Anthony van Dyck’s arresting portraits of Jacob de Witte and Maria Nutius were painted in around 1628, at the beginning of the artist’s second Antwerp period. These beautifully preserved paintings are supreme examples of Van Dyck’s proficiency as a portrait painter, allowing us to appreciate his extraordinary talents at one of the most fertile moments in his career. Since the appearance of the paintings at the 1899 exhibition marking the tricentenary of the birth of the artist, they have been exhibited only five times, and just once in the last forty years. Since the nineteenth century, the portraits have hung among other masterpieces that formed the nucleus of two significant private collections of Old Master paintings, including that of South African mining magnate, Sir Joseph Benjamin Robinson (1840–1929). The somewhat dormant existence of the two paintings, whose history prior to 1899 is unknown, is visibly reflected in the works’ excellent state of preservation, as noted by Alfred Scharf following their display at the Royal Academy of Arts in 1958.¹ Recent restoration, in preparation for the present sale, reveals the remarkable condition of the paintings, allowing us to appreciate the sitters in all their glory for the first time in 120 years.

In spite of the near obscurity of these portraits, scholars have uncovered valuable information about the identity of the two sitters since their appearance in the exhibition of the Robinson Collection at the Royal Academy in 1958. Before this date, the pair had been referred to as ‘M. and Madame de Witte’, with little further investigation. However, the publication of Ludwig Burchard’s article in The Burlington Magazine following the exhibition, has enabled us to identify the sitters as Jacob de Witte (b. 1591; d. 23 November 1631) and his wife, Maria Nutius (b. 1601; d. 2 September 1661). Burchard based his identification on a commemorative monument in Antwerp Cathedral, erected to the memory of them both, and of their son Jacobus Anthonius de Witte.² Jacob de Witte was the son of Margarethe van der Heerstraeten and Adriaen de Witte, a prominent notary in the Antwerp community. Upon the death of his father in 1616, Jacob inherited his positions as notary and clerk to the Chamber of Orphans of the City of Antwerp (the body that administered orphans’ property), attorney to the Council of Brabant, and Squire of Buerstede Vekene. It is likely that the de Witte’s association with the Chamber of Orphans led to the acquaintance, and subsequent marriage, of Jacob and Maria. Maria Nutius (a Latinization of the Flemish ‘Nuyts’), was the daughter of Anna Tempelaers and Martin Nutius, a printer and bookseller in Antwerp. Both her parents died in 1608, when she was just seven years old, and it was Jacob’s father who drew up the inventory of their belongings. Adriaen had in fact known Martin Nutius for at least eleven years previously, having notarised his will in 1597. Maria would have been seventeen on the occasion of her marriage to Jacob de Witte on 21 January 1618 and assuming a date of around 1628 for the present portraits, is aged 27 as seen here.

It was originally thought that the paintings were conceived on the occasion of Jacob and Maria’s marriage, as is so often the case for male and female pendant portraits. Erik Larsen was of the opinion that the works dated to Van Dyck’s early years in Antwerp, around 1618–19, on account of the ‘sober style of Van Dyck at this period’.³ However, in the publication of the complete works in 2004, Horst Vey agreed with Nora de Poorter’s suggestion that a date at the beginning of Van Dyck’s ‘second Antwerp period’ is more convincing, a view with which Hans Vlieghe and Katlijne Van der Stighelen concur.⁴

According to the anonymous eighteenth-century Life of Van Dyck, the enquiring young artist left Antwerp for Italy on 3 October 1621 on a horse given to him by Rubens. He was back in Antwerp by July 1627, having enjoyed fruitful visits to the cities of Genoa, Rome, Florence, Bologna, Venice and Palermo. It appears that he stayed in Antwerp – aside from some local travel to other cities in the Low Countries – until his departure for England in 1632. Van Dyck’s return to Antwerp was bittersweet: shortly after his arrival, he suffered the death of his sister, and the effects of the war between the States-General and the Spanish Netherlands, following the expiry of the twelve years truce around the time of Van Dyck’s departure to Italy, continued to weigh heavily on the city. Despite the complicated political situation, painters continued to receive commissions for portraits as well as both religious and secular paintings. Van Dyck’s earlier success had not been forgotten, and Rubens’s increased responsibilities as diplomat to the Spanish Habsburgs meant that Van Dyck was able to emerge fully from his master’s shadow as the most skilled and sought-after portrait painter in the Southern Netherlands. By the end of 1628, he had already produced portraits of the Archduchess Isabella (1566–1633) and Geneviève d’Urfé (d. before 1656), lady-in-waiting to Queen Marie de’ Medici, and had been appointed court painter to the former on a salary of 250 guilders a year. His first recorded portrait from this period is of the wealthy cloth merchant, Peeter Stevens (c. 1590–1668), signed and dated 1627, shortly followed by a painting of his English wife, Anna Wake (1604–before 1669; figs 1 and 2).⁵ This pair provides a useful benchmark for dating the present works. Striking similarities are the placement of the sitters in the ‘wrong’ heraldic order – that is, the man on the right and the woman on the left, as the viewer sees it – their three-quarter-length format, plain backgrounds, and the heraldic symbols visible in the upper right and left corners. It is possible that Van Dyck’s portrait of Nicholas Lanier may also have been executed around this time (Kunsthistorisches Museum, Vienna), when the ‘Master of His Majesty’s Music’ was in Antwerp in the summer of 1628 with a part of the Mantua collection purchased for King Charles I.⁶

Van Dyck’s decision to reverse the ‘heraldic’ rule was not without precedent. In the sixteenth century, Maarten van Heemskerck had painted portraits of Pieter Bicker and his wife in this order, both dated 1529 (Rijsksmuseum, Amsterdam), and Anthonis Mor’s later portraits of Jean Lecocq and his wife from 1559 (Gemäldegalerie Alte Meister, Kassel) follow suit. Lorenzo Lotto also chooses to reverse the order, this time in a double portrait in his Portrait of Giovanni della Volta with his Wife and Children, 1547 (National Gallery, London).⁷ Van Dyck’s full-length portraits of an unknown man and woman with their children in the Louvre, dated to around 1628–29, depict the sitters in reverse, their architectural setting and full-length representation showing the recent effects of sixteenth-century Venetian painter, Titian, with whose work Van Dyck was by 1628 thoroughly familiar.⁸ Van Dyck’s full-length portraits of Philippe Le Roy and Marie de Raet in the Wallace Collection, painted in 1630 and 1631 respectively, also position the sitters in this way, with coats of arms visible in mirroring upper corners.

Unlike the aristocracy of Brussels, Van Dyck’s clientele in Antwerp appear to have preferred three-quarter-length portraits. It is curious that Van Dyck reverts to ‘closing up’ the backgrounds in these paintings; this may be a result of his prolific output at this time, focusing on the likeness itself rather than on accoutrements and elaborate settings, but more likely, however, is the connection of this format to Van Dyck’s Iconographie, a series of portraits of important male figures at the time, which he started to produce around 1627, on his return from Italy. Without a complete copy of the Iconographie preserved, it is difficult to establish a complete list of sitters. It has not been suggested anywhere that the portrait of Jacob de Witte could be related to an engraved portrait from this series, but it is worth mentioning that similar portraits, such as that of Peeter Stevens and Jan van Montfort, both reproduced in the Iconographie, follow the similar three-quarter-length format and sombre style.⁹ Could it be that the engravings in the Iconographie informed Van Dyck’s decision to reverse the traditional heraldic order? Once engraved, the male portraits would be mirrored, thus giving rhythm to Van Dyck’s ‘pantheon’ of nobility, military figures, scholars, artists and businessmen.

Jacob de Witte and Maria Nutius’s stiff and voluminous millstone ruffs were historically thought to support an earlier date for the paintings, as opposed to the softer Flemish ruff worn by Stevens. While these modern ruffs appear to have taken over from their earlier, more rigid counterparts, examples of those worn by Jacob and Maria are found in other portraits from this period, and Van Dyck’s confidence and sensitivity to the sitters’ features also support this later date.¹⁰ Maria is modestly dressed. She wears the vlieger, traditional dress worn by married women in the Low Countries, and simple adornments: pearl earrings and bracelets (similar to those worn by Anna Wake), and a wedding ring on the index finger of her right hand. Her cuffs are trimmed with Flemish bobbin lace and she holds a leafy branch in her right hand, perhaps from a rose bush, representative of the joys and sadness of love. It is unusual for Van Dyck not to have included a rose in bloom, however, and it has also been suggested that this could be an orange branch, symbolic of chastity and purity.¹¹

Both portraits are exemplary of Van Dyck’s virtuoso and efficient manner. In them, it is clear to see how years in Italy had inspired in him the sprezzatura, or ‘certain nonchalance’, of his Italian counterparts of previous generations. Van Dyck worked from loose sketches on a light grey ground, which he then built upon with just one or two layers of paint. Jacob de Witte looks directly out at the viewer, his confidence expressed by his steady and unfaltering stare. Giving the impression of a portrait that has been painted ad vivum, Van Dyck faithfully records de Witte’s pursed lips and dewy complexion. The soft folds of papery skin above his left eye and a receding hairline reveal his age of around 38 years. Tousled strands of hair catch the light that is reflected in his watery eyes, characteristically conveyed by Van Dyck with just a few tiny dots of white paint. Cleaning the painting (see below) has removed unnecessary strengthening in de Witte’s beard and moustache to reveal perfectly intact coarse blonde hair, the strokes of paint laid in with effortless precision. De Witte’s intense expression is enhanced by the overall simplicity of the composition. His left hand clasps his black cloak, his gold wedding ring glinting on his little finger, and the rigid ruff feels at odds with his smooth yet rumpled silk doublet. His costume is enlivened by small, regular slashes to the front, and by the intricate embroidery on his right sleeve. Van Dyck has left Jacob’s right hand and cuff unfinished, mirrored by the unfinished hand of Maria, which lacks any modelling and lighter tones. The schematic representation of Maria’s left arm, compared with the careful rendering of her right sleeve, effectively conveys a recession of space in the painting. Maria stands in a more frontal position than her husband, and her body is shifted further to the edge of the composition. Wispy strands of fine, dark hair frame her smooth, pale face, and her tentative smile reveals a certain fragility. Van Dyck layers the coloured glazes thinly onto the grey ground in certain areas of her right hand and face – notably around the bridge of her nose and eyes – suggesting her delicate and translucent complexion. Curiously, Van Dyck has chosen to light Maria from the upper left, while Jacob is lit from the upper right. However, the couple are unified by their plain costume and by the homogeneity of the coarse open brushstrokes used to fill in the background, indicating that the paintings were most likely executed in quick succession, one after the other.

The portrait of Jacob de Witte displays a visible coat of arms, upper left, showing the Witte Leverghem crest which includes a chevron argent between three seagulls, capped with crowns and a seagull with outstretched wings. A variation of the same crest is recognisable in the upper right corner of the portrait of Maria, however here it is combined with her own family emblem of a blue duck. The recent cleaning of the paintings has confirmed the crests to be contemporary, but not in Van Dyck’s hand. It is likely that they were added later, by someone else, either before or after leaving Van Dyck’s studio.

Note on Provenance

We know remarkably little about the paintings’ history in the centuries immediately following their completion, but in the light of their remarkably good state of preservation it seems more than likely that they remained in Antwerp in the family of the sitters. Their only recorded son was Jacobus Anthonius de Witte, Heere van Leverghem (1629–1688), who married Anna Catharina Malliaert, and they in turn produced one recorded son, Joannes Adrianus de Witte (1658–1719), who had one son: Joannes Gullielmus Nicolas Josephus de Witte (1695–1737).

The first concrete record of them is when Arnold de Pret Roose de Calesberg lent them to the Van Dyck exhibition in Antwerp in 1899. This is probably Arnold Ferdinand Charles de Pret Roose de Calesberg (Antwerp 1842–1902), in which case their sale to Robinson may have been occasioned by his death (but it cannot be ruled out that he was Gaston Arnold Honoré de Pret Roose de Calesberg [Antwerp 1839–1918 Reading]).

Their subsequent history is much clearer. Sir Joseph Robinson, unlike other Randlords who emigrated to South Africa, was born there in what is now the Eastern Cape at Craddock in the Cape Colony in 1840. He was a self-made man, initially trading in wool with the Boers, alongside whom he fought in the name of the Orange Free State, but subsequently acquiring diamond-bearing land in Kimberly and elsewhere, which formed the basis of his fortune. He became a great deal wealthier following the discovery of gold in Witwatersrand in the Transvaal, one of his mines producing 146,000 ounces of gold per annum by 1894. He was briefly in partnership with another Randlord collector, Alfred Beit, with whom he fell out, though less acrimoniously than he did with Cecil Rhodes.

Robinson’s collecting life was solely in London. In 1894 he took out a lease on Dudley House at 100 Park Lane, and most of his pictures, including the present pair, were acquired for the house over the following decade (fig. 3). Dudley House had been remodeled in 1855 by William Ward, 1st Earl of Dudley, to house his own collection, giving it an 81-foot Picture Gallery and considerably larger ballroom, where Robinson subsequently entertained, hosting performances by Dame Nellie Melba, among others. Dudley House was ideal for the display of Robinson’s pictures. In a relatively short span of well under a decade, and advised by the dealers Charles Davis and Sir George Donaldson, he assembled a significant collection of some 170 Old Masters, which included works of his own day such as Landseer, as well as a distinguished group of grand English portraits including works by Gainsborough and Reynolds, and Old Masters of the Italian, Dutch, Flemish and French Schools (the latter including four large canvases by François Boucher and his School, of which two are in the present sale as lots 25 and 26).

In 1910, shortly after securing a baronetcy from the Liberal government of the day, Robinson left London and returned to South Africa, and the collection went into storage. He instructed that the majority of the works be sold at Christie’s in 1923. He returned to London from Cape Town shortly beforehand, and in an incident notorious in the history of auctioneering, he ensured that the sale would be a disaster. He resolved to say goodbye to his pictures on the eve of the sale, and arriving at Christie’s in a wheelchair, he ‘fell in love with the collection he had not seen for so long’ and put high reserve prices on every picture in the hope that he would be able to keep them.¹² His tactic was successful, with only eleven of the 116 paintings sold. The remaining pictures, including the Van Dyck portraits, returned to storage and were eventually inherited by Robinson’s daughter Ida upon his death in 1929. It was under the supervision of Princess Ida Labia (she had in 1921 married Count Natale Labia [1877–1936], Italian Minister Plenipotentiary to the Union of South Africa) that the Van Dycks were lent to the Royal Academy and subsequently to the South African National Gallery in 1959, where they were described as being among the ‘peaks of excellence’ in Robinson’s collection.¹³ It was Princess Labia’s elder son who sold the paintings at Sotheby’s in London in 1976, when they were acquired for the present family.

Conservation

It was immediately apparent that these portraits appeared to be in remarkably good condition. They were covered in a thick and rather coarsely-applied layer of varnish which had discoloured considerably. On top of this were a very few retouchings to the beard and moustache of Jacob de Witte and the wispy detached hairs of Maria Nutius. These, easily visible under ultra-violet light, took the form of almost rhythmic hatched lines, as if the work of an engraver, and appeared to be unnecessary, perhaps occasioned by the dulling of the original brushwork due to the discoloured varnish. They most likely dated from the 1960s or 1970s and were in turn covered by a thinner layer of varnish. Removal of both layers of varnish and the retouchings by Etienne Van Vyve in Brussels in October 2021 indeed showed the latter to be completely superfluous, requiring no retouching at all. The unfinished left and right hands of Jacob de Witte and Maria Nutius had some limited overpaint, now removed, and a very few small areas of abrasion in the costume of De Witte required limited retouching.

For a full conservation report, please see here.

We are most grateful to Dr Hans Vlieghe and Dr Katlijne Van der Stighelen for inspecting the paintings first-hand, and for proposing a dating of both works to c. 1628, at the beginning of Van Dyck’s second Antwerp period.

¹ Scharf describes the works as ‘beautifully preserved’; see Scharf 1958, p. 303.

² The inscription on the monument reads: Deo Optimo Maximo/ Sepulchrum/ JACOBI DE WITTE/ Toparchae de Buerstede Vekene […]/ et/ MARIAE NUTIUS/ conjugum/ Praecessit ille 23 9bris 1631/ secuta est illa 2 7bris 1661/ …; see Inscriptions funéraires et monumentales de la Province d’Anvers, vol. I, Antwerp 1856, p. 299. Vey in Barnes, De Poorter, Miller and Vey 2004, under no. III.143, refers to the couple having two sons, although we have found no confirmation of this.

³ Larsen 1988, vol. II, p. 42, under no. 74.

⁴ Vey remarks: ‘These portraits have hitherto been assigned to Van Dyck’s first Antwerp period. However, as Nora De Poorter recognized, the handling points to the second Antwerp period. The collars do not conflict with the later dating.’ Vey also notes that a copy en buste of Jacob de Witte was formerly with C. Brunner, Paris (according to a photograph in the Witt Library, London), but it has not been possible to locate this work; see Vey in Barnes, De Poorter, Miller and Vey 2004, p. 363, under no. III.143. We are grateful to Dr Hans Vlieghe and Dr Katlijne Van der Stighelen for suggesting a date of c. 1628 for the portraits, based on first-hand inspection; verbal communication, October 2021.

⁵ Vey in Barnes, De Poorter, Millar and Vey 2004, p. 353, nos III.133 and III.134, reproduced pp. 354–55.

⁶ Inv. no. 501; https://www.khm.at/en/objectdb/detail/655/?lv=detail

⁷ Inv. no. NG1047; https://www.nationalgallery.org.uk/paintings/lorenzo-lotto-portrait-of-giovanni-della-volta-with-his-wife-and-children

⁸ Inv. nos 1242, 1243; see Vey in Vey Barnes, De Poorter, Millar and Vey 2004, p. 375, nos III.170, III.171. The male figure in this pair has a similar air to Jacob de Witte, and is described by Christopher White as ‘stiffly presented in an old-fashioned millstone collar’ and ‘with a sideward glance at the spectator’; see Christopher White, Anthony van Dyck and the Art of Portraiture, London 2021, p. 158.

⁹ For Jan van Montfort, see Vey in Barnes, De Poorter, Millar and Vey 2004, p. 335, no. III.109.

¹⁰ See, for example, the portraits of Adriaen Stevens and Maria Bosschaert, both dated 1629, and the portrait of Jan van Montfort of 1627–28; Vey in Barnes, De Poorter, Millar and Vey 2004, pp. 352–53, nos III.131 and III.132, and p. 335, no. III.109.

¹¹ The portrait of an unknown woman in the El Paso Museum of Art, Texas, includes a similar leafy branch without a bloom. The portrait of Marie de’ Medici in the Musée des Beaux-Arts, Bordeaux, shows a rose branch in bloom. For the former, see Vey in Barnes, De Poorter, Millar and Vey 2004, p. 132, no. 1.152; for the latter see Vey in Barnes, De Poorter, Millar and Vey 2004. p. 332, no. III.105.

¹² According to Waterhouse 1958, p. ix.

¹³ The Joseph Robinson Collection, exh. cat., National Gallery of South Africa, Cape Town 1959.