January 2009

Explore the Apollo archive

Look back over two vibrant years of Apollo: browse every issue from January 2006 to the present day.

During the 1890s and early 1900s the American industrialist P.A.B. Widener purchased 93 paintings, primarily Old Masters, from the Dutch-born art dealer Leo Nardus, who operated out of New York and Paris.1 According to two dossiers of unpublished correspondence in the archives of the Philadelphia Museum of Art,2 Widener began to suspect that there was something wrong with these pictures in late 1907, when he noticed that the three Vermeers he had acquired from Nardus were not included in Hofstede de Groot’s newly completed catalogue raisonné (Fig. 3-5).3

In the wake of this discovery, a succession of experts – including De Groot, Bernard Berenson, Roger Fry and Wilhelm Valentiner – paid visits to Widener’s home outside Philadelphia to assess the ‘Nardus pictures’. The experts informed Widener that, while a few of these paintings were true master-pieces, the vast majority were copies, fakes, or comparatively minor works incorrectly attributed to great artists such as Vermeer, Velázquez, Rembrandt, Turner, Botticelli, Raphael and Leonardo da Vinci. Upon hearing the news, Widener reportedly became ‘frightfully enraged’ and fulminated that he had been ‘swindled in the most shameful manner’.4 The resulting turmoil would yield important changes not just for the Widener collection and the career of the mysterious Leo Nardus, but also for the standards and practices of the American art market.

Nardus is seldom mentioned in the annals of art history, although, from his arrival in New York in 1894 until his downfall in 1908, he was an extremely active and successful dealer in the United States.5 Fluent in four languages, a champion swordsman and an internationally known chess master, Nardus was a highly charismatic individual, and he used his charm to devastating effect in his business dealings. He was able to sell unexceptional Italian and Flemish primitives to American millionaires for very high prices – sometimes through the mail – simply by employing such seductive phrases as, ‘It is difficult to explain these pictures to you in every detail, but I have the conviction that they belong to the highest works that art has produced’.6 As for provenance, all that Nardus generally offered was a vague disclaimer such as, ‘The pictures mentioned here, have been chosen by me, out of a famous Collection in Italy.’

Contemporary accounts of the American Old Master market during this relatively early period in its development suggest that adventurers of the Nardus type were not unusual. Indeed, they were a source of sardonic pride in certain corners of the European art world. ‘Avant tout, tant pis pour les Yankees’, as the critic Paul Eudel put it in his classic Trucs et Truqueurs (1907).7 ‘They have simply to learn more or else hire better experts.’ By virtually any standard, however, Nardus constituted a breed apart from other disreputable operators of the era. His clients included not only Widener, but also the influential Philadelphia corporation lawyer John G. Johnson; the financier J.P. Morgan; the textile magnate M.C.D. Borden; Sir William C. Van Horne, president of the Canadian Pacific Railway; and other prominent individuals.

In establishing these valuable social connections, Nardus possessed an important credential that most art-world impostors did not: he appeared to be associated with a perfectly reputable European gallery. Nardus, who was the son of a prosperous Utrecht antique dealer, had served his apprenticeship in the art trade with the Paris branch of Bourgeois Frères,8 and circumstantial evidence suggests that he may, at least initially, have been dispatched to America in the firm’s service. Nardus sold only a small number of genuine masterworks in the United States, and the majority can be traced back to Bourgeois Frères. Although the consistent success Nardus enjoyed in selling minor paintings as landmarks of Western art was a complex phenomenon, it is likely that his fine, ex-Bourgeois pictures – Rembrandt’s Lucretia among them – helped to establish credibility for his less plausible offerings.

Vague hints of Widener’s difficulties with Nardus are scattered in the early-20th-century literature on Dutch art and connoisseurship, although Nardus’s name is never mentioned. In his Alt-Holländische Bilder of 1918, for example, Wilhelm Martin cites Widener’s early forays into the art market as a cautionary tale for aspiring collectors.9 Likewise, in a footnote to his Levensbericht van Dr. C. Hofstede de Groot (1931), H.E. van Gelder mentions De Groot visiting Philadelphia in 1908, at Widener’s behest, and condemning the majority of Widener’s pictures.10

The matter is also taken up, in an elliptical way, in the memoirs of Widener’s grandson, who admitted that the elder Widener had been ‘taken many times’ by unscrupulous dealers, but maintained that the offending parties had always been happy to take back the ‘bum merchandise’ when asked to do so.11

The Philadelphia correspondence tells a pointedly different story. P.A.B. Widener did not, in fact, adopt a relaxed attitude towards being cheated, and Nardus felt no conciliatory impulse when caught. Perhaps the most striking item, in this regard, is a letter from Widener to Nardus dated 5 October 1908, in which Widener announced his intention to sue:

I had hoped that you would save yourself from the very serious consequences that threaten by voluntarily endeavoring to do all in your power to make good the outrage you have perpetrated on me. [...] I do not intend that you shall carry away and continue to enjoy the money that you have obtained from me under gross false pretenses.12

Widener’s assessment of Nardus’s character was quite blunt. ‘There is not the slightest good in your trying to brazen out the matter by acting and talking as if it was a question about which experts might differ’, Widener wrote, leaving little room for a forgive-and-forget resolution.

After charging what would be excessive prices for genuine pictures, you supplied me with false ones. Some of the pictures which I obtained from you were exchanged with dealers in Europe, who have condemned the same in the most unreserved manner and have offered to sell back to me the pictures at one twentieth of what I paid for them. All these pictures were falsely attributed. A few were old pictures, wrongly labeled. Many of them were modern imitations.

The anger that Widener expressed in this letter was apparently all the more volcanic for having been so long delayed. Initially, when he learned that his three Vermeers had been omitted from Hofstede de Groot’s catalogue, his response had been one of mild annoyance – not at Nardus’s perfidy, but rather at De Groot’s evident carelessness. Widener simply arranged to send word to De Groot requesting that the three ex-Nardus Vermeers be included in future editions of the catalogue.13 However, when De Groot failed to accede to this request, Widener grew more concerned. According to a letter by a third party, dated 6 May 1908, Widener came to France to sort things out:

When I arrived to take lunch with Mr. W., Mr. Nardus was there, and it seems they had in the morning quite a discussion about pedigree, etc.

Mr. W. was much disturbed and I took him out for a drive with Harry [Widener’s grandson] to give him some distraction; before he left, he arranged with Mr. Nardus to meet Hofstede de Groot who is in Paris, to explain to him about his Vermeers of Delft about which he is writing a book.14

The proposed meeting between Nardus and De Groot did indeed take place, and De Groot agreed to suspend judgment on the Vermeers until he had seen them in person.15 Given De Groot’s incomparable knowledge of the art market, it is possible that he already suspected what he would find in Pennsylvania before making the trip. P.A.B. Widener had published a catalogue of his collection in 1900 – two folio-sized volumes bound in red leather. Of the 250 copies printed, most were distributed to friends and family, so that the catalogue itself was seen by few people outside Widener’s immediate circle. The illustrations, however, had been engraved by Goupil in Paris, and it seems that word of Widener’s indiscretions began to circulate among the French dealers. Although Widener was unaware of the gossip, his sons George and Joseph evidently heard snippets of it. Just prior to De Groot’s intervention, there were reports that the two brothers – both middle-aged men – had become ‘increasingly suspicious’.16

Once De Groot arrived at Lynnewood Hall – Widener’s imposing 110-room residence, just north of Philadelphia – and saw the full extent of Nardus’s handiwork, he promptly resolved to convene a panel of experts to conduct a comprehensive enquiry.17 Bernard Berenson was called in to assess Nardus’s Italian paintings; Roger Fry, the English ones. Photographs of the Flemish primitives were submitted to Max Friedländer and Wilhelm von Bode in Berlin. With regard to the Dutch works, the young Wilhelm Valentiner was on hand to confirm De Groot’s own findings. Privy to these deliberations, John G. Johnson (Fig. 6) – Widener’s friend, neighbour, personal attorney and fellow victim of Nardus – would later note that all the experts stood united in their opinions and that the things he over- heard at Lynnewood Hall could only be characterised as ‘shocking’.18 ‘I am mortified that it was through my recommendations that Mr. Widener bought the pictures,’ Johnson remarked.19 ‘My belief in Nardus was a blind one and I saw only with his eyes... I thought Nardus was a great expert and honest. Honest he certainly was not.’

Widener was not the only one of Johnson’s friends who had done business with Nardus. Once De Groot understood the outlines of the story, he dispatched Wilhelm Valentiner to survey the damage. One of Valentiner’s first stops was Montreal, where he called on Sir William C. Van Horne, who was apparently somewhat confused by the visit. ‘I am naturally much interested in what Dr. Valentiner says about the Rogier van der Weyden,’ Van Horne wrote to De Groot at Lynnewood Hall. ‘I am however unable to understand what he means by “totally wrong”. Surely, he cannot mean anything like a modern copy.’20

Valentiner also investigated the relatively small New York collection of J.P. Morgan – who then kept the vast majority of his pictures in London, loath to pay the 20% American import duty. (Congress would abolish the tax in 1909, largely at Morgan’s insistence.) Valentiner discovered that Morgan owned a questionable Holbein whose execution resembled that of an equally questionable Dürer belonging to Widener.21 Hofstede de Groot’s archive in the Hague contains a curt letter from Morgan’s private secretary saying, ‘Mr. Morgan would like to see you on the subject at his library, 33 East 36th Street, NY City, some day next week after Monday’.22 The result of this conversation is not known. Nardus’s grandson (who lives today in the south of France) has, however, confirmed that Morgan was one of Nardus’s clients.23

It took years to uncover all of Nardus’s misdeeds, with most of the detective work being done by Valentiner – who would stay on in America as a curator at the Metropolitan Museum of Art. For instance, in preparing a catalogue of the private collection of the textile millionaire M.C.D. Borden in 1910, Valentiner discovered that Borden owned six Nardus fakes, including a fraudulent Rembrandt self-portrait.24 He expunged these works from the catalogue, but the story has an intriguing footnote. Aside from the fakes, Nardus had also sold Borden a masterpiece, which had come from Bourgeois, namely Rembrandt’s Lucretia (Fig. 10).25 Neither Valentiner nor anyone else ever entertained any doubts about Lucretia, but Nardus’s involvement with it – not a major selling point after the Lynnewood Hall incident – was prudently left unmentioned when Borden’s estate sold the painting at auction in 1913.26 The National Gallery of Art in Washington, which now owns Lucretia, has no record that it ever belonged to Nardus – a man whose name has been written into history with the point of an eraser.

The legal resolution of the Nardus-Widener entanglement was complex, but ultimately anti-climactic. Nardus was not one to get bogged down in details, and dealing with angry clients was a task he preferred to delegate to his long-time friend and business associate Michel van Gelder. Van Gelder was a somewhat more respectable figure than Nardus himself, although not, apparently, a person above making veiled threats. Declaring that a trial would be ‘very annoying for both parties’27 and could prompt ‘a great deal of mischief’, Van Gelder proceeded to negotiate a settlement that gave Widener only very limited restitution – and if it hadn’t been for John G. Johnson, Widener might not have obtained even that.

Johnson was acutely aware that public exposure of his and Widener’s ‘foolishness’ might lead to embarrassment and ridicule. Although Widener declared that he would ‘have satisfaction no matter what it shall cost’, the cool-headed Johnson – who had taken almost as big a financial trouncing as Widener - advocated that they instead ‘bear the loss in silence’.28 Gallantly, Johnson agreed to forego reparations for himself if Van Gelder would at least make some gesture towards placating Widener.29

With this generous offer on the table, Van Gelder convinced Nardus to take back 12 of Widener’s pictures – ascribed to Old Masters but apparently modern – and, in exchange, Nardus gave Widener a dozen replacement pictures of better quality.

Of the works thus received, Widener chose to keep only two – the Rembrandtesque panels Head of Saint Matthew and Head of an Aged Woman, now in the National Gallery of Art in Washington (Figs. 8 and 9) – which Nardus had obtained from a reputable dealer.30 For reasons that are not entirely clear, Widener immediately auctioned off the rest. Unfortunately, the sale – held in Amsterdam on 30 June 1909 – did not attract the expected interest from the trade. The star attraction, a Rembrandt tronie (character head) fetched only $5,200, and four of the lots failed to sell at all.31 Nonetheless, for a man of Widener’s wealth, the amount of money generated was probably less important than the principle involved in getting some form of recompense.

For his part, Nardus remained unrepentant. When he took delivery of the returned fakes, he wrote to Widener saying, ‘At last I have received the pictures, and now, after I have had a fresh look at them, I have to state loyally that all of those pictures are correct’.32 The Turner ‘made a big impression’ on him. The Cuyp he loved more than ever. As for the Hobbema, he was shocked that anyone could have doubted it. ‘Here, it is unnecessary, to go into any details’, he wrote. ‘Simply believe me that it is really a beautiful work of Hobbema and that I guarantee it genuine, and you can mention in my name to the man who said it to be a copy or something like it that he is an ignorant of pictures’. Summing up, Nardus bid Widener an aggrieved adieu: ‘I see now very clear in this affair, and I frankly say that it is very sad that you listen to the arguments of men who state such aberrations.

I repeat, you can enjoy your collection and you can be proud of it.’ As far as the available records show, this was the last that Widener ever heard from Nardus.33

Change being more often evolutionary than revolutionary, it is perilous to speak of watershed moments in the history of any field of study. Nonetheless, given the sheer size of the fraud Nardus perpetrated in America, one may well wonder to what extent the practices that came to characterise the highest end of the us art market in the post-1909 era were a response to the abuses of Nardus and others like him. For instance, the increasingly important role of bona fide experts in the decision-making of American collectors seems to have received direct encouragement from the Nardus matter – a notion borne out by the experiences of Nardus’s sometime business partner, Michel van Gelder.

Van Gelder came out of the Widener negotiations with his reputation more or less intact, and tried, for a few years, to go on operating much as Nardus had, not realising that the rules of the game had changed. Like Nardus, he always claimed to be a gentleman, not a merchant. ‘I am not dealing in pictures,’ he once told Johnson, ‘but simply ceding to you some works of art which I know you love and which help to enrich your collection.’ After what happened at Lynnewood Hall, however, this line of reasoning was no longer effective. If Van Gelder wanted to sell Johnson, for instance, a newly-discovered masterpiece by Carpaccio, he would have to show it first to Bernard Berenson for approval. In a letter to Johnson from 1909, Van Gelder shows how the process worked:

While in Florence, I met with Mr. Berenson who told me that he has received a letter in which you wrote about the Carpaccio. I was quite astonished to hear that he would not decidedly say that it was by the master but very near! [...] As a result of this I have now decided to make absolutely sure not to have any more disagreement in that respect.

Any Italian picture which I may show you in future shall be known to B. through a photo or otherwise and will be accompanied by a written assurance on the photo.35

Over time, Berenson’s ‘written assurance on the photo’ – or certificate of expertise – became indispensable in the marketplace, as no major American collector was likely to buy an Italian painting without one. To a limited extent, the certificate system was already known in Europe – certainly through Bode’s work with the Sedelmeyer firm – but in America it would evolve into a fixed principle in the art collector’s catechism: Old Masters had to come with a signed opinion from an expert in the field. Although certificates were hardly a foolproof solution to the manifold problems of authenticity and value, they did at least help cut down on the worst abuses.

As a case in point, Van Gelder’s so-called Carpaccio did not meet with approval: Berenson would not attribute it, and so Johnson returned it.36

Given that the pre-eminent position that Joseph Duveen attained in the American art market derived, in no small part, from his semi-secret partnership with Berenson, the much-studied rise of Duveen, likewise, takes on a subtly different appearance when seen in the light of the Nardus swindles. Under a contract struck in 1912 – not publicly disclosed at the time – Berenson received a 10% commission (later 25%) on all works he attributed for Duveen, as well as a handsome retainer from the firm.37 Today this relationship is sometimes viewed disapprovingly from a moral standpoint, but it is worthwhile to note that Berenson and Duveen colluded to make money selling genuine masterpieces, not false ones à la Nardus. In an atmosphere where collectors had the recent example of Widener before them, the authority – and, in most cases, the rightness – of Berenson’s attributions carried so much weight that Duveen quite reasonably sought to monopolise Berenson and shut out the competition, an aggressive business manoeuvre by any measure, but a shrewd one nonetheless.

Of course, even the art world’s most renowned authorities make mistakes. Interestingly, while all the experts involved in the events of 1908 knew immediately that Nardus had been selling works under false colours, even these truly great scholars

had trouble sorting out some of the subtle questions of connoisseurship raised by Nardus’s most vexing pictures. It was obvious, for instance, that the ‘Leonardo’ (Fig. 1) sold to Johnson by Nardus and Van Gelder was wildly misattributed, but a small ‘Rembrandt’ (Fig. 11) in Johnson’s possession, by contrast, gave De Groot pause for thought. He seems to have been inclined initially to dismiss this Old Man at Prayer along with most of Nardus’s other Dutch pictures, but, then, in a reversal of opinion, he accepted it as authentic.38

Some months later, visiting Van Gelder in Belgium at the insistence of Franz Kleinberger, De Groot was surprised to discover that Van Gelder owned a second version of the same subject (Fig. 12), but, curiously, declared it ‘no less genuine’.39 One wonders if he might have altered his opinion had he known that Nardus owned yet a third variation on the same image (Fig. 13).40 This was, moreover, not the only instance in which works developed twins after passing through Nardus’s hands.41 In time, this strange phenomenon led to the inevitable dark speculations.

There were persistent rumours that Nardus – a gifted portraitist on top of all his other accomplishments – was an active forger of Old Masters.42 Although Nardus’s grandson disputes that notion vehemently, it is clear that Nardus did sometimes sell forgeries. He sold several to Johnson, and many more to Widener, including, according to De Groot’s notes, a ‘Rembrandt’ Standard Bearer for $95,000 43 – a sum equal to roughly 200 times the annual income of the average American household around 1900. And, whether or not Nardus engaged in forgery himself, he and Van Gelder did employ a reputed art forger as their primary picture restorer in the Netherlands.44 The forger in question, Theo van Wijngaarden – seen here in a portrait (Fig. 14) by Leo Nardus – was later a mentor to the infamous Han van Meegeren, a chain of connections suggesting that art fraud, like other fields of artistic endeavour, has its own schools, masters, traditions and lineages.45

In the end it mattered little to Nardus’s American victims whether they were duped by modern forgeries or merely by minor pictures with grossly inflated attributions: the treachery and deceit were distressing all the same. Johnson, a sensitive and very decent man, would long regret having introduced Nardus and Van Gelder to his circle of friends. In particular, he felt guilty ‘for the ruin done to Mr. Widener’.46

As it turned out, however, Widener neither wanted nor needed anyone’s pity. Having witnessed first-hand the power of the world’s real art experts, he sold most of his ‘Nardus pictures’47 and then put Valentiner, Berenson, and De Groot to work. With

their advice and counsel, Widener, in just the eight years of life remaining to him, assembled one of the finest collections in the world – probably the most dramatic turnaround in the history of collecting. In 1911, he even acquired a genuine Vermeer, Woman Holding a Balance – from Knoedler – thus wiping away some of the painful memory of the three ex-Nardus ‘Vermeers’ that had directly triggered the day of reckoning at Lynnewood Hall.48

In 1918, three years after Widener’s death, the famed art dealer René Gimpel visited Philadelphia to meet Joseph Widener. ‘He and Frick have the most beautiful collections in the United States’, Gimpel noted in his diary, impressed by the array of Rembrandts, Turners and Van Dycks. He hastened to add, however, that ‘des véritables escrocs’ – ‘some real crooks’ – ‘had tricked P.A.B. Widener at the beginning and sold him fake paintings for as much as a hundred thousand dollars a piece.’49 Word had got around.

Jonathan Lopez has written a biography of Han van Meegeren, The Man Who Made Vermeers, forthcoming from Harcourt.

Print this article   |  Email to a friend |  Subscription offers | Comment