Hans Holbein the Younger 1497/8–1543
NG 2475 
Christina of Denmark, Duchess of Milan

Susan Foister, 2024

Extracted from:
Susan Foister; with Rachel Billinge, Marika Spring and Lea Viehweger; and contributions by Lorne Campbell and Allison Goudie, The German Paintings before 1800 (London: National Gallery Global and Yale University Press, 2024).

Oil on wood (oak, identified), 180.2 × 82.8 cm, painted surface 178.5 × 82.8 cm

Provenance

Since Holbein is documented as taking the portrait of Christina (1522–1590) for Henry VIII at Brussels in 1538 (see The Commission), the earliest reference to the portrait is presumably that in the inventory of Henry VIII in 1542.¹ It is doubtless the same portrait that is listed again at Westminster, after the king’s death in 1547, in the charge of Sir Anthony Denny, Keeper of the Palace of Westminster, as ‘Item a greate Table with the picture of the Duchyes of Myllayne beinge her whole stature’.² A note made in February 1551, during the reign of Edward VI, on an inventory of property that had been the responsibility of Sir Anthony Denny recorded that the portrait was given by the king (Henry VIII?) to Henry FitzAlan, 19th Earl of Arundel (1512–1580): ‘This table Therle of Arundell hathe of the kinges maiesties gift as he saithe’.³ Arundel may have passed the portrait directly to his nephew John, 1st Baron Lumley (1533–1609), who undoubtedly possessed it by about 1590 (see further below).⁴ However, the portrait may rather next have come into the possession of William Herbert, 1st Earl of Pembroke (1501?–1770), since an inventory of his possessions at Baynard’s Castle in 1561/2 lists: ‘The picture of the Duches of Lorayne’. This might provide corroboration for the assertion of Holbein’s biographer Van Mander that the painter Federico Zuccaro (about 1542–1609, known to have been in England for six months in 1574/5) saw a portrait apparently of Christina by Holbein at the Earl of Pembroke’s house, described as ‘a portrait of a certain Countess, dressed in black satin, life size, a full length figure, unusually pretty and well painted by Holbein’.⁵ The National Gallery portrait formerly bore the characteristic cartellino of works in the collection of John, 1st Lord Lumley (fig. 1; see Technical Notes)⁶ and is by this means securely identifiable with the portrait listed in the Lumley inventory of around 1590 as ‘The Statuary of the Duches of Myllayne, afterwards Duches of Lorreyn daughter to Christierne King of Denmarke doone by Haunce Hollbyn’.⁷ It probably passed with other Lumley pictures to the collection of Thomas Howard, 1st Earl of Arundel.⁸ It appears to be the portrait sketched by Anthony van Dyck (1599–1641) with others in Arundel’s collection when he was first in England in 1621 (fig. 2).⁹ The artist and writer Joachim von Sandrart (1606–1688) records seeing a life‐size portrait of a countess by Holbein, which must be that of Christina in Arundel’s collection in 1627 (and formerly in Pembroke’s).¹⁰ After Arundel’s death in 1646 the portrait is recorded in the inventory of his collection made at Amsterdam in 1655 as ‘Duchessa de Lorena grande del naturale’.¹¹ These pictures were claimed by Arundel’s only surviving son, William, Lord Stafford (1614–1680). The portrait is next recorded by George Vertue (1684–1756) in 1720 at Stafford House (Tart Hall), London, the home of Arundel’s grandson, Henry Stafford‐Howard, 1st Earl of Stafford (1648–1719).¹² It was evidently inherited by Stafford’s cousin, Henry Charles Howard of Greystoke (died 1730), for it was seen by Vertue in about 1744 at the house in Soho Square, London of his son Charles Howard (1720–1786, from 1771 the 10th Duke of Norfolk).¹³ According to Vertue, the portrait was removed to Greystoke Castle (Cumbria) but by 1768 it was at Worksop Manor (Nottinghamshire), both houses owned by the Dukes of Norfolk.¹⁴ The picture was cleaned, repaired and varnished [page 441] [page 442] when it was at Worksop in 1774: a bill for the work carried out on this and other pictures by Messrs Hayman & Pugh records that the pictures were all ‘in a very bad condition and greatly mildewed and the painting scaling off which made it very difficult to restore’.¹⁵ It was still at Worksop Manor in 1813.¹⁶ The house was sold by Bernard Howard, 12th Duke of Norfolk (1765–1842), in 1839, and it was probably then that the painting was transferred to another seat of the Dukes of Norfolk, Arundel Castle,¹⁷ where it was seen by Waagen in 1854.¹⁸ From 1880 until 1909 it was on loan to the National Gallery (except when lent to exhibitions; see below). In 1908 the 15th Duke of Norfolk informed the trustees that he was being offered substantial sums for the picture but gave an undertaking to first offer the portrait to the Gallery for the same price; on 22 April 1909 he told the trustees he had been offered £61,000 for the picture, giving them until 1 May to find the matching sum. As they were unable to do so, the portrait was sold to Colnaghi’s, who offered it to the American collector Henry Clay Frick (1849–1919) for £72,000 on condition that the National Gallery would be allowed to try to match the sum by 1 June. At the last minute it was bought by the National Art‐Collections Fund with the aid of an anonymous donation of £40,000, supplementing a Treasury grant for £10,000 and other contributions, and presented to the National Gallery in 1909 (for details of the sale and subsequent acquisition see Appendix).¹⁹

Copies and Versions

• (1) Sketch in pen and ink by Anthony van Dyck; present collection unknown (fig. 2).²⁰

• (2) Unknown artist, half‐length; Royal Collection (RCIN 403449), oil on panel, 44 × 32 cm (fig. 3).²¹

• (3) Unknown artist, three‐quarter‐length; National Trust, Dunham Massey (inv. 932355), oil on panel, 101.6 × 72.4 cm.²²

• (4) Unknown artist, three‐quarter‐length; Florence, Uffizi (inv. 1890 n. 4219), oil on canvas, 88 × 68 cm.²³

• (5) Otto Haslund (1842–1917), full‐length, including the Lumley cartellino, 1892; Frederiksborg, Denmark, Nationalhistoriske Museum (inv. A603; size not given).²⁴

• (6) Full‐length, inscribed ‘Katherine, 4th wife of Charles Brandon Duke of Suffolk, Mrs R[aymond]’, ffennell sale, London, 5 October 1956, lot 28.

• (7) Full‐length, 181.6 × 82.5 cm, anon. sale, Sotheby’s, 12 April 1961, lot 132.

• (8) Full‐length, Mrs Hewitt, Worthing.²⁵

• (9) Full‐length, Mr Willmett, Trewsbury House, Gloucestershire.²⁶

Exhibitions and Loans

RA 1880 (177); London 1890 (92); London 1894 (4); Manchester 1897 (47); on long‐term loan to Cheltenham Art Gallery and Museum, 3 June 1915–5 December 1918 ; on loan to National Art‐Collections Fund, London, 27 May–[page 443]17 June 1928 ; Picture of the Month, National Gallery, 3–30 November 1943 ; NG 1945–6 (24); NG 1975 (20); NG 1995; NG 1997–8 (30); Tate Britain 2006–7 (157); NG 2014b.

Technical Notes

The painting is quite badly worn, especially in the shadows of the face and hands, the dress and the background, and has been extensively retouched. There are losses associated with the movement of panel joins and old splits but elsewhere, despite a history of flaking paint, there are only a few small isolated losses.

The portrait was recorded as having been cleaned and repaired in 1774 (see Provenance). In 1968, after a careful study to establish the status of the paint on the background, the picture was cleaned, removing several layers of overpaint and the Lumley cartellino, and restored. In 1989 the support was treated to secure the joins, losses were then filled and retouched.

The support is an oak panel consisting of four boards with vertical grain, 1.2 cm thick, butt joined and aligned with dowels; three of the boards are 25–30 cm wide, but that on the extreme right is only 2.6 cm wide. The overall width of the support is 82.5 cm at the top edge and 82.8 cm at the bottom. The oak originates from the Baltic/Polish region. Only the middle board could be dated using dendrochronology; the youngest heartwood was formed in 1512.²⁷ The back of the panel has been roughly finished using hand tools and the panel seems to be its original thickness. The panel is in good condition apart from some areas of woodworm damage.

The panel was prepared with a ground of chalk (calcium carbonate) in glue.²⁸ This process took place with an engaged frame in place: there is a tiny fragment of the wood from the frame near the top edge of the panel. A raised barb, formed when the ground continuing onto the frame was broken during frame removal, is present at the top of the panel, and there is also a strip of unpainted wood across the bottom, but the ground continues to both side edges. There is a thin pale grey priming on top of the ground.

Underdrawing, consisting of simple outlines, is visible in infrared images (fig. 4). The drawing is careful and closely followed for the head, but freer with some changes around the hands and gloves (figs 5, 6), and a line across the floor suggests that the skirt was first planned to be longer.

Infrared reflectography also shows that the black clothes were painted leaving approximate reserves for the fur, hands, gloves and cuffs but the arrangement of the cuffs and gloves was changed during painting, so parts of the gloves are painted over black. There are also slight changes to the outer contours of the skirts. Short, curved black marks, apparently made at random in the background, may be evidence of the artist wiping his brush to modify the amount of paint on it as he worked.

Examination of paint samples was undertaken in 1950 and 1967 and was restricted to establishing the status of the various layers of paint on the background at that time, [page 444] which, having been found not to be original, were removed in the 1968 treatment.²⁹

The original paint of the blue background was found to contain azurite, of high quality and large particle size. The effect of the shadows was created by using a brownish‐grey modelled underpaint beneath the azurite.

Examination of the surface with the microscope reveals that the other pigments appear to be as used in other portraits by Holbein: black of small particle size (possibly lamp black), lead white, various earth pigments, lead‐tin yellow, vermilion and red lake.

The stone of the ring is painted with vermilion and red lake, the setting has raised highlights containing lead‐tin yellow, and there are traces of powdered gold used as paint (shell gold; figs 12, 13).

Holbein blended his colours wet‐in‐wet and feathered the brushstrokes to soften transitions (figs 8, 9, 11).

Reports by A.E. Werner (1950) and Joyce Plesters (1967) state that the medium is linseed oil but there is no indication of how this was discovered.

The Sitter

The 16‐year‐old Christina is shown wearing black robes and a black cap, evidently the Italian‐style mourning dress she had adopted as her regular garb following the death of her first husband, Francesco II Maria Sforza, Duke of Milan (1495–1535).³⁰ John Hutton, Henry VIII’s ambassador in Brussels, referred in a report of 9 December 1537 to her wearing of ‘moorrnyng aparell after the maner of Ytalie’.³¹ Christina has brown eyes and full, red lips; her hair is concealed by her black cap; this has a grey lining and is trimmed at the back with rows of braid, set one above the other, radiating outwards (fig. 10). Her black robe is probably of silk taffeta, a lustred fine twilled silk used for mourning, its sleeves full below the shoulder, tight below the elbow.³² The robe is lined or trimmed with brown fur (probably marten or sable, fig. 11) and appears to be fastened with a black satin ribbon tied into a bow. Under her robe she wears a high‐necked black gown, trimmed with red on the collar, with the frills of a white chemise edged with black visible at the neck and wrists (fig. 7). She holds a pair of buff‐coloured leather gloves and wears a single ring with a red stone on the third finger of her left hand.

Christina was said to have been tall: John Hutton described her in his report of 9 December as of the age of 16 yeres, very high of stature for that age. She is highar than the Regent, a goodly personage of boddy, and compytent off beawtie, of favor excellent, sofft of speche, and very gentill in countenance. She werythe moorrnyng aparell after the maner of Ytalie … She resembllythe myche Mystris Shelton, that somtyme watid in Court uppon Queyn Anne. She ussithe most to spek French, albeit that as it is reportid she can Ytalian and High Almeyn³³ He wrote in addition of the dimples evident when she smiled: [page 445] [page 446] Ther is non in theis parties off parsonage, beawtie, and byrthe lyke unto the Duches off Myllayn. She is not soo pewre whyt, as was the late Qweyn, whois soal God pardon; but she hathe a syngular good countenaunce, and when she chancesithe to smyl, ther aperithe two pittes in hir cheikes, and wone in hyr chyne, the wiche becommythe hyr right excellently well³⁴ Hutton reported on 21 February 1538 that Christina enjoyed hunting, spoke French and ‘in hir spekyng she lispithe, wiche dothe nothing mysbecom hir’.³⁵ Another Englishman, Thomas Wriothesley (1505–1550), Hutton’s successor as Henry VIII’s ambassador in Brussels, wrote on 1 February 1539: Very pure, fair of colour she is not, but a mervelous good brownishe face she hathe, with faire redd lippes, and ruddy chekes; and onneles I be deceyved in my judgement, which in all thinges, but specially in this kynde of judgement, is very basse, she was yet never soo wel paynted, but her lyvely visage dothe muche excel her poincture.³⁶ She was also reported to have been known for her exceptionally beautiful hands.³⁷

Christina was the younger daughter of Christian II of Denmark (1481–1559) and his queen, Isabella of Austria (1501–1526), sister of the Emperor Charles V, and was born in November 1521. She is shown aged four with her elder brother, John, and elder sister, Dorothea, in Gossaert’s portrait of about 1526 (fig. 14); after their father was deposed as king in 1523 the family left Denmark to live in exile at the court of their Habsburg relative Margaret of Austria, aunt to Isabella. In September 1533, aged 11, Christina was married by proxy to Francesco II Maria Sforza, Duke of Milan; she was widowed just two years afterwards. Christina continued to live in Brussels at the court of her aunt, Mary of Hungary (1505–1558), Regent of the Netherlands, and in 1538–9 negotiations for her marriage to Henry VIII took place but came to nothing. No objection seems to have been made to this by Christina herself: in February 1539, asked by Thomas Wriothesley about her own inclinations, she ‘blusshed excedingly, and said: “As for myn inclination”, quod she, “what shuld I saye? You know that I am at thEmperrurs commaundement”, and again, “You knowe I am thEmperours poore servaunt.”’³⁸ Soon afterwards Henry VIII’s relations with her uncle the Emperor Charles V deteriorated to the extent of preparation for war and the marriage negotiations were abandoned. In 1541 Christina instead married François, duc de Bar (1517–1545), who became duc de Lorraine in 1544 and died the following year. Christina then became Regent of Lorraine but was banished from the duchy of Lorraine in 1552, when it was occupied by the French. In 1554 she was again residing in Brussels, when it was proposed that she might marry Emmanuel Philibert, Duke of Savoy (1528–1580). She went to England in 1555 to meet him but again the marriage did not proceed. She returned to Lorraine in 1559 but finally in 1579 settled in Tortona, Piedmont, part of the Duchy of Milan, which had formed part of [page 447] her marriage settlement. Christina died at Alessandria in Lombardy on 10 August 1590.³⁹

In addition to the portrait of Christina as a child by Gossaert mentioned above, and her portrait by Holbein, her portrait was taken on several further occasions. A payment for a full‐length portrait by Bernard van Orley is recorded in the accounts of Mary of Hungary in 1533, when Christina was 11, presumably in connection with her marriage to the Duke of Milan.⁴⁰ Ambassador John Hutton mentioned on 21 February 1538 a portrait of Christina that had been commissioned at the Brussels court by the ‘Ladie Marquis of Barrough’: she said, yf I had seyn hir owt of hir morning apparel, so gorgeously as she had seyn hir the day before, I wold have marveillid, for she said, to tell me in secret, she cawssid hir piktire to be made, wiche being fenisshed, the Duches had promissid to give it unto hir, soo that she of her own motion said, assone as it cam to her hands I shuld have a sight therof.⁴¹ Hutton also mentions in a letter of 14 March to Thomas Cromwell a portrait he was to have sent to England on 9 March but which he judged unsatisfactory (see below).⁴²

A portrait of Christina by Titian is recorded in Mary of Hungary’s inventory.⁴³ There is a drawing after a lost portrait of Christina in the Recueil d’Arras, a collection of drawn copies of portraits that probably dates from 1567.⁴⁴ A portrait signed by Michiel Coxcie (died 1592) of 1545, depicting a woman identified as Christina, is in the Allen Memorial Museum at Oberlin, USA.⁴⁵ A portrait of Christina by Anthonis Mor (1516/21– about 1576) painted in 1554, when she was to marry Emmanuel Philibert, is now in the Royal Collection (fig. 15).⁴⁶ Another group of portraits, the work of French artists, appear to be connected with the negotiations for the peace of Cateau‐Cambresis of 1559: a drawing in the British Museum by François Clouet (about 1515–1572), a painted portrait dated 1558, another similar one and a miniature in the Uffizi.⁴⁷ A portrait with the date 1589 attributed to an anonymous Flemish painter in the collection of the Prado shows Christina with her daughter‐in‐law, Claude of France, Duchess of Lorraine (1547–1575), and granddaughter Christine of Lorraine, Grand Duchess of Tuscany (about 1571–1637).⁴⁸

The Commission

The circumstances surrounding the commission of NG 2475 are the best documented of Holbein’s career. The artist was sent to Brussels in March 1538 to portray Christina of Denmark in order to obtain the likeness of a potential wife, a task he was to repeat several times with other subjects until Henry’s marriage to Anne of Cleves – portrayed by Holbein in 1539 – was fixed upon. A portrait by another artist had already been despatched to the king by English ambassador John Hutton on 9 March, when Holbein arrived in Brussels on 10 March. Hutton recounts in his letter of 14 March to Cromwell: havyng the day before [Holbein’s arrival] sent wone of my sarvandes towards youre Lordshipe with a picture of the Duches of Myllain, I thought it very necessary to stay the same, for that in my opinion it was not soo perffight as the cawse requyrid, neither as the said Mr Haunce coold make it. Uppon wiche determination I dispatched another of my servaundes, in post, to returne the same, wich your Lordshipe shall receve by this berrar.⁴⁹ The earlier portrait was considered unfit for purpose: ‘the other is but sloberid in comparison to it, as by the sight of bothe your Lordshipe shall well aperceve’; although Hutton had despatched it he had sent a servant to return it.⁵⁰ The prospect of comparison between this image and Holbein’s, however, implies that the poorly made portrait would remain until Holbein’s arrival in London on 18 March.

Holbein, identified in the documents as ‘Mr Hanns’, left England, accompanied by the courtier Sir Philip Hoby, on 2 or 3 March and arrived in Brussels on 10 March. Hoby had detailed instructions supplied to him by Thomas Cromwell on how to proceed, advising him to beg Christina ‘to take the pain to sit that a servant of the King, who is come thither for that purpose, may take her physiognamy; and shall ask when Mr Hanns shall come to her to do so’.⁵¹ Hutton wrote in a letter to Cromwell on 14 March that he had commended Holbein to Christina as ‘a man very excellent in makyng off phisnaymies’ and had succeeded in obtaining an appointment for him to take her portrait. On [page 448] 12 March at one o’clock in the afternoon he was conducted to the duchess’s presence, and Christina sat to Holbein, ‘who having but thre owers space hathe shoid hym self to be master of that siens, for it is very perffight’.⁵² After Hoby and Holbein had departed, Mary of Hungary wrote to Eustace Chapuys (about 1490–1556), the Imperial Ambassador in London, of Henry VIII’s interest in a marriage to Christina, and informed him that Cromwell had requested a man be allowed ‘to take her portrait in order to show it to the King and give him greater desire to see her. This I have allowed, and the man has actually returned to England with the portrait, well satisfied with the personal appearance and manners of my said niece’.⁵³ Chapuys recorded the king’s delight in the likeness Holbein had produced in a letter written to Mary of Hungary from London on 23 March 1538, five days after the artist’s return to England: on the very same day, the 18^th, the painter sent by this King to Flanders came back with the Duchess’ likeness, which, I am told, has singularly pleased the King, so much so that, since he saw it he has been in much better humor [sic] than he ever was, making musicians play on their instruments all day long.⁵⁴ No information survives documenting Holbein’s normal methods of proceeding with a commissioned portrait, and [page [449]] in any case the circumstances of this one were exceptional. His usual method was to make use of preparatory drawings, in most cases for the heads and shoulders alone; the outlines of the drawings were transferred to the panel to preserve the proportions of the sitters’ features exactly as originally observed, although Holbein might then make small modifications.⁵⁵ It can be presumed that he would have made one or more drawings of Christina in her presence during his three‐hour sitting. As this is the sole instance of a documented sitting with Holbein, it is not known if three hours was the normal span he expected for such a portrait; but since in this case there was to be no possibility of future access to the sitter, he may have wanted to make further drawings, for example of Christina’s hands (such as those he made in 1523 for the portraits of Erasmus, now in the Louvre) and of her dress (see, for example, his drawing of the dress of an Englishwoman seen from both front and back, now in the British Museum).⁵⁶ It would have been easy for Holbein to return to London carrying more than one drawing but it is highly unlikely that he would have set out to cross the Channel with a panel the size of the National Gallery portrait of Christina, and he would not have needed to do so. The likeness Henry saw on Holbein’s return, and which pleased him, is therefore likely to have been established in one or more drawings.

It has been suggested that Holbein also referred to a smaller, head‐and‐shoulders portrait of Christina in mourning dress now in the Royal Collection, which was evidently owned by Henry VIII in 1542 (fig. 3; see Copies and Versions), but this appears to be a version made after Holbein’s portrait or studies for it. There are some differences in the rings worn: in the Royal Collection portrait Christina wears three rings, one with an emperor cameo, whereas in NG 2475 she wears only one, different ring.⁵⁷ However, since Holbein was expressly sent to Brussels to take her portrait from the life, it is unlikely that he would have needed to refer to any other portraits of Christina.

The full‐length format was apparently thought desirable in portraits of sitters who were candidates for marriage with the kings of England. Not only are there a number of references to Henry VIII’s desire to see as much of his potential brides as possible, with no defect covered, it is recorded that in 1446 Henry VI specified that portraits of his potential French brides should be full length.⁵⁸ Christina’s mourning costume obviously precludes any sight of her limbs, as would any other sixteenth‐century dress, but Holbein has ensured that what is visible – face and hands – are clearly presented against the black clothing (figs 16, 17). He has also suppressed any clear sense of the space Christina occupies: the train of her dress blurs the line where wall and floor meet, and the blue‐green background implies outdoor rather than inner space, though [page 450] this is belied by the way in which her shadow is cast against the background on the left. With the strong vertical shadow to the right perhaps implying an entrance nearby, the effect is to suggest a subtle sense of movement as though Christina sways forward towards the viewer.⁵⁹

Attribution and Date

The painting is unsigned and undated but its provenance and outstanding quality have led to its universal acceptance as the portrait commissioned from Holbein by Henry VIII in 1538. It formerly bore the characteristic cartellino used for works in the collection of John, Lord Lumley and is identified as the work of Holbein in his inventory of about 1590. It seems reasonable to conclude that this is the portrait painted by Holbein on his return from Brussels in March 1538 and retained in the king’s collection despite the failure of the marriage negotiations.

Appendix: The Acquisition of the Portrait in 1909⁶⁰

By 1908 Holbein’s portrait of Christina of Denmark, owned by Henry Fitzalan Howard, 15th Duke of Norfolk (1847–1917), had been on loan to the National Gallery for 28 years (although the loan had originally been seen as a temporary measure while the duke was carrying out renovations to Arundel Castle). The Gallery already owned one Holbein, The Ambassadors (NG 1314), which had been purchased in 1890 from the 5th Earl of Radnor with the assistance of a number of donors (see p. 30). The period had seen the sale of numerous works of art in aristocratic collections as a result of the flexibility given by the 1882 Settled Land Act, which allowed the sale of land and chattels.⁶¹ In this context it was perhaps not entirely surprising that the Director of the National Gallery, Sir Charles Holroyd (1861–1917), was concerned about the potential sale of the Holbein portrait: in a letter to Holroyd of 12 February 1908 the Duke of Norfolk explained that he had had several enquiries in the past ten years as to whether he would sell it, but ‘as I knew it sometimes happened that a price was offered for a picture far in excess of its real value, I was not prepared to say that it was impossible to conceive of an offer which might tempt me to sell’.⁶² In the letter the duke also discouraged the notion of a public subscription, which he saw as the only means the Gallery would have of raising money to buy the painting in the improbable event of agreement on a price. At the National Gallery board meeting on 10 March 1908 the duke’s letter was read out, and the director also reported on his conversation with the duke at Norfolk House on 14 February, from which he concluded that the duke ‘might feel it his duty for public object to sell’.⁶³ He mentioned that Mr Pierpont Morgan had previously promised not to compete for the picture, and indeed to contribute to its purchase for the nation, and that he had informed the duke that Dr Bode of the Berlin Museum would not purchase the painting if the National Gallery were offered it, ‘as he considered we had acquired a right in it’; however, he would take it if the Gallery did not, as ‘he would rather have it go to Berlin than to America’.⁶⁴ Holroyd reported that the duke had been persuaded to say that he would be inclined to accept the sum of £50,000 for the painting (having already been offered £60,000, he baulked at only £40,000, which Holroyd thought he might be able to raise): ‘I gathered that if we would get promise of help from the government and private subscription he would consider an offer of £50,000’.⁶⁵ The duke agreed that should he have another offer he would try to let the Gallery know and would also try to grant a delay of a month, but he would be unwilling to risk the loss of a purchase in such a case.⁶⁶ The trustees agreed to permit the director to obtain the promise of donations in case the painting came to market.⁶⁷ Holroyd’s memorandum gives some insight into the Duke’s motivation for a possible sale: ‘if an extravagant price was offered for it he would not refuse to sell it as he felt he could do so much good with the money and that such large sums were needed for educational purposes and for dealing with the poverty around us’. It also reveals his own strong feeling that the painting ‘ought not to leave on any account’: it was one of his list of 30 such works in Britain. In a concluding passage deleted from the memorandum he wrote: ‘I am sure there will be quite an outcry if the picture goes and the Duke himself will be blamed as well as the Trustees and Director’.⁶⁸

The following year Holroyd’s worst fears were realised: a bidding war for the picture ensued, pitting Duveen Brothers against the firms of Colnaghi and Knoedler. William McKay for Colnaghi wrote to their bank on 13 March saying he had seen ‘a letter addressed to a friend in which the Duke said he had an offer of £39,000 but that he would consider £60,000’.⁶⁹ On 23 March the antique dealer Charles Davis wrote to the duke to say he had ‘a friend’ (presumably Duveen) who was prepared to offer £50,000 for the picture; the Duke refused, stating the sum needed was £60,000, and the Gallery must also be offered the painting.⁷⁰ The offer was confirmed in another letter of 13 April.⁷¹ On 18 April Otto Gutekunst of Colnaghi wrote to Charles Carstairs at Knoedler, evidently reporting on this exchange and informing him that Joseph Duveen, who told him £39,000 had already been offered by ‘somebody’ (whom he wrongly assumed to be Colnaghi), had therefore offered the duke £50,000 (after advice from the art historian Bernard Berenson).⁷² With the knowledge that the duke was standing firm on the need for £60,000 to secure the painting, on 21 April 1909, trumping Duveen, Otto Gutekunst offered the winning sum of £61,000 on behalf of Colnaghi, to which the duke agreed.⁷³ The following day, 22 April, the duke sent a letter to the Gallery stating that he had sold the Holbein for £61,000 but was prepared to accept the Gallery as a purchaser on the same terms. He informed them: ‘I last night came to an agreement with a purchaser whose name I am not at liberty to mention’. He gave the trustees only nine days, until 1 May, to find a matching sum in order to purchase the painting.⁷⁴

The events of this period were reported at the meeting of the trustees on 11 May: the Duke of Norfolk had spoken with the Secretary to the Board and to the trustee Lord Carlisle on 22 April. Lord Carlisle did not see that anything could be done in the absence of the director, who was in Portugal, returning on 28 April, and that in any case it would be impossible to raise £61,000 in this period.⁷⁵ On his return Holroyd reported that [page 451] he tried to see Lord Balcarres, the chairman of the National Art‐Collections Fund (NACF), which had been founded six years before, in 1903. However, he was unable to do so until the Friday, owing to the debate on the budget, and the option expired on the Saturday.⁷⁶ Following Colnaghi’s acquisition of the painting, Charles Carstairs of Knoedler immediately agreed the sale of the painting to Henry Clay Frick for £72,000 ($360,000), which, with the inclusion of the £11,000 dealers’ profit, then became the sum the NACF had to match.⁷⁷ It should be noted that in his letter of 22 April setting out his offer to the Gallery, the duke had specified that if the Gallery did not accept the offer to purchase the painting for £61,000 by 1 May, the subsequent purchaser should again offer it to the Gallery voluntarily but that there was no attempt to bind the purchaser to any specific price for this second offer, hence the sum to match now became £72,000. Colnaghi set a further deadline of 1 June, one month hence, and pledged £2,000 to help the Gallery’s fundraising efforts.⁷⁸

A public appeal for £72,000 was immediately launched by the NACF. Following a letter written by the Fund that appeared in the Morning Post, Daily Telegraph and The Times, the sale was extensively debated in the press.⁷⁹ The letter ran as follows: Sir – The committee of the National Art‐Collections Fund feel it their duty to make a strong appeal to all lovers of art in favour of the preservation of this celebrated picture for this country … The sum is large; the moment is not the most favourable, but the committee nevertheless hopes that, although the time is short, contributions will be forthcoming from men of wealth and from others according to their means, sufficient to save the country from the great loss that is threatened. The Times wrote on 1 May 1909: This is, of course, the very worst conceivable moment for the question to have arisen for a Government with a deficit of 16 million cannot think of buying pictures at this price, and all the rich men are hit so hard by the Budget that to raise £60,000 for the purchase even of a masterpiece will be a very difficult task.⁸⁰ The Times further reported on 5 May 1909 that the Duke of Norfolk had ‘informed the National Gallery that he might at any moment receive an offer which, in view of the anxious financial future for all landowners, he might feel obliged to accept for the reduction of his liabilities’. This contradicts what is known of the duke’s motives from Holroyd’s memorandum of their meeting, cited above, and ascribes to him the motivation of a pressed estate owner rather than his avowed motive of funding the alleviation of poverty. It has also been suggested that the sale was ‘a calculated move on the Duke’s part to dramatize the consequences of threatening the landed classes’.⁸¹

The Morning Chronicle responded with disapproval of the sale: The Duke of Norfolk belongs to a family which has given many great names to English history. He above all therefore, should understand that the ordinary laws of private property do not apply to his picture by Holbein which has been a national asset. An ordinary millionaire who has made money by gambling on a large scale may be excused if he puts one of the world’s masterpieces into the public market, using the ordinary tricks of the trade to secure a fancy price. But that the premier Duke of England should so far forget his family pride as to expect the highest possible price from his country under the threat of selling his picture to some American or Jewish millionaire abroad is almost incredible and certainly shameful.⁸² The National Gallery was accused of holding the country to ransom and some commentators thought the money would be better spent on battleships.⁸³

Letters to The Times denounced the Duke of Norfolk and Colnaghi as well as the buyer, who was assumed (correctly) to be an American. Philip Burne‐Jones warned in The Times of 3 May 1909 that if the panel painting should ‘find a home in America … its days are practically numbered … No painting can survive many years in the overheated atmosphere of American rooms or galleries’.⁸⁴ The fears of loss to Americans were considerable, and encapsulated in the famous cartoon of Uncle Sam attempting to drag a terrified Christina from the frame of the painting (fig. 18).

The press debate intensified: the Daily Graphic of 4 May 1909 termed the response to the potential loss of the painting a collective ‘condition of mourning’ in Britain, with aesthetes holding daily vigils at the National Gallery.⁸⁵ It appears that some journalists were surprised to learn that the Gallery did not own the painting outright.⁸⁶ Others expressed hostility to foreigners: Colnaghi was seen as a foreign firm and anti‐German sentiment came to the fore in hostility to Gutekunst as well as to the Berlin Museum; it was also noted that Holbein was a Bavarian.⁸⁷ Countering this, the NACF stressed the relationship of artists and subject to the court of Henry VIII: in an article in the Saturday Review of 8 May, D.S. McColl asserted that ‘The Christina, woven into the texture of our English history and into our memories of near thirty years at the National Gallery, is irreplaceable’.⁸⁸

The story had therefore already been broken and much debated in the press, and the campaign to save the painting from its American fate begun, when, at the board meeting of 11 May 1909, the letter dated 22 April from the Duke of Norfolk was read. As already mentioned, this stated that he had sold the Holbein for £61,000 but was prepared to accept the Gallery as a purchaser on the same terms, with nine days, until 1 May, to find a matching sum in order to purchase the painting.⁸⁹ With this date now behind them, the trustees focused on the new price and new deadline: a letter dated 6 May 1909 from Colnaghi offered the picture to the Gallery for £72,000 and set the date for their response as 31 May.⁹⁰ Despite the fact that the trustee Alfred de Rothschild (1842–1918) – a serial objector to the purchase of paintings – had in 1897 expressed his opinion that the Holbein was ‘unworthy’ of the collection, the trustees were able to agree as required that the attempt to acquire the painting should proceed.⁹¹

A letter dated 5 May 1909 from the Chancellor of the Exchequer, David Lloyd George (1863–1945), was read to the same meeting, stating that he had discussed the matter with the Cabinet and that the government would provide a sum not exceeding £10,000 in the event of the public appeal needing additional funds to make up the balance for the purchase.⁹² This contribution had in fact been swiftly organised by Holroyd, who reported to the trustees that he had had the opportunity of speaking with Lloyd George at the Royal Academy banquet concerning the need for government support.⁹³ An offer from Francis Howard (1874–1954) to hold an exhibition of old masters at the Grafton Galleries for two to three months to raise funds was also discussed: he pointed out that the Galleries’ Fair Women exhibition took over £8,000 but the period proposed would have been beyond the deadline of 1 June.⁹⁴

A public subscription was now being organised through the NACF. Early in May Holroyd had attended the annual meeting of the NACF, where he was able to announce the promise of government support and assist in organising the public subscription. The Daily Graphic of 6 May 1909 reported that Holroyd swore that if Mary Tudor had Calais written on her heart, he had Holbein written on his.⁹⁵ His plea was also reported by the Connoisseur: ‘The pathetic appeal made by Sir Charles Holroyd at the annual meeting of the fund will never be forgotten by those who were privileged to be present’.⁹⁶ The public campaign started to take effect: the Gallery’s records include a letter from Hawes Turner to the NACF, enclosing £1 towards the Holbein fund from a Miss B.N.⁹⁷ There were a few dissenters; one editor argued against retaining the painting: It is not as though the picture speaks any lessons of patriotism, raises any historic memories of which we are proud, nor inspires any exaltation of morals. It is a portrait of a woman, executed by a foreigner long dead, and of no conceivable advantage to the nation except as a museum specimen of mastery in technical art.⁹⁸ The Sheffield Independent published a series of unflattering comments by Gallery visitors on Christina’s lack of physical charm, with one man observing: ‘Wot an old puddin’‐face’.⁹⁹ By contrast, the critic Claude Phillips (1846–1924) wrote in the Daily Telegraph on 22 May: let us not give the world the humiliating spectacle of a mighty and enormously wealthy nation allowing one of the greatest treasures in its National Gallery to be torn from its wall, while it stands as motionless, helpless onlookers, muttering curses not loud but deep, as Wotan the contemptible fallen Jupiter does when the all‐powerful Giants, flouting him to his very face, tear away Freia, the life‐giver, the goddess of beauty and of love.¹⁰⁰ At the National Gallery board meeting of 25 May 1909 Holroyd reported that he had been informed by the NACF that the total amount raised fell short of £10,000.¹⁰¹ He had agreed with Colnaghi a 12‐hour extension of the Gallery’s deadline to Tuesday 1 June, as 31 May was a Bank Holiday, but he felt that there was little chance of success.¹⁰² By Friday 28 May the NACF had raised only £32,000.¹⁰³ But an unexpected and dramatic last‐minute anonymous donation of £40,000 from a lady writing from a German spa ensured that the painting stayed in the National Gallery. On 1 June Carstairs of Knoedler in London had to tell Frick he had lost the picture: ‘Nation retain Holbein balance money subscribed today regret can’t be yours’. Frick immediately cabled back: ‘consider have been trifled with. No authority extend option which expired yesterday.’¹⁰⁴

Sir Isidore Spielmann (1854–1925), one of the honorary secretaries of the NACF, an engineer who was prominently involved in international exhibitions, wrote vividly of the saving of the painting in his memoir of 1924: We were becoming increasingly despondent and unhappy, and felt that only a miracle could save us…. On Tuesday morning June 1 something did turn up … Among the letters was one written on Sunday May 30 (Whit Sunday) from a lady at a health resort in Germany … I read the letter again and again and wondered if we were being hoaxed or if the lady was [page 453] mentally sound. ‘Holroyd, the picture is saved’ I said. ‘Cheer up my dear fellow; the picture is ours’, said my colleague. He looked at us from one to another as if dazed, and then followed the most pathetic incident in the whole of these proceedings. He buried his face in his hands and burst into tears.¹⁰⁵ The donor’s identity was to be kept a secret in perpetuity, and there are few clues to her identity.¹⁰⁶ Spielmann, writing in October 1924, noted that she had by then passed away.¹⁰⁷ It was much later asserted by the Earl of Crawford (1871–1940) that the cheque constituted a third of her whole estate.¹⁰⁸

A week afterwards there was astonishment that the donor was a woman, a ‘patriotic lady’.¹⁰⁹ Henry James (1843–1916) wrote to Edmund Gosse (1849–1928): ‘The Holbein Duchess has been saved – by a veiled lady who has bought her off for £40,000. Can you lift the veil?’¹¹⁰ Speculations on the identity of the donor included Lady Tate and Lady Harriet Wantage. Punch in the summer of 1909 satirised irate letters of denial from Britain’s wealthiest heiresses. One suggestion was that the donor was the American Consuelo Vanderbilt, Duchess of Marlborough, fortifying her social position but attributing her support to her ‘love of art’.¹¹¹ Rosalind Howard, Lady Carlisle (1845–1921), was the only one of the women suggested who did not deny the rumour.¹¹² It has been observed that women and noted feminists played a prominent role in the subscription drive to meet the price of £72,000. The Lady commented: but for her unparalleled generosity, the ‘Duchess’ would have gone to America. The presence of women at the Chemistry Conference, and the appearance of this anonymous lady coincide, and emphasize the lesson that everywhere women are taking an active part in public work. Only the other day they received what is known as the ‘commercial’ vote in Italy. But this is dangerous ground.¹¹³ Others were less supportive: Vanity Fair said that the money was wasted, since Britons could see their own live duchesses for free and ‘to any patriot they are quite as good. Why, then, pay £72,000 to see dead duchesses. It is all nonsense’.¹¹⁴ Punch included a fictive letter from the novelist Marie Corelli (1855–1924) stating that she could have given £40,00 but it was ‘of no interest to me to provide the nation with pictures at which ignoramuses and toads are free to look’.¹¹⁵ Cartoons depicting the saving of the picture included one in the Westminster Gazette for 7 June 1909, with Sir Francis Carruthers Gould showing Christina clinging to John Bull asserting ‘I ne‐ver – will – desert – mr Bull’ and ‘something HAS turned up’ (fig. 19). Another, in the Western Mail, depicted ‘American Covetousness’ with the donor as a faceless man climbing out of Atlantic with the painting, rushing to meet John Bull, who exclaims ‘Saved’.¹¹⁶

The Times published a list of subscribers on 14 June 1909, which included the painter John Singer Sargent (1856–1925), writer Edmund Gosse, art historian Robert C. Witt (1872–1952, secretary to the NACF) and the playwright George Bernard Shaw (1856–1950). The partners of Colnaghi gave £2,000, Edward VII 100 guineas and the Prince of Wales £125; a message from the Prince and Princess of Wales was published in The Times on 1 July 1909.¹¹⁷ The Duke of Norfolk himself made an anonymous donation of £5,000.¹¹⁸

At the National Gallery board meeting of 22 June 1909 a letter dated 3 June 1909 from the NACF addressed to the Treasury was read, announcing the purchase of the Holbein and requesting the government grant of £10,000.¹¹⁹ On 10 June 1909 the Gallery sent a letter to the NACF, in reply to theirs of 9 June, stating that the trustees had made a formal application for the £10,000 from the Treasury.¹²⁰ On 11 June 1909 the trustees received a letter from the Treasury requesting them to submit a Supplementary Estimate for £10,000 and enclosing a letter from the NACF regarding the purchase of Holbein’s Duchess of Milan.¹²¹ On 21 June the NACF acknowledged receipt of the £10,000.¹²²

On 9 November the NACF announced the completion of the purchase and the official presentation to the National Gallery. A letter dated 9 November 1909 from the NACF was read, formally presenting the picture to the nation. The director was to thank members of the NACF at an assembly to be held at the Grafton Galleries on 11 November.¹²³ On 10 November 1909 the trustees wrote to the NACF, in reply to their letter of 9 November, to express their gratitude.¹²⁴

The Burlington Magazine (edited by Charles Holmes, who would become the next Director of the National Gallery in 1916) noted: the crisis will at least have done the country one distinct service. It will remove definitely and finally any doubt which the public may have felt as to the reality [page 454]of the peril which threatens the important works of art remaining in private possession in England … no other masterpiece in the country, however great its owner, can henceforth be regarded as safe.¹²⁵ The episode was used as the basis for Henry James’s play The Outcry of 1909, which was made into a novel published in 1911.¹²⁶ James started work on the play in early autumn 1909, completing the first draft by 17 December. Plans to stage it the following summer were dashed by the death of King Edward VII, which closed all theatres. James therefore recast the play as a novel. The plot presents Lord Theign, who plans to sell a painting by Reynolds to an American collector, Mr Breckenridge Bender, which causes an outcry in the press. The character of the young connoisseur Hugh Crimble was based on James’s friend the novelist Hugh Walpole. He meanwhile discovers another painting in Lord Theign’s collection said originally to be a Moretto but possibly now a Mantovano worth £100,000, ten times as much as under the first attribution. James presented this painter as fictional but had to admit in a letter to Robert Witt: ‘I had never heard (in I am afraid my disgraceful ignorance) of the painter the two specimens of whom in the National Gallery you cite’.¹²⁷ In the novel Lord Theign desires to sell the Reynolds to pay one daughter’s gambling debts, while his other daughter is involved in a love triangle with Lord John and Hugh Crimble, who, despite the prospect of achieving both financial and reputational benefit from his attribution of the Moretto to Mantovano, which might allow him to marry Lord Theign’s daughter Lady Grace, is against its sale too. In the final chapter Lord Theign decides not to sell the Reynolds, tearing up a cheque for £10,000, and to give the Mantovano to the National Gallery. He addresses his friend Lady Sandgate, who herself owns a Lawrence: ‘“Will you then join me in setting the example of a great donation–?” “To the What‐do‐you‐call‐it?” she extravagantly smiled, “I call it,” he said with dignity, the “National Gallery.”’¹²⁸

Select Bibliography

Levey 1959, pp. 54–7; Rowlands 1985, no. 66, pp. 116–17 (with incorrect reference to cartellino); Foister 2004b, pp. 200–4, 266–7.

Notes