Lot 17
  • 17

SIR PETER PAUL RUBENS | Head of a young man wearing armour

2,500,000 - 3,500,000 GBP
3,135,000 GBP
bidding is closed


  • Sir Peter Paul Rubens
  • Head of a young man wearing armour
  • oil on oak panel


Probably Jean Nicolas Ribard, Rouen (1694–1758); Thence by descent to Jean Philippe Nicolas Ribard, Rouen (1724–1798);

Thence by descent until acquired by Heinz Kisters, Kreuzlingen, until 1958;

From whom acquired by Dr Günter and Anne Liese Henle, Duisberg;

From whose Estate sold (The Henle Collection), London, Sotheby's, 3–4 December 1997, lot 40;

When acquired by the present collector. 


Cologne, Wallraf-Richartz-Museum, Die Sammlung Henle: aus dem grossen Jahrhundert der niederländischen Malerei, 1964, no. 30; Greenwich, Bruce Museum; Berkeley, Berkeley Art Museum; Cincinnati, Cincinnati Museum of Art, Drawn by the Brush: Oil Sketches by Peter Paul Rubens, October 2004 – May 2005, no. 4.


J.B. Descamps, Catalogue et Estimation des Tableaux Appartenants au Feu Monsieur Jean Philippe Nicolas Ribard, 1798, MS, no. 17: 'Deux Têtes d'apôtres par Rubens'; H. Vey, Die Sammlung Henle: aus dem grossen Jahrhundert der niederländischen Malerei, exh. cat., 1964, no. 30, reproduced;

J.S. Held, The Oil Sketches of Peter Paul Rubens. A Critical Catalogue, 1980, vol. I, pp. 609–10, no. 443, vol. II, reproduced fig. 429;

M. Jaffé, Catalogo Completo. Rubens, 1989, p. 224, no. 406, reproduced; 

E. McGrath, Corpus Rubenianum Ludwig Burchard, part XIII: I. Subjects from History, 1997, vol. II, pp. 303, 307, note 47, reproduced vol. I, fig. 207;

M.E. W[ieseman[, in P. Sutton & M.E. Wieseman, Drawn by the Brush: Oil Sketches by Peter Paul Rubens, exhib. cat., New Haven and London 2004, pp. 98–101 (reproduced in reverse);

T.M[eganck] with H. D[ubois], in J. van der Auwera & S. van Sprang, Rubens.  a Genius at Work, exhib. cat., Brussels 2007, pp. 75-6, under no. 1, reproduced p. 76, fig. 4 (reproduced in reverse).

Catalogue Note

No painter before the nineteenth century used oil sketches as such an essential part of his working method, nor painted so many of them, as Rubens. He did not invent the oil sketch, but he is the artist with whom they are most closely associated. They perfectly suited his energetic creative process, although it should not be forgotten that he also made many drawings, and these were as integral a part of his working practice as his oil sketches. On the whole Rubens did not make drawings of heads when working up ideas for paintings and other projects, but relied on oil studies of heads, such as this one. Many of these were painted ad vivum, often of the same model seen from different angles, and kept for future use, while others, also based no doubt on models, were created with a particular figure in a painting or tapestry in mind, and it is not always easy to be sure which was which.  Rubens used this characteristically vivacious study for the head of the third warrior from the left in his massive painting of Saint Ambrosius of Milan barring Emperor Theodosius from entering the Cathedral in Milan, painted circa 1615–17, and now at the Kunsthistorisches Museum, Vienna (see fig. 1 and detail, fig. 2).1 The Vienna painting, probably in the collection of Frans von Imstenraet in 1678, and first recorded in the Imperial collections in the Stallburg in 1733 during the reign of Kaiser Karl VI, was until the 1960s thought by some scholars to be by Van Dyck, on account of the younger artist's sketch-like free copy in the National Gallery, London, generally dated circa 1617–18.2 While Van Dyck may well have worked on it during his tutelage with Rubens, Rubens’ authorship of it is no longer challenged.

A dating of circa 1614 or 1615 for the present study, slightly earlier than the Vienna painting, has been suggested on the basis of comparison with drawings probably by Willem Panneels in the so-called Cantoor, a large assembly of drawings after Rubens and other artists in his circle, now kept in the Print Room in Copenhagen, many of which record oil sketches of figures and of heads.3 These studies are likely to have remained in the studio as props throughout Rubens’ career, and the making of the Cantoor copies after them underscores their importance for Rubens, and reminds us that they were integral to his studio working practice.4 One of these sheets represents the same head as seen in the present sketch, along with a second view of the same model, in profil perdu, presumably taken from lost oil sketch (fig. 3). This secondary profile head was used by Rubens for a subsidiary figure at the extreme left of his full-scale Christ and the Adulterous Woman, datable circa 1614–1615, probably once kept in Rubens' house, and now in Brussels.5 The Brussels painting is significant for the model that served for the present sketch, since he is recognisable in the heads of other figures in the painting: the two youths near the column, above and behind all the other participants, looking forward, and in a head that appears in between the adulterous woman and one of her accusers, turned to his left.  Oil sketches of the same model seen from different angles probably existed, but they have not survived, and are not recorded in Cantoor drawings.

If we are to believe, as Julius Held and Justus Müller Hofstede did, that these sketches of head studies of the same youthful model were all painted at the same time, the present work must also have been created similarly a couple of years before its use in the Vienna painting.  The difficulty in a precise dating is due to the varying views on the dating of related paintings: the Brussels Christ and the Woman taken in Adultery for example has been dated as widely as circa 1610-15, although the most recent consensus puts it to the end of that period.  In any event it is likely to be at least a year or two earlier than the Vienna painting, and rather close in date to the Brussels one.   

Apart from the present painting, other oil sketches by Rubens connected with the finished painting are known. His study of the Head of an Old Man in Edinburgh was probably made in preparation for the figure of Saint Ambrose in the painting, and was not a sketch made earlier brought into use, because it shows the bearded Saint wearing a bishop's cope, albeit of different design to the one in the finished painting (see fig. 4), although Liz McGrath thought the cope was added later.6 Given that this is one of the two principal figures however, it is not surprising that Rubens would have made a specific sketch for it, and not drawn on an existing repertoire.

Another oil sketch related to the Vienna picture, for the head of the warrior to the extreme left of the painting, was sold at Sotheby's London in the same sale as the present sketch in 1997, though not from the Henle collection (fig. 5).7 In it, Rubens has concentrated on the curly hair of the older warrior, and has blocked in the face with a few peremptory strokes of the brush. As with other of Rubens’ figural sketches, that head was also used in another Rubens composition, also earlier than the Vienna painting: for the soldier in armour to the upper left of Rubens' Death of Seneca in Munich, which Elizabeth McGrath dates to circa 1615, and thus in line with the present sketch, but which others have placed earlier, to 1612–13.8

The present sketch was not used in any other work that we know of. While that might suggest that it, like the Edinburgh sketch for Saint Ambrose, was made in preparation for the Vienna painting, its energetic and free modelling and warmer tones, which differ from the Edinburgh sketch, indicate that it belongs to a slightly earlier period in Rubens’ career, circa 1612–15, and probably, as Betsy Wieseman and others have suggested, circa 1614-15.9 In adapting the fresh-faced youth for an armoured Roman soldier in the Vienna painting, Rubens has given him extended mutton-chop side whiskers, presumably to make him appear more martial: the two older soldiers nearer the viewer have full beards, one greying. It therefore seems likely that this sketch and the one for the nearer and older soldier were made at about the same time, and brought out together for the Vienna picture, a complex work requiring a wide repertoire of figures; so much so that as Arnout Balis pointed out, paintings like the St Ambrose ‘seem not just to make use of the available heads, but to be altogether conceived on the basis of them’.10 Both sketches are on panels of similar size (excluding a small strip added to the left edge of the present work), and both panels are of similar composition (see diagram, fig. 6). 

The argument that Van Dyck may have painted parts of the Vienna picture, perhaps subsidiary figures or their heads, and that in consequence this and the other head sketch sold in 1997 might conceivably be from his hand, needs to be addressed. Until late in his career, Rubens painted oil sketches exclusively on panels, often, like the present one, formed of different pieces of oak, whereas Van Dyck, even when in Rubens's workshop (and like Jacob Jordaens at the same time), almost always painted sketches on canvas or paper that were subsequently glued to panels, probably for sale. Furthermore, the present sketch shows none of the hallmarks of Van Dyck's style or technique, either around 1615–17, or at any other time, and is entirely characteristic of Rubens at a slightly earlier date, for example in the crimson lake used in the tear ducts and to build up the modelling of the cheeks (see Catherine Hassall's report summarised below). It is of course perfectly possible that Van Dyck was asked to contribute heads to the Vienna painting using Rubens’s head sketches, but there is little internal stylistic evidence to suggest this, and it would run rather counter to Van Dyck’s character, independent from an early age, even while working under Rubens’ direction. It is also worth noting that in Van Dyck’s free copy in London, the head corresponding to the present sketch, and the one to the extreme left, are very different in style to the Vienna prototype.

Many of Rubens’ oil sketches are constructed of several smaller pieces of oak, perhaps off-cuts, glued together to form the requisite rectangle. This is true, for example, of the present panel, as well as the study of the bearded man second from left that Rubens used not just in the Vienna painting but for numerous figures in paintings from about 1612–18. The sketch for the curly haired bearded centurion at the far left is another example (fig. 5), executed on a panel of very similar configuration to the present one; as the diagram of their formation shows (figs 6 and 7), the horizontal join occurs at almost exactly the same place and the two planks in each work are of near identical measurements, a further argument that they were painted at the same time.11 

Rubens’ sometimes bizarre but intricate panel constructions have long been understood as an idiosyncrasy of his work. Even very large landscapes painted much later in his career conform to this pattern or technique of construction: the famous Castle of Het Steen at the National Gallery London, is a confounding arrangement of some twenty separate planks of oak; the Rainbow Landscape in the Wallace Collection is constructed from nineteen planks; the Watering Place in the National Gallery from eleven; and so on.

When sold at Sotheby's in 1997 this picture included a hand resting on a staff in the lower right corner. As the catalogue entry noted then, this was a later addition, and has been removed in the interim. Another example of such an embellishment is the statuette and hand added by Rubens’ student Jan Boeckhorst to the Bearded man in profile formerly in the Schoenborn collection and sold in 2013.12 

An examination of the paint layers conducted by Catherine Hassall on 22nd May 2019 confirms that the ground layers of the strip of wood to the left and the main panel are the same, so the panel was not added to subsequent to the painting of the sketch.  She could find no separation between the ground and the layers of paint of the armour and the vermilion red used to outline the left shoulder before being covered with black, and certainly no varnish or dirt, which suggests that the armour was part of the original conception of the sketch, or added very shortly afterwards, although the underlying vermilion would be an odd choice for armour, and may indicate that Rubens envisaged a different costume at the outset.  the same paint was used for the grey underlay of the armour highlights and the first application of paint for the curved white highlight between the armour and the base of the neck of the boy.  Crimson lake was used for the shading around the eyes. 

1 Oil on canvas, 362 x 246 cm. Kunsthistorisches Museum, Inv. 524; see Jaffé 1989, no. 424; McGrath 1997, vol. II,  pp. 297–310, reproduced vol. I, figs 204, 208 and 210; and W. P[rohaska] in J. Kräftner et al. (ed.), Peter Paul Rubens 1577. The Masterpieces from the Viennese Collections, Vienna 2004, pp. 114–18, no. 25, reproduced.
2 National Gallery, London, Inv. 50; see McGrath 1997, vol. II, pp. 297–98, reproduced vol. I, fig. 205.
3 It was probably, but not certainly, Willem Paneels, pupil, steward and secretary, employed in Rubens's studio up to 1630, who drew copies of many of Rubens’ studies, some of them with the inscription ‘taken from Rubens’ cantoor’ (a sort of chest of drawers); the set of drawings have since become known as The Cantoor drawings, of which about two-thirds are attributable to one hand, probably Panneels.  The Cantoor includes drawings made after paintings, oil sketches and drawings made by Rubens over a span of decades, including for example many drawn copies after the antique that he drew in Italy, copies after more recent artists such as Raphael, and many studies of ecorché figures, but they cover virtually every aspect of Rubens' activity up to 1630.  For a detailed discussion of the Cantoor see P. Huvenne & I. Koeckelbergh (eds.), Rubens Cantoor, exhib. cat., Antwerp 1993.
4 An inventory at his death records ‘Une quantite des visages au vif, sur toile, & fonds de bois, tant de Mons. Rubens, que de Mons. Van Dyck.’ Some went straight onto the market and were recorded later the same year in a list compiled by the Antwerp dealer Matthjis Musson of the items he had acquired from Rubens’ estate. Others of them were ‘finished’ after his death, probably for the purpose of selling them on to unknowing buyers as complete Rubenses, turning them from head studies into bust-length portraits. 
5 Musées Royaux des Beaux-Arts Belgique, Brussels, inv. 3461; see Jaffé 1989, p. 197, no. 262, reproduced; Van der Auwera & Sprang 2007, pp. 71-6, no. 1 (the authors were unaware that they had reproduced the present sketch in reverse, and were sonfused as a result), and K. Bulckens, Corpus Rubenianum Ludwig Burchard, Part V (2), The Ministry of Christ, London and Turnhout 2017, pp. 135–40, no. 30, reproduced fig. 138.  A second version, possibly autograph, also including the same figure at the extreme left, is in a private collection in Toledo, Ohio; see Jaffé 1989, p. 197, no. 262, reproduced.  A painting of this subject is recorded in the Estate settlement of Rubens' first wife, Helena Fourment (though an odd subject for a virtuous wife to own or to have been given by her husband), and was probably still in Antwerp in the 18th Century.
6 Edinburgh, National Galleries of Scotland; see McGrath, 1997, vol. II, p. 304, reproduced vol. I, fig. 211.
7 Private collection; see Jaffé 1989, p. 227, no. 423; McGrath 1997, vol. II, p. 303, reproduced vol. I, fig. 206; and W. P[rohaska] in Vienna 2004, pp. 117–18, reproduced fig. 3.
8 Alte Pinakothek, Munich.
9 See Wieseman, 2004, p. 100.
10 H. Vlieghe, A. Balis and C. Van de Velde (eds), Concept, Design, and Execution in Flemish Painting (1550–1700), Turnhout 2000, p. 141.
11 It does not include, however, the 3 cm-wide vertical strip at the left of the present panel.
12 Sold, London, Christie’s, 2 July 2013, lot 30.