and of course had worked on many of the great masterpieces of Italian art that hang in museums across the world. But a further thought occurred. Over his career he had acquired a unique intimacy with the paintings of both Leonardo and the so-called Leonardeschi, the master’s close followers. In that context, he was perhaps best known for his role in the 1967 acquisition of Leonardo’s Portrait of Ginevra de’Benci by the National Gallery of Art. Mario had been tasked by the gallery’s director, John Walker, and president, Paul Mellon, to travel to Liechtenstein to examine the painting, both to verify Leonardo’s authorship (which had been doubted by some) and assess its condition. His approval on both counts led to the NGA’s purchase of the painting (with funds from Mellon’s sister, Ailsa Mellon Bruce) and Mario’s subsequent assignment to accompany Ginevra to Washington. ‘She’ travelled in a climate-controlled suitcase next to Mario on a flight to New York, in a seat famously booked for ‘Mrs. Modestini.’⁴

The trip from my apartment to the Modestinis’ was only five minutes by taxi, and to protect the painting sensibly but anonymously, I carried the picture in a plain black plastic bag—what a New Yorker would call a trash bag, a Londoner a bin liner—a practical solution, but in retrospect not the most respectful one. Mario examined the painting on an easel; he was clearly impressed, but said little. Dianne later reported his thoughts: that this was a painting by an artist in the immediate following of Leonardo, but not by him—a picture that achieved an unusual sense of power by virtue of its being slightly larger than life.⁵ He suggested the name of Andrea Solario as its possible author. When I heard Mario’s opinion the next day, I was intrigued and pleased, both by the possibility that this might be an attributable painting and that it could be by one of Leonardo’s most gifted followers.

With the painting on the easel, I asked Dianne for advice in how to proceed with its conservation treatment. As we spoke she proceeded to ‘wet down’ the painting, rubbing cotton wool saturated with mineral spirits across the surface of the painting—a process that for a few seconds can clarify a cloudy varnish and remove some surface dirt. Perhaps daunted by the evidence of what seemed several campaigns of old restoration and repaint, she began by cautioning me about which conservators in New York not to bring the picture to. But as some of the passages of the painting became more legible, she volunteered one or two colleagues that she thought might be up to the challenge. Then, after a few more minutes of study, she asked whether I would like her to do a few cleaning tests. I enthusiastically agreed and Dianne proceeded to probe a few areas with a mild solvent.

The layers of discoloured varnish, dirt, and modern retouching were palpably thick, but even with the relatively gentle solution Dianne was using, dramatic results were immediately apparent. As her cotton wads quickly darkened, areas covered by murky repaint gave way to passages of great quality alongside distinct, but limited areas of paint loss. Dianne proceeded from one area to another, each time asking permission to proceed. Most of the small cleaning ‘windows’ revealed great promise beyond. Others did not—especially one or two along the left side of Christ’s face, which seemed to expose nothing other than a broad white fill material, presumably compensating for a complete loss of the paint.

After an hour of this investigation, Dianne proposed that I leave the painting for her to examine and explore over the next several days. Although it had not been my intention—consciously, at least—in bringing the painting to her, that she had initiated preliminary work on the painting, and with such intriguing results, was unexpected and exciting.

I did not expect to hear from Dianne for a week at least, but at the end of the next day, she called. ‘Bob, I cleaned the painting: come over!’ Standing again in front of the picture, I saw a painting greatly transformed (Fig. 1.6). Gone was the cloud of dark varnish that had given the work some sense of harmony, even if perverted, and in its place was a raw surface disrupted by losses, cracks, disturbing variations in colour, and a band of gesso fill that followed an arc around the left side of Christ’s head and continued to the bottom of the painting. But gone as well were the most disturbing features of the figure—the strange red colour of his hair, a partial beard that looked as if it were theatrical make-up, and a pair of mismatched beady eyes. What now emerged were brilliant passages of delicacy and complexity, colours that with slight saturation were beginning to recover a vibrancy that had been long diminished, and, with Christ’s face, a haunting expression of power and solemnity.

Fig. 1.6Salvator Mundi, after initial cleaning (May 2005).

Treatment in Earnest

Dianne’s initial cleaning was followed by mechanical excavation of the gesso fill on which the most egregious restorations had been painted. The reason for its initial application was now evident, as it had been applied to disguise a valley formed by two longitudinal sections of the panel that had split and had each become slightly bowed. The fill had been intended to create a new level surface that would cover the underlying break and simulate the appearance of an undamaged panel. But as the fill was in places more than two inches wide, a broad area of the composition had been reconstructed atop the new gesso layer—and, as was painfully evident, by an artist far less talented than Leonardo.

With the removal of the repaint and fill came the discovery that much of the original paint surface beneath was miraculously intact. But there were localized areas of profound damage: thin, vertically oriented bands where the surface had been scraped down to the preparation, including two in which the raw wood of the panel was visible. It was now evident that, in the effort to create a flat surface, the front of the panel had once been brutally planed, essentially shaving the peaks off the highest points of the split and bowed panel.

Before any further work could be done on the paint surface it was imperative that the panel itself be treated. For this Dianne suggested that the painting be shown to George Bisacca, a specialist panel conservator at the Metropolitan Museum of Art. A few days later I brought the painting (now more carefully wrapped) to George’s studio at the museum. His examination of the panel confirmed what Dianne had already observed. At some point in its history the panel had warped and checked, forming the long, curved crack. To resolve this problem the panel had been thinned and then glued to another panel (a technique known as marouflage), then cradled. George recommended a treatment for the painting that would involve the removal of the supplemental panel and cradle so that the original panel could be properly repaired and conserved. Unfortunately, George’s commitments to the Metropolitan Museum precluded his undertaking the project, and it would not be until the following year that work would begin on the necessary structural work on the panel.

For that Dianne was able to recommend Monica Griesbach, who had been her student at the Conservation Center of the Institute of Fine Art and who had gone on to specialize in the treatment of panel paintings. She had been working abroad, but returned to New York in early 2006 to open her own conservation studio. The Salvator Mundi was one of the first projects to enter her new space on West 35th Street; it was to remain there for four months as the delicate and meticulous process proceeded. Both the cradle and the auxiliary panel that had been affixed as a support were made of pine and these were slowly removed mechanically until the original panel, which could now be seen to be walnut, was revealed (Fig. 1.7). The long, curved split was exposed and seen to be the most significant of the faults that had plagued the panel over the years. Other old fractures became evident, all radiating from a knot in the wood located beneath the ascending band across Christ’s chest, just to the left of the crossing. The panel, now separated into two large pieces and five small fragments, was then rejoined in a manner that permitted the gentle natural curve of the panel to return (Fig. 1.8). Splits were re-glued, and hand-fitted pieces of comparable wood—salvaged walnut from the sixteenth-century—were inserted in areas of weakness. After the panel was stable, a secondary spring-based system was devised and installed to provide support to the delicate panel while permitting it to flex minutely in response to changing atmospheric conditions (Fig. 1.9).

Fig. 1.7Salvator Mundi, verso of panel during conservation. The back of the original panel is at the left; the remnants of the auxiliary panel with the pattern of the removed cradle, at the right (May 2006).

Fig. 1.8Salvator Mundi, the original panel rejoined in the studio of Monica Griesbach (July 2006).

Fig. 1.9Salvator Mundi, verso of panel with new auxiliary support (September 2006).

Twentieth Century History

After the auction, Alex had inquired of St. Charles Gallery whether they had any information concerning the painting’s prior history. They kindly informed him that the picture had been consigned from the estate of a Basil Clovis Hendry, Sr., a sheet-metal contractor of Baton Rouge, Louisiana. What little could be discovered of his life gave no indication of an interest in Renaissance art. The only cultural associations mentioned in his obituary noted that he had led a big band orchestra called the Musical Monarch in the 1930s and 1940s, and for many years had directed the Men’s Choir at Our Lady of Mercy Catholic Church in Baton Rouge.⁶ An online real estate listing for his former residence in East Baton Rouge called ‘Le Maison-sur-le-Lac’ included a photograph showing the Salvator Mundi hanging over a staircase (Fig. 1.10). By its less-than-prominent placement and what we had discovered of Mr Hendry’s religious commitment, we could only assume that the painting had been owned and cherished more for devotional purposes than artistic ones.

Fig. 1.10The Salvator Mundi hanging in the residence of Basil C. Hendry, East Baton Rouge, Louisiana, 2005.

However, there was a clue on the painting itself that pointed to its earlier history. Before its removal, the cradle bore an inscription in white paint: ‘CC 106’; while on the lower left corner of the frame the same number, but without letters, appeared in black: ‘106’ (Fig. 1.11). The significance of these inscriptions was soon evident.

Fig. 1.11Salvator Mundi, painted inscription on cradle (above); painted inscription on frame (below).

Over the previous months I had begun a review of the literature on the Salvator Mundi. The most significant publication on the subject had been a 1964 article, ‘Leonardo’s ‘Salvator Mundi’’, by the German scholar Ludwig Heydenreich, which first appeared in the journal of Leonardo studies Raccolta Vinciana.⁷ While various scholars had earlier referenced individual paintings and drawings of the Blessing Christ from Leonardo’s circle, Heydenreich was the first to group what was known visually of the composition and to provide an iconographic context for it. He began with the Hollar etching of 1650 (Fig. 1.3) and associated with it seven paintings of the same type. Six of these were illustrated in his article—paintings in the Detroit Institute of Arts and the National Museum of Warsaw, as well as others known to have been at one time in private collections: the Vittadini Collection at Arcore, near Monza; the Viktor Stark Collection in Zurich; the Trivulzio Collection in Milan; and the Yarborough Collection, London (Fig. 1.12).

Fig. 1.12(a-f) Versions of the Salvator Mundi composition, composite. Upper row, left to right: Detroit, Detroit Institute of Art; Warsaw, Wilanów Palace Museum; Milan, private collection, formerly Arcore, Vittadini Collection. Lower row, left to right: formerly Zürich, Stark Collection; Milan, Trivulzio Collection; formerly London, Yarborough Collection.

Although the basic composition of each of these paintings is comparable, the variety in detail, expression, and style one from another, and from our painting is striking. The seventh painting of the group listed by Heydenreich was a work once in the Cook Collection in Richmond. Heydenreich had declined to reproduce this version, noting ‘Because of the ruinous condition of the painting we forgo an illustration.’⁸ However, I had found a photograph of the work in the files of the Witt Library, back when I began my initial investigation at the time of the painting’s purchase, and providentially, I had made a photocopy of it for my study (Fig. 1.13). Heydenreich was, if anything, polite in his assessment of the painting. The painting as reproduced in the photograph presented an alarming version of Christ, featuring piercing eyes, a crooked moustache, unkempt beard, and a bold series of corkscrew curls in the hair descending along the left side of the face. What was also apparent from the photograph was a pattern of slightly curved diagonal lines passing through both the tunic and sphere that Christ holds at the right. That these lines did not correspond in any way to the design of composition suggested that they reflected the pattern of the wood grain of the panel that had been used.

Fig. 1.13Salvator Mundi, photograph by Gray, prior to the 1913 catalogue of the Cook Collection, London, Witt Library.

That pattern was unmistakable and all too familiar: it was identical with the wood grain pattern in our painting. At first it seemed inconceivable that the frightening Christ of the Cook photograph could be the same man as ours, but then an examination of the blessing hand, which had essentially remained untouched over the centuries, assured me that the photograph recorded the same painting at an earlier moment in its history.

I had known of the Cook Collection from several great paintings that had once famously been part of it. The great Adoration of the Magi begun by Fra Angelico and completed by Fra Filippo Lippi in the National Gallery of Art in Washington is in fact known as the ‘Cook Tondo,’ (Fig. 2.1) while other celebrated alumni of the collection include Jan Van Eyck’s Three Maries at the Sepulchre (now in the Boijmans van Beuningen Museum, Rotterdam), Velazquez’s Old Woman Cooking Eggs at the National Gallery of Scotland, Edinburgh, Antonello da Messina’s Christ at the Column in the Louvre, and Rembrandt’s Portrait of a Boy (Titus?) in the Norton Simon Museum, Pasadena. But I was unaware of the full scope of the collection, as well as its history, until I began looking into the past of our Salvator Mundi.

The collection had been formed by Sir Francis Cook (1817–1901) largely in the last quarter of the nineteenth century and had descended through his son Frederick to his grandson Sir Herbert Cook, a barrister and a sometime art historian (1868–1939).⁹ At the time of Francis’s death the collection consisted of about five hundred paintings, as well as antiquities, Renaissance sculpture, majolica, bronzes, tapestries, and decorative objects. Under Herbert’s care and direction, the collection of paintings housed at Doughty House in Richmond was enlarged, growing to as many as 630 works.¹⁰ What is astonishing was not the size of the collection alone, but the quality and importance of the paintings, with representation from the most celebrated masters, as well as superior examples by the secondary ones. There is no doubt that through the twentieth century it was the most important private collection of old masters in Britain (after the Royal Collection). That it was from an early date ‘freely open to genuine students’ raised its prominence among both scholars and the interested public.¹¹

Herbert was closely involved with the study and care of the collection and edited a lavish three-volume catalogue of the paintings, published between 1913 and 1915. The volume of Italian works, written by Anglo-Finnish art historian Tancred Borenius, was the first of these: A Catalogue of the Paintings at Doughty House, Richmond, & Elsewhere in the Collection of Sir Frederick Cook, Bt., Visconde de Monserrate. The set, amply proportioned and elegantly designed, was a model catalogue for its time. ‘All but the least significant works are illustrated’, and it is a measure of the low esteem in which the Salvator Mundi was held that no photograph of it was included.¹² Nonetheless, it was not difficult to find the painting listed on page 123 under the name of Leonardo’s follower Boltraffio (Fig. 1.14):¹³

Boltraffio (free copy after) / 106. Salvator Mundi … Bust of Christ who has dark auburn hair and wears a blue tunic, with brown bands. He raises his right hand in benediction and holds in his left a transparent glass globe. The flesh tints are reddish. Dark background. / Panel, 25 ½ × 17 ½ inches (0·64 ×0·445 m.) / Doughty House, Long Gallery, no. 1A. / Photo Gray 28992

Fig. 1.14Catalogue Entry from the 1913 catalogue of the Cook Collection.

Any lingering doubts that our Salvator Mundi was the Cook painting were eliminated once I realized that the number 106 printed in Borenius’s catalogue was in fact the catalogue number of the picture, and that the painted numbers on the frame and panel’s verso corresponded to it and alluded to its ownership (‘CC’ for Cook Collection) and catalogue number for both reference and visitor identification.

Borenius’s brief comments acknowledged the parlous state of the painting, while associating it with the version of the composition in the Vittadini Collection, which had recently been cited by Berenson as a work by Boltraffio: ‘Purchased for £120 from Sir J. C. Robinson in 1900, under the name of Luini. This picture, which has suffered both by over-cleaning and repainting, appears to be a free copy after the Salvator Mundi by Boltraffio in the collection of the late Signor G. B. Vittadini at Arcore, near Monza.’ To which Herbert Cook, as editor, added, ‘I should prefer to say a parallel work by some contemporary painter of Leonardo’s school.’¹⁴

The few factual details given in the entry were of some interest. The notation ‘Photo Gray 28992’ established that this was the negative number associated with the photograph of the painting that I had found at the Witt Library. The photographer was William Edward Gray, a London photographer known for documenting artworks.¹⁵ The painting’s location in the Long Gallery at Doughty House is given as no. 1A. From the later Abridged Catalogue of the collection, which lists paintings in the order of the hang, we know that the Salvator Mundi was placed immediately at the left as one entered the Long Gallery.¹⁶ A photograph of the Long Gallery, thought to have been taken around 1905, shows the Salvator Mundi glimpsed at an extreme angle, hanging high on the wall, skied with two paintings below (Fig. 1. 15).¹⁷

Fig. 1.15View of the Long Gallery, Doughty House, Richmond. The arrow indicates the Salvator Mundi

The immediate provenance for the painting is given too. It was purchased in 1900 from Sir John Charles Robinson (1824–1913), perhaps best known for his work in forming the sculpture collection of what was to become the Victoria and Albert Museum.¹⁸ The two had met in the 1860s, and in 1868 Cook had begun his collection by purchasing several paintings from Robinson, both directly and at auction in Paris, where Robinson was constrained to sell pictures following his departure from the South Kensington Museum.¹⁹ Over the next thirty years Robinson served as advisor to Cook in the acquisition of works of art for his collection, while at the same time serving as a source for more than half. Robinson acquired his paintings largely at auction, for the most part in London, and a review of his extant account books indicates that by the 1890s he was acting more as dealer than collector, as he generally did not hold on to works for more than a year or two.²⁰ Unfortunately, those records end in 1897, and as there is no mention of the acquisition of the Salvator Mundi, we may assume he purchased the painting sometime between that year and its 1900 sale to Cook.²¹

A review of auction catalogues during those years reveals what was likely to have been Robinson’s source, a painting sold at a Christie’s sale on 31 March 1900. Lot 33 of the ‘Pictures and Drawings’ auction that day was a ‘Salvator Mundi’ catalogued as by ‘B. Luini’—auction house shorthand for a painting attributed to, but not by, B[ernardino] Luini, one of Leonardo’s most distinguished followers. It was sold for £6 6s. and purchased by someone named ‘Gribble’. The dimensions as given (24 ½ × 17 inches) are slightly smaller than the Salvator Mundi’s today (25 13/16 × 17 7/8 inches), a relatively minor variance which may be accounted for by the imprecision of the cataloguer, a less revealing frame, or perhaps the panel treatment (marouflage and cradling), that may date from this time. The purchaser at the auction, ‘Gribble’, is to be identified with Thomas James Gribble (1851–1905), a dealer and ‘fine art commission agent’ who on at least one occasion acted as Robinson’s agent at a Christie’s auction.²² And significantly, it was ‘under the name of Luini’ that Cook, as Borenius noted, acquired the painting.

The consignor of the painting was the estate of the ‘Late Joseph Hirst, Esq., JP of Wetherby, Wimbledon, and formerly of Leeds’. Hirst (1814–1900) had been a cloth merchant in Leeds, moving to Wimbledon by 1891, but is not known to have been a collector or otherwise involved in the arts. However, his eldest son, also named Joseph, born in 1843 but deceased five years before his father, was a noted archaeologist and writer, educated and ordained into the priesthood in Italy, and later rector of Ratcliffe College, a distinguished Catholic school in Leicestershire. His career and commitment to his faith might make him a more likely source for the painting than his father, to whom the Salvator Mundi would have passed on his death.²³ In any case, we have no knowledge of the painting’s immediate whereabouts prior to Joseph Hirst.

On the death of Sir Francis Cook in 1901 the collection was divided between his two sons, with the elder, Frederick (1840–1920), receiving the ‘pictures and drawings, the antique sculptures and marbles, the tapestries, glass and terra cottas.’²⁴ Frederick, more concerned for business than art, left the stewardship of the collection to his son Herbert, who was actively involved with its care throughout his life, until financial difficulties necessitated the first of many sales in 1939, by which time the collection had passed to Herbert’s son Francis Ferdinand Maurice Cook (1907–78).²⁵

While we know of the Salvator Mundi’s acquisition by Sir Frederick Cook in 1900, the painting appears not to have been immediately exhibited at Doughty House, as the first catalogue of the collection, issued in 1903, omits it.²⁶ Perhaps this was the period in which the painting underwent its unfortunate restoration. However, from soon after, as indicated by the Long Gallery photograph and the 1913 catalogue, the painting remained on view, evidently in the same location high on the wall. It is there listed in the 1932 Abridged Catalogue, although by then demoted to ‘Milanese School.’²⁷ Scholarly attention to the painting was, in any case, negligible. Other than in Cook Collection publications, the only citations of the painting in its fifty-eight-year tenure in Richmond were two by Wilhelm Suida, simply listing the picture as one of the treatments of Leonardo’s composition, and one from Sir Kenneth Clark, who termed the painting one of the versions ‘less close to the original.’²⁸

This art-historical obscurity was amplified by curatorial neglect. S. C. Kaines-Smith, who was hired as curator of the collection in 1941, annotated the entry on the Salvator Mundi in copies of the 1932 catalogue with the remarks ‘frightfully overpainted’ and ‘Not to be restored (after test).’²⁹ In a report prepared by the painting restorer W. H. J. Drown on 2 June 1942 the picture is listed under the heading of ‘Pictures mostly in bad condition which must be tested to ascertain whether of sufficient quality & value to justify the necessary outlay for restoration’ with the comment ‘most amateur state of repaint.’³⁰ With the onset of World War II, the bulk of the paintings collection was evacuated from Doughty House—most to the Cook properties at Cothay Manor in Somerset, a few to Porthallow in Cornwall.³¹ The more cumbersome and what were considered lesser works, including the Salvator Mundi were stacked in the basement of Doughty. The adjacent galleries were in fact hit by German bombs in 1944, but, miraculously, the painting escaped injury. However, the damage rendered Doughty House unusable for the display of art and the Salvator Mundi was consigned to storage. The painting emerged briefly for a touring exhibition titled Sacred Art; A Selection of Paintings from the Cook Collection, which toured three English museums in 1947 and 1948, but was then placed in a leased storage facility at Addison Crescent in London.³² John Somerville writes, ‘It is still there on a 1949 list and though dealers and collectors came and went there throughout the 1950s to see “what they wanted or that might be for sale,” it must have stayed there unnoticed or unwanted until it was consigned to Sotheby’s in 1958.’³³

Sales from the collection had continued through the 1950s and 242 paintings had been sold through dealers by 1952; the residue, consisting of what were considered minor works, 136 paintings in total, was consigned to Sotheby’s for an auction sale on 25 June 1958.³⁴ The catalogue preface stated ‘the present catalogue lists all the paintings which remain except for a group of about fifty pictures which the Trustees intend to retain indefinitely. The sale of the pictures listed in this catalogue brings therefore to a close the dispersal of what can fairly claim to be the greatest assemblage of paintings formed in this country during the nineteenth century.’ Kenneth Clark attended the sale, later writing to Bernard Berenson, ‘In a fit of madness I even bought some pictures at the sale of the last remnants of the Cook Collection.’³⁵ Despite his manifest knowledge of Leonardo’s works, he ignored the Salvator Mundi. The highest price achieved was £3,800 for the large Botticelli workshop Descent of the Holy Spirit, soon acquired by the Birmingham City Art Gallery. At the other end, nine paintings sold for less than £50, the Salvator Mundi being one of them. As lot 40 and catalogued as ‘Boltraffio,’ the surname alone signifying that the painting was considered a later work ‘in the manner of’ the named artist, it brought £45 (Fig. 1.16). The purchaser was given as ‘Kuntz,’ a figure who remained unidentified until journalists for the Wall Street Journal were able to establish a plausible connection to the last known owner of the painting in Louisiana.³⁶

Fig. 1.16Sotheby’s sale catalogue of June 25, 1958 (p. 19).

They associated the Kuntz who purchased the painting at the Cook sale with Warren E. Kuntz and Minnie Stanfill Kuntz, a New Orleans couple known to have collected art and antiques on their European travels. The two are recorded as having departed from England within two weeks of the Cook auction, sailing home on SS Zoella Lykes from Southampton.³⁷ They lived in a small c. 1849 house on Bourbon Street in the historic French Quarter of the city, where the Salvator Mundi was presumably kept.³⁸ Warren Kuntz died in December 1968; his wife, who in 1973 had moved to the Audubon neighbourhood of the city, in 1987. Minnie’s elder sister Clotide had been married to Basil Earl Hendry; their son Basil Clovis Hendry, the nephew of Minnie, inherited their art holdings, and hung the Salvator Mundi in his house in East Baton Rouge (Fig. 1.10). Following his death in 2004 his heirs engaged a local appraiser to value his personal property. The Salvator Mundi appears in the estate appraisal, without any reference to Leonardo, as ‘Continental School (19^th Century), “Portrait of the Head of Christ.” Oil on panel. Framed, 28 × 17 ½ inches. Poor Condition’, with a stated value of $750.³⁹ The painting was sent for sale to New Orleans and at some point before the auction in April of 2005, an association with Leonardo’s lost composition, though tentative, must have been made, as evidenced by the essentially accurate cataloguing by the auction house. Yet neither the painting’s pre-sale estimate, nor its ultimate purchase price would suggest that it could have been Leonardo’s lost original.

Four months after the auction, New Orleans was devastated by Hurricane Katrina, which caused unprecedented flooding over much of the city. Reflecting on the timing now, and considering how the painting had also escaped destruction when Doughty House was bombed during the war, one can only wonder at the strange fortune that permitted the Salvator Mundi to survive, even as we must acknowledge the fragility and tenuousness of all works of art and the dangers that have caused so many to be lost forever.

The Game’s Afoot

While I now had some indication of the painting’s twentieth-century provenance, the determination of what the painting actually might be—who might have painted it and under what circumstances—remained elusive. There was, without doubt, a commercial motive in pursuing such research. But in truth this was secondary to clarifying the mystery that enshrouded the painting and the prospect of directly studying an unknown work associated, even if tangentially, with the greatest of all masters. It was with full empathy, but with what would later seem awful irony, that I came across the opening lines of an article written by Herbert Cook, once the painting’s owner, published in The Burlington Magazine in 1911:

The interest taken in Leonardo da Vinci was never greater than it is today. A whole library of comment has grown up round the inexhaustible subject of a man to whom the title of Universal Genius may most fitly be applied. Anything which can throw even a sidelight on this wonderful Superman is of cardinal value, and therefore any material, even though in itself relatively insignificant, which helps us to that end, becomes of importance. In the sphere of painting so little comparatively remains to us of Leonardo’s own creations that what is vaguely called the Leonardesque demands our closest study. His pupils, followers and imitators reflect the glory of their master and teacher, and at times it is only through them that we get glimpses of the invisible source of inspiration. Admitting, as we well may, the artistic inferiority of what is called the Post-Vincian school of Milan, we may be thankful that it has preserved for us themes, compositions and motives of the master himself which but for it would have been for ever lost.⁴⁰

Leonardo has always been a passion for me, personally and professionally. In this I am hardly alone, and my Leonardo résumé is one duplicated by thousands worldwide. I had seen all of his known paintings in person; viewed many hundreds of his drawings; attended exhibitions devoted to his artistic, scientific, literary, and mechanical work; and visited the shrines marking his life, from the putative casa natale in Vinci to his last residence, the Chateau du Clos Lucé in Amboise—finally paying my profound respects at the nearby Chapel of St Hubert, where Leonardo’s mortal remains are believed to lie. As an art historian, I had a good knowledge of his career, but only limited familiarity with the huge specialist literature devoted to him. The process that my research necessarily followed was straightforward: to read everything I could find on the subject of the lost Salvator Mundi, and to examine, for the most part through photographs, all the related images—paintings, drawings, and prints—that could be connected, even tangentially, with Leonardo’s treatment of the subject.

Heydenreich’s article on the Salvator Mundi provided an excellent introduction to the problem. His conclusion, based on the differences among the versions of the composition and the manifest lack of distinction of any of them, was that Leonardo had never completed a fully realized painting of the subject. Rather, he conjectured, a drawing, now lost, was likely to have been the source of the replicated image.⁴¹ While this proposal was fundamentally contradicted by the emphatic subscription on Hollar’s etching of the Salvator Mundi (‘Leonardo da Vinci pinxit, Wenceslaus Hollar fecit Acqua forti, secundum Originale. A^o 1650’), it could not be argued with, given the lack of any painting known to Heydenreich that could conceivably be associated with the Master’s hand.

Such a painting did subsequently appear, a work of considerable quality that then belonged to the Marquis de Ganay in Paris (Plate 8). Briefly mentioned as a copy by Carlo Pedretti in 1973, the painting was advanced as the lost original by Joanna Snow-Smith, first in an article in 1978, then in a catalogue to accompany an exhibition of the painting in 1982.⁴² Snow-Smith’s research, both regarding the provenance of the painting and its relationship to Hollar’s etching was thorough, but essentially flawed by unprovable suppositions. And although the composition of the Ganay Salvator Mundi appeared to be the closest known to Hollar’s etching, the marked difference in style between it and Leonardo’s known paintings resulted in her attribution’s finding little support among scholars.⁴³ Snow-Smith’s 1982 publication did, however, illustrate and discuss the ex-Cook Collection painting that was the focus of my interest, although she discounted it as the possible source for Hollar: ‘That the Cook painting could not have been the prototype of Hollar’s etching is evidenced, however, by the fact that there is no jewel in the center of the stole nor an accommodation in the orphrey for the insertion of one; the embroidery merely continues the hexagonal pattern without interruption whereas Hollar includes a change in the pattern around a mounted jewel.’⁴⁴ In making this judgement, Snow-Smith necessarily based her opinion on the evidence of the c. 1912 Witt photograph. But the centre of the crossed bands was precisely the area then covered by fill and repaint. Dianne Modestini’s removal of the overpaint and excavation of the fill had in fact revealed the presence of the jewel and the precise design seen in the Hollar etching that appeared to be absent.

Over the following months I pursued my planned immersion in the Leonardo literature. Kenneth Clark’s 1935 association of two red chalk drawings at Windsor Castle (Plates 6 and 7) with the Salvator Mundi composition (and its possible relationship to a painting of the young Christ requested by Isabella d’Este) was significant, but generally the few more recent publications that referenced the lost painting added little of substance.⁴⁵ However, a dogged review of citations, museum catalogues, auction records, research websites, and online databases permitted me to find several more painted versions and variants of the Salvator Mundi composition. None of them came anywhere near the level of quality of our painting and many were notably amateurish. Nonetheless, they reflected the fact that the composition had some currency, although limited, in the early Cinquecento. Furthermore, through a kind of philological review of the constancies, changes, and embellishments of each picture, the salient aspects of the original were made evident. In addition to those paintings already cited by Heydenreich and Snow-Smith, these included paintings in museums in Quimper, Bergamo, Padua, Toronto, Rome, Mantua, and London, a significant version in the Church of San Domenico in Naples, and other pictures at various times on the art market.

E Pluribus Unum

In September 2006—with the panel now conserved, brought into plane, and safely supported—the Salvator Mundi was returned to Dianne’s studio at the Conservation Center of the Institute of Fine Arts in New York. It joined one of six or seven paintings being treated as part of the Kress Program in Paintings Conservation, which Dianne directed. Students and faculty, including those from the art history program of the Institute of Fine Arts, freely circulated in the studio, but, remarkably, no one seems to have taken special notice of the bust-length figure of Christ blessing. When time permitted, Dianne continued the cleaning of the painting—a process slowed not only by her teaching commitments, but by her growing awareness of the significance of the painting—and the deliberative, probing, and meditative considerations on how best to treat it.

I would make periodic visits to the studio as the remaining passages of overpaint were removed, some with solvents, others laboriously and mechanically with a scalpel. Although only small areas were involved at this point, each recovered area was a precious find and, significantly, each area was of the highest quality. However, the background of the painting proved less revealing. Test cleans indicated that under the muddy grey-brown background lay remnants of green paint that had largely been scraped away. The green layer itself was an addition, likely from the sixteenth century, and had covered an original near-black background, one which, however, appeared not to have survived. For these reasons, there seemed little point in removing the later background, since it would only need to be reconstructed.

One afternoon that autumn Dianne called me on my mobile phone and asked me to come over to her studio as soon as I could: there was an anomaly with the painting that she wished to discuss with me. I arrived at her studio toward the end of the day. Her students had all left, and the painting sat on an easel in the cool light of dusk. She reported that she had been working on the blessing hand and had found that remnants of the later background colour still covered the edge of one of the fingers. She had proceeded to remove the overpaint and decided to extend the cleaning better to clarify the spatial relationship between the hand, the background, and Christ’s cascading hair. That area had been visually confusing, in part due to what seemed an irregular light-brown abrasion that extended from the right of Christ’s thumb into his hair. However, what was now evident after this small area of overpaint was removed, was that the left portion of the problematic area was quite distinct from the hair in both shape and colour. It was in fact pink in tone and, although not fully realized, unmistakably thumb-shaped (Plate 4). While this new ‘second’ thumb shared its lower end (the proximal phalanx) with the visible thumb, its upper end (the distal phalanx) was oriented more vertically than in the finished hand, where it is cocked to the left. In retrospect it was discernible after the initial cleaning (cf. Fig. 1.6), but its distinct shape and future significance were not appreciable amid the visual “noise” of the surrounding area at that time. We studied the painting in silence as we each attempted to account for this revelation and understand its significance. All the other versions of the composition, now numbering over twenty, as well as Hollar’s etching, consistently portrayed the thumb bent to the left. We tried and failed to come up with an explanation other than the most obvious one: that this first placement of the thumb was a pentimento, an initial idea for the orientation of the thumb, one never fully completed, but carried out in pink underpaint and later obscured by the background once the final position of the thumb had been determined. The frightening part of this conclusion was that the painting before us would then necessarily have been the prime version of the composition, the earliest of all, since the others followed the revised orientation. This painting was Leonardo’s original.

While the idea that our painting might in fact be the lost Salvator Mundi had been an unspoken fantasy, one which we had methodically resisted due to its fundamental improbability, we had in an instant convinced ourselves of Leonardo’s authorship. With that came a host of issues, concerns, and challenges. The first, most practical one, came from Dianne, who no longer felt comfortable leaving the painting exposed and, even if in only a slight degree, vulnerable in her studio. She would now keep the painting locked in the Conservation Center’s safe when not actively working on it. And for me, research on the painting would have to be supplemented by a strategy as to how best to introduce the painting to the community of scholars.

Dianne completed the cleaning of the painting and documented its state with photography (Plate 1) as we decided how best to proceed with the restoration. While perhaps alarming to a layman, the appearance of the painting devoid of later additions and alterations was encouraging. We considered the possibility of an ‘archaeological’ restoration—leaving all losses unretouched—and later a simulation in Photoshop was prepared to judge the possible effect. But while such a restoration would have the benefit of easily distinguishing Leonardo’s hand from any interventions by Dianne, the integrity of the painting, its ability to function and communicate with the viewer, would be irretrievably lost. Instead Dianne suggested a plan of restoration in which limited defined losses would be retouched to provide continuity and legibility to the figure, while any broader areas of loss, particularly along the longitudinal crack and fill, would be toned to harmonizing colours but not otherwise reconstructed. The goals were several: to allow the painting to live once again as a work of art, to direct the viewer’s eye to the original passages from Leonardo’s hand, and not to perplex the viewer—especially the scholarly viewer who would judge the painting’s authorship—with areas of restoration that could never achieve the level of quality of Leonardo’s original.

The revelation of the thumb pentimento suggested that we look more carefully for other changes that might have occurred during the process of the painting’s creation. Soon after I had begun studying the painting in 2005, I attempted taking infrared photographs using my own digital camera, but the results had been of limited quality due both to the amount of overpaint then present and the technical limitations of the photographer. With the painting now in its cleaned state and in the controlled lighting environment of the Conservation Center, I proceeded to take a new series of infrared photographs, which were then digitally assembled by Nica Gutman-Rieppi, then associate conservator for the Kress Collection Conservation Program. Nica later followed up with imaging using an Infrared Vidicon camera (IRR), as well as X-ray. The X-ray revealed little information—only confirming the areas of loss that were already visible. As we would later discover, Leonardo’s use of lead white in the preparation of the panel limited its efficacy.⁴⁶

The infrared imaging was, however, revelatory as several notable changes could be observed. Perhaps most prominent was a shift in the placement of the crossed bands which had first been sited higher and to the right of its final location (Fig. 1.17). The band or orphrey that serves as a hem across Christ’s chest was repositioned as well. And the fingers in the hand holding the orb had clearly been rethought. They were originally longer by more than a centimetre, their adjustment reflecting a better understanding of the size and mass of the orb they grasp (Fig. 1.18). A more subtle change was observable. The infrared imaging revealed that the central jewelled ornament was first conceived on a larger scale, both broader and longer than ultimately realized. The pattern of dark preparation revealed in the infrared image corresponded to some degree with the reserve for this area summarily sketched by Leonardo in one of the two preparatory Windsor drawings (12525; Plate 7).

Fig. 1.17Salvator Mundi, infrared reflectogram, detail of chest and hand.

Fig. 1.18Salvator Mundi, infrared reflectogram, detail of hand and orb.

Additionally, another change was noticeable in daylight. Through the rigorous rectilinear geometric pattern of the orphrey bands a different motif was visible: gently curving elliptical arabesques, painted in a pale-yellow or white colour, which must have reflected an initial concept for the decoration (see Fig. 5.1).⁴⁷ These changes, all subtle but significant, further indicated that the design elements of the composition had not been copied or transferred intact from another painting or cartoon, but had developed and been refined on the panel itself.

Sharing the News

Dianne completed her work on the Salvator Mundi in July 2007 following her plan of selective retouching, and the painting was then professionally photographed (Fig. 1.19). The transformation was staggering: the combination of localized retouching and the ‘quieting’ of the distracting losses achieved a balance in which the power and the conception of the figure could be palpably felt. On another late afternoon visit to her studio, as I gazed intensely at the painting for minutes on end in the waning light, I had a strange, almost visionary experience: volumes seemed to fill the head of Christ and this two-dimensional object seemed transformed into a corporeal being. The apparition lasted but an instant, and I do not attribute its cause to anything other than a fusion of excessive personal involvement and visual fatigue, but when others would later tell me of their intense emotional and spiritual responses to the painting, I would not be a disbeliever.

Fig. 1.19Salvator Mundi, mid treatment (December 2007).

In these months my own research continued, in New York and elsewhere. I was able to take several trips to Europe, both working in libraries and studying related paintings, drawings, prints, and sculpture in galleries and museums. What I was not able to do was to share the focus of my study with colleagues, friends, or art historians of the period—many of whom I had come to know well over my career. It seemed to me that, as much as I absolutely believed in Leonardo’s authorship of the painting, any suggestion of it at this point could only be met with the greatest scepticism, if not hostility. When asked the habitual questions of what I was working on or whether I had come across any interesting paintings lately, I would reply with some wearily ambiguous response, while suppressing a jubilant scream.

In September 2007 I was in Florence and called on Mina Gregori, the great connoisseur of Italian painting, with whom I had had friendly and helpful correspondence over the years. In her apartment, we discussed various paintings I had been working on and I decided to test the waters by showing her the recent photograph of the Salvator Mundi. She studied it in silence, nodded, muttered a few words of which the only one I could clearly understand was ‘leonardesco’ and then said to me, rather disappointingly at the time, ‘Bisogna vederlo’: it needs to be seen.

Providentially, Mina had already planned a visit to New York and so, two months later, she was able to examine the painting first-hand in Dianne’s studio. As she scrutinised the panel, Dianne explained the conservation treatment that had been undertaken and I spoke briefly about what I had discovered concerning the painting’s history. Mina commented on some specifics of the painting, noting particularly brilliant passages, the iconography of the painting (especially the essentially beardless Christ), and the emphatic use of sfumato, the subtle use of delicate shading intimately associated with Leonardo and his followers. She made no statement about attribution, however, and as we left the room, I could not resist asking her who she thought had painted the picture. She looked at me with an expression that seemed to suggest that the answer was painfully obvious, if arrived at only after some internal debate. ‘È lui,’ she said: ‘It’s him’.

Mina’s endorsement of the attribution was significant, not alone because of her stature as an art historian, but as it gave me confidence that I was not indulging in some self-created fantasy of a type that I had witnessed in owners of paintings for my entire career. An attribution to Leonardo is a priori unbelievable—something I well understood—and having my rational (or so I believed) conclusion of Leonardo’s authorship confirmed by an independent authority assured me that I was not deluding myself.

I now considered to whom the painting should next be shown. Many criteria informed that decision. I sought an art historian intimately familiar with Leonardo’s paintings, but one whom I knew to be independent and fair-minded; someone who was recognized as a sophisticated connoisseur of Renaissance paintings, but also who might not see me primarily as an opportunistic art dealer. I feared approaching a Leonardo specialist directly as I knew that ‘new’ Leonardos were the bane of their profession. I needed to contact someone whom I absolutely respected, but someone who would not dismiss me as a lunatic. I decided to write to Nicholas Penny, who was then serving as senior curator of sculpture at the National Gallery of Art in Washington. Despite my confidence in Penny’s probity and objectivity, I thought that mentioning Leonardo’s name or sending an image of the picture might result in my email’s being ignored. Instead I wrote asking whether he might be in New York in the near future as I wanted to show him a Renaissance painting ‘of extraordinary importance’. His curiosity was no doubt piqued; he replied, promising to visit on his next trip to New York in December.

On the day we had fixed I retrieved the painting from Dianne’s studio and brought it to my gallery on East 78th Street. Penny arrived and after a brief greeting I walked him to the easel on which the picture rested. I offered no background information regarding the painting’s provenance, conservation history, or what I had discovered concerning its relationship to the Leonardo’s Salvator Mundi composition. In retrospect, it seemed to take only a nanosecond for him to understand what he was looking at; and only a minute or so for him to decide how to express it to me. But he soon sat down and briefly shared his immediate thoughts on the painting. He firmly believed in Leonardo’s authorship, but recognized that acceptance of the attribution would require a consensus of scholarly opinion. He wished to consider how I might proceed to achieve that and promised to call me the next day with his thoughts. In this, I could see that, as affable as he was to me, his motivation was dictated by a sincere conviction in the importance of the painting and his awareness that any improper introduction of such a major a work of art would likely lead to unwarranted controversy.

When he phoned, he informed me that, just a few days before his visit, he had been appointed director of the National Gallery in London, where he had served as Clore Curator of Renaissance Paintings through the 1990s. He described his ambition that the gallery grow in its role as a place for scholarly inquiry and suggested that once his tenure began I bring the painting to London for comparison with the National Gallery’s Leonardo, the Virgin of the Rocks. At that time he would invite leading Leonardo scholars, all appropriately wary of putative Leonardo discoveries, to study the two paintings together and discuss the attribution of the Salvator Mundi. That the venue for this meeting would be on the ‘neutral ground’ of the National Gallery would, he hoped, foster objective analyses of the painting. Until then, he suggested that my ‘home institution,’ the Metropolitan Museum of Art, be made aware of the painting.

Museum Visits

After the Christmas holidays the painting was brought to the Metropolitan Museum for study and safekeeping. Curators Carmen Bambach, Andrea Bayer, Keith Christiansen, and Everett Fahy—each specialists in the field—examined the painting while the conservation staff, headed by Michael Gallagher, used their advanced equipment to produce more revealing infrared reflectograms than had previously been obtained. These showed the pentimenti in the design of the composition with greater clarity, while demonstrating aspects of the artist’s technique and process: his practice of brush drawing with a carbon black medium; his habit of manual blotting to smooth the paint surface, as evidenced by the presence of palm or hand prints on Christ’s forehead; and the likely use of a partial cartoon for the design of Christ’s head, as evidenced by a row of spolveri along the top edge of the upper lip.⁴⁸

Nicolas Penny took up his position in London in February 2008 and soon after asked Luke Syson, curator of early Italian paintings at the National Gallery, to travel to New York to view the Salvator Mundi. In the meantime, he was contacting Leonardo specialists to coordinate a time when each might be in London. It was not until May that such a meeting could be organized. Logistical arrangements were made—not the least of which was securing travel insurance for the painting; even as a work ‘attributed to Leonardo’ the cost was substantial. I collected the painting from the Metropolitan Museum and fitted it into a carrying case that had been specially constructed for it. The Salvator Mundi spent the evening in my home—in retrospect, another incredible occurrence—and the next day I hand-carried it on a plane to London.

The precious cargo was deposited at the National Gallery and the Leonardo scholars arrived: Carmen Bambach, David Alan Brown (who came a few days later), Maria Teresa Fiorio, Martin Kemp, and Pietro Marani—in addition to curators and conservators from the National Gallery. Each was naturally cautious; some spoke openly, others among themselves or not at all. But the questions that emerged—at least, those I was privy to—concerned dating, iconography, condition, and the relationship to other works by Leonardo. Attribution seemed not to be an issue. There was no debate.

Although I was present for this informal symposium—mostly to answer any questions that might arise—the individual scholars subsequently conveyed their opinions in confidence to Luke Syson and Nicholas Penny. A few days later Penny called to tell me that the sought-after consensus had in fact been achieved—at least among the participants. I returned to New York with the confidence that the cautious prefix ‘attributed to’ might be deleted from future descriptions of the painting.

For the study session in London, Leonardo’s Virgin of the Rocks was taken down from exhibition, removed from its frame, and temporarily brought to the Painting Conservation Laboratory (Fig. 1.20). The degree to which the varnish of the painting had discoloured since its last treatment was immediately evident and initiated a more thorough examination that soon resulted in the decision to clean the painting. And shortly after that, the first ideas for undertaking a major exhibition devoted to Leonardo’s career in Milan began to take shape. It was, however, not for another year, that the framework for the exhibition was established, and not until January of 2010 that the National Gallery formally requested the loan of the Salvator Mundi for what was to be Leonardo da Vinci; Painter at the Court of Milan, to open in November 2011.

Fig. 1.20Salvator Mundi and the Virgin of the Rocks in the National Gallery Conservation Lab (May 2008).

The Phoney War

In the interim, research continued, but now with contributions from others. Martin Kemp, who came to New York in December 2008 to study the painting at length and to meet with Dianne Modestini, began to prepare an extensive study of the painting, now published in expanded form in the present volume. He also introduced me to one of his former students, Margaret Dalivalle. I had come across two intriguing references to what appeared to be paintings of Christ by Leonardo in the collection of King Charles I.⁴⁹ Margaret, whose own work focused on English collecting in the seventeenth century, agreed to pursue that reference through archival research. Her important discoveries are presented in this volume.

With the Salvator Mundi back in New York and with the prospect of its public exhibition in the future, we reconsidered its presentation. For its examination by scholars the painting had been minimally inpainted, glazes on the orb not replaced, and the eyes left unresolved. And, as discussed, visible pentimenti—the ‘second’ thumb and the variant patterns of the crossed bands—were left plainly visible. But these choices limited the aesthetic appreciation of the painting. For this reason, Dianne resumed her work on the painting. She was also reconsidering her decision to have left the muddy grey background intact. Not only did that colour seem incorrect—especially considering the intense and well-preserved browns and blacks that were being revealed by the cleaning of the London Virgin of the Rocks—but the thickness of the background paint, even if only infinitesimally greater than that of the figure of Christ, compromised the illusionary space that Leonardo had created. Dianne removed the grey background, the traces of the green layer beneath, and some attendant fill material mechanically under the microscope. As anticipated, there were areas, especially near the top of the panel, where no remnants of the original background survived. But, remarkably, much was recoverable, and what survived was a rich black, well in accord with backgrounds in other Leonardo paintings such as La Belle Ferronière, St. John the Baptist (Plate 9; both Musée du Louvre, Paris) and the Portrait of a Musician (Pinacoteca Ambrosiana, Milan). After much discussion and photographic documentation, she decided to mask the ‘first’ thumb and the arabesque pattern in the orphrey—fascinating details, but ones that made the painting more an object of art-historical curiosity rather than a viable work of art. Other, more subtle revisions to her restoration, including a resolution of Christ’s eyes, succeeded in creating the vibrant image that one sees today (frontispiece and Fig. 1.21).

Fig. 1.21Salvator Mundi, after treatment (February 2011).

An ultraviolet fluorescence photograph taken after the completion of the conservation treatment clearly shows both the extent and limitations of the restored passages (Plate 5). The inpainting, which fluoresces a dark, purple-black colour, closely corresponds with the areas of loss and damage visible in the cleaned-state photograph (Plate 1).

Scientific and technical studies of the painting were undertaken at this time. Nica Gutman-Rieppi took eight microscopic cross-sections from inconspicuous yet representative places on the painting; these were analysed and studied by her and later subjected to a variety of advanced examination techniques and processes by conservation scientists at the Philadelphia Museum of Art, Beth Price and Ken Sunderland, by Thomas J. Tague, Jr, of Bruker Incorporated, and Richard Newman of the Museum of Fine Arts, Boston.⁵⁰ The results provided important information about Leonardo’s materials and techniques: that the panel was prepared with an organic sizing, likely glue, followed by two similar preparatory layers, both bound with oil—the lower ground containing lead white mixed with particles of ground glass, while the upper ground consisted of finer particles of lead white and glass with the addition of a bit of lead tin yellow; that he employed a limited number of pigments—three reds (vermilion and red iron oxide, and perhaps a lake), lead tin yellow, ultramarine blue, umber, lead white, and two blacks (carbon black and bone black), as well as a transparent additive in some of the paint mixtures (soda-lime glass); that he created flesh tones by using successive layers of extraordinarily thin washes of colour; and that he used walnut oil as his painting medium, although perhaps not exclusively.⁵¹ From examination and study of the x-radiographs it became evident that Leonardo laid in part of the composition (the crossed orphreys) with lines incised into the ground and that he utilized a compass to draw both the orb (as indicated by a hole made by the needle point that corresponds with the centre of the sphere, although no inscribed lines describe the circumference) and a semi-circle used to establish the shape of Christ’s head.

The use of ultramarine blue for the tunic of Christ merits further comment. As the Raman spectroscopic analyses have indicated, the ultramarine pigment utilized by Leonardo came from lazurite mined in Sar-i-Sang, Afghanistan—a rare mineral in the Renaissance, more costly per ounce than gold. Significantly, no other version of the Salvator Mundi features a tunic of the blue; most are red, with a few tending toward brown, mauve, or grey. When prepared following the complex extraction recipe recorded by Cennino Cennini, quartz is leached from the pulverized stone, and the distilled pigment acquires the brilliance and chroma associated with the blue of Fra Angelico’s paintings.⁵² However, the ultramarine pigment that Leonardo used for the Salvator Mundi retained large particles of quartz, indicating that the lazurite was simply ground without further purification. Like innumerable other paintings, including other works by Leonardo, the ultramarine has degraded and become more transparent over time, permitting the dark underpaint to assert itself.⁵³ With the associated phenomenon known as ‘ultramarine sickness’ the blanching of the medium produces a cloudy effect so that the modelling becomes less legible, with little distinction between the lights and darks. Such details as the sharp pleats beneath the hem and the omega-shaped drapery fold that bunches over the lower left band of the stole, clearly articulated in Leonardo’s preparatory drawing (as well as in many copies of the painting), have become more difficult to read over time.

Interestingly, recent studies of ‘ultramarine sickness’ have pinpointed its cause and confirmed that saturation of the paint can partially, if only temporarily, restore the earlier appearance of an affected area—a phenomenon already reported in the nineteenth century.⁵⁴ A manifestation of such changes can be observed by comparing photographs of the Salvator Mundi from its time in the Cook Collection, through the stages of its recent conservation, to the present, as shapes and folds in the blue tunic seem repeatedly to intensify and recede, their visibility corresponding to the successive application and absorption of varnish over the years.⁵⁵

With the restoration complete we now approached the task of presenting the Salvator Mundi in an appropriate period frame. A review of Leonardo’s paintings in museums revealed no frame type that seemed suitable—most were later in date—and while a search among dealers internationally found several handsome candidates more or less contemporary with the painting, all proved too flashy or distracting to adorn the sombre and serious blessing Christ. Eventually an early sixteenth-century Tuscan cassetta frame, with a delicate gilt design over an ebonized panel, was found in New York; fortunately, it required only a slight adjustment to fit the painting.

Practical considerations followed. A protected climate-controlled environment was needed for the painting, which had spent the years since its acquisition between conservation studios and art storage facilities. Alex Parish and I came to an arrangement with a friend and colleague, Warren Adelson, a specialist in American painting who would become our partner in the painting. His gallery in a townhouse a few steps from the entrance to the Metropolitan Museum had private secure space and provided the ideal combination of access and security for the painting. There we were now able to show the painting to scholars and friends. David Ekserdjian visited and contributed valuable insights, many subsequently formulated into an article on the painting.⁵⁶ David Alan Brown visited and shared astute observations, as did David Rosand, my long-time friend and dissertation advisor at Columbia; Patricia Rubin; and Vincent Delieuvin of the Louvre. Responses from others outside of academe—memorably, the painter Jamie Wyeth—reminded me of the extraordinary impact of the painting, something which years of my focused attention had inevitably dulled.

Revelation

The painting’s existence was known to but a few people until 2008; then, as scholars and museum personnel were introduced to it, its circle grew. Over the next three years rumours spread about the discovery of a lost masterpiece, and many in the art world, in academe, and even in the art press came to know of the re-emergence of the Salvator Mundi but remained discreetly silent. The National Gallery exhibition Leonardo da Vinci: Painter at the Court of Milan was announced in July 2010, but no mention of the Salvator Mundi was then made; it was intended that the picture would be ‘unveiled’ shortly before the opening of the National Gallery exhibition, scheduled for November 2011. In association with the exhibition, a documentary on Leonardo was being prepared by the BBC to be shown on the eve of the opening. The discovery of the Salvator Mundi would be part of that film, and so in March 2011 the presenter Fiona Bruce, accompanied by a crew, arrived in New York to film the painting and to discuss it on camera with me and Dianne Modestini. Da Vinci: The Lost Treasure, as it would be titled, accurately detailed the secrecy with which the painting was held. And in seeking verisimilitude, the first encounter by Ms Bruce with the painting was captured on film. Her response—finding the image powerful, spiritual, and disturbing—would be echoed by many visitors to the exhibition in the coming months.

The plans for a late-October announcement of the painting’s discovery were derailed by the appearance of an article in the July issue of the American magazine Art News. Its author, the magazine’s editor and publisher Milton Esterow, had written a short but well-informed piece about the discovery of the painting and its planned inclusion in the National Gallery exhibition.⁵⁷ The response was enormous and worldwide, as the press, both in print and online, seized on the story. The paucity of information provided in the original article yielded a host of increasing inaccuracies as journalists struggled to put an individual stamp on what they wrote. Esterow had used the 1912 photograph from the Witt Library (Fig. 1.13)—the only available image of the painting—which caused several publications to suggest that it represented the current state of the painting, while others confused the oft-reproduced Ganay version, which had sold at auction in 1999, with it.⁵⁸ The ‘spooky’ painting was on panel or canvas, for sale or slated for auction, found in a jumble sale or an estate clearance, a remarkable discovery or, for one distinguished critic, ‘launched in a sophisticated marketing operation’.⁵⁹

A response was required, and so within a few days we issued a press release summarizing the ownership, critical, and conservation history of the painting, together with a photograph of it after its recent treatment. Each scholar who had studied the painting at the National Gallery in 2008 was asked to confirm his or her support for the attribution to Leonardo in writing; all did. At the same time the National Gallery confirmed the painting’s inclusion in the forthcoming exhibition. Speculation concerning the ‘new’ Leonardo continued throughout the summer, accompanied by a wealth of media coverage, which only added to the anticipation of what would have been, even without the Salvator Mundi, one of the greatest art exhibitions mounted in modern times. In advance of the opening, The Sunday Times Magazine was granted special access to those involved in the discovery and research, publishing a well-informed account (Fig. 1.22).⁶⁰

Fig. 1.22Sunday Times Magazine, cover, 9 October 2011.

In preparation for its travel to London, George Bisacca created a microclimate environment for the panel behind non-reflective glass and a new flexible ‘featherboard’ support that gently held the panel inside the frame. In November, I served as courier for the painting, not hand-carried as it had been in 2008, but extraordinarily packed, prepared, and crated for transport utilizing the most sophisticated shipping techniques for rare and delicate works of art. The painting arrived at the National Gallery as an honoured guest and was examined by conservator Larry Keith, in the company of Luke Syson and myself—the condition check recorded as part of the film Leonardo Live, which documented the opening of the exhibition on 8 November and the preparations for it (Fig. 1.23).

Fig. 1.23Condition check from Leonardo Live. Conservator Larry Keith examines Salvator Mundi, observed by Luke Syson and Robert Simon.

The Salvator Mundi was hung between the two preparatory drawings from the Royal Library at Windsor in the final gallery of the exhibition (Fig. 1.24). The catalogue entry authored by Luke Syson presented salient and significant information about the painting, but the constraints of the format necessarily prevented a fuller examination of this major addition to our knowledge of Leonardo’s career. Discussion of the painting, both in the popular and scholarly press, then began in earnest, and will no doubt continue for years to come.⁶¹ The present volume is the first to treat the painting monographically. As such it should provide both an important reference and an invitation to further consideration of this most spectacular art discovery of our time.

Fig. 1.24Installation view of the Salvator Mundi and two preparatory drawings from the Royal Collection at the National Gallery.