Lot 28
  • 28

Sir Peter Paul Rubens

2,000,000 - 3,000,000 USD
2,994,500 USD
bidding is closed


  • Sir Peter Paul Rubens
  • the adoration of the magi

  • oil on oak panel
  • 19 7/8 by 25 1/2 in.; 50.5 by 64.8 cm


Possibly the Counts of Collalto, Vienna (according to the 2010 sale catalogue);
Baroness Auguste Stummer von Tavarnok, Vienna, by 1895;
Thence by descent in the family until acquired by the present owner;
Sale, Zurich, 17 September 2010.


T. von Frimmel, Verzeichnis der Gemälde im Besitz der Frau Baronin Auguste Stummer von Tavarnok, Vienna 1895, p. 65, no. 156 (as Rubens);
J. Foucart, in Le siècle de Rubens dans les collections publiques françaises, exhibition catalogue, Paris 1977-78, p. 175, under cat. no. 128 (as a copy).

Catalogue Note

This lively sketch or modello remained unknown for the entire 20th century, only in 2010 emerging from the Viennese family who had owned it since the 19th century, and it is here definitively restored to Rubens for the first time in over a century.  A number of specialists have recently seen this work and accepted it to be by the hand of Rubens, including a group of scholars at the Rubenianum who had the opportunity to study it firsthand.

Von Frimmel catalogued it when it was in the Stummer von Tavarnok collection in 1895, but it was overlooked by subequent Rubens scholars. More recently, it has been confused with another version by Rubens which, in the physical absence of the present lot, had been assumed to be the Stummer von Tavarnok painting.

This other version by Rubens, of nearly the same dimensions and similar in most but not all details, was formerly with David Koetser, Zurich, and is now in a private collection. It was widely seen in the exhibition devoted to Rubens oil sketches in 2004-5 (see fig. 1) curated by Peter Sutton, Betsy Wieseman and Nico van Hout.  It was examined by Julius Held in 1980 and confirmed by him as a modello for the Lyon painting, and was subequently studied at the Rubenianum in Antwerp, by, inter alia, Hans Devisscher who plans to publish it in his forthcoming Corpus Rubenianum volume.1  Since its was rediscovery in Switzerland in 1980 it has been assumed by Rubens scholars to be the ex-Stummer von Tavarnok painting and was published as such in subsequent literature.  The reappearance of the present painting in 2010, however, when it was consigned for sale by a descendant of Baroness Stummer, firmly dispelled the Stummer von Tavarnok connection that post-1980 Rubens scholars had assumed for the ex-Koetser version. Earlier, Foucart cited it in the 1977 exhibition catalogue Le siècle de Rubens, but he was referring only to von Frimmel's reference and had no knowledge of either this or the Stummer von Tavarnok version.

The present painting is preparatory for a large canvas now in the Musée des Beaux-Arts in Lyon (see fig. 2). Although not so fully resolved in detail as the larger work in Lyon, and as is typical of Rubens' working method, the modello reveals numerous alterations. Overall the composition of the Lyon version is more compressed, with the figures moved closer to the picture plane, imbuing them with a greater monumentality. To mention but a few of the particular alterations: the Virgin, whose head is turned a few degrees towards the viewer, so that it is no longer in pure profile, stands a little more stiffly upright in the Lyon painting. Conversely, St. Joseph is no longer seen in three-quarter profile but is turned away from us in pure profile. The page that tends Melchior's train in the extreme lower left is straightened up, relinquishing the whimsical gesture he makes with his head in the modello; the youth who leans in surprise towards the Christ Child is turned into a stern-faced, bearded older man; the black servant behind Balthasar is clothed in a new striped tunic that covers his arms; and the space itself is better defined, so that it is now clearly a ramshackle stable. Rubens thus formalizes the composition for the Lyon canvas and in so doing abandons much of the fluidity and characterization that makes the modello so instantly alluring. The apparent stiffness of the larger Lyon canvas may be partly to blame for its being long overlooked as a major work by Rubens until Jacques Foucart chose it for the exhibition Le Siècle de Rubens in 1977-78.

Several of the characters here represented are based on existing worked-up head studies that Rubens clearly kept in his workshop for just such a purpose as this commission. Unaware of the modello, Julius Held likened the head of Balthasar in the Lyon canvas to that of the leftmost of the two Levantine head studies in the painting formerly in the Anton Philips collection in Eindhoven (see fig. 3).2  The resemblance of Balthasar to the ex-Philips study, though obvious in the corresponding figure in the present modello, is stronger still in the finished canvas in Lyon where he is likewise looking almost directly at us, his head on a very slight right-leaning tilt. The modello however shows Balthasar's head un-tilted and his eyes look up to his left, as if distracted by the bearded soldier next to him. That Rubens used the ex-Philips study for this figure does however seem most likely, just as he clearly used other worked-up head studies for each figure either side of Balthasar: the head of the black servant over his shoulder is with some certainty to be linked to the left hand study in what is perhaps Rubens' most famous head study group, the Four studies of a man's head in the Musées Royaux des Beaux-Arts in Brussels (see fig. 4).3  Again, this is perhaps clearer still when the study is compared to the figure in the Lyon canvas which so obviously replicates the Brussels study; with the modello, Rubens has, with just half a dozen strokes of the brush, signified his intention to use the first of the four studies for this figure, which he then works up in far greater detail in the finished canvas. Held surprisingly failed to notice this link, connecting the Brussels studies only with the Diogenes in Frankfurt and the Allegory of Nature in Glasgow. For the head of the old, bearded man immediately to the right of Balthasar, Rubens used another head study, seemingly now lost but surviving through a probable copy in the Hermitage, St. Petersburg (see fig. 5).4  That particular head was first conceived as a tiny drawing, along with other studies of the same model's head in different positions, on a sheet now in the collection of the Duke of Devonshire at Chatsworth (see fig. 6). This sheet, when considered together with the lost painted study, the present modello and the finished canvas in Lyon allows us a fascinating and complete understanding of Rubens' working practice from start to finish through the four stages of this project's development.

Prior to cleaning, which was carried out subsequent to the 2010 sale, the right hand section that now appears only summarily sketched in, had been manipulated to include a diagonal roof beam and a section of sky above the Virgin. The figure of St. Joseph had, furthermore, been crudely worked up in a most un-Rubensian manner, and the boy at the center left who lifts the lid off the gilt vase had also been clumsily "finished." This overpaint now removed, it is possible to fully appreciate the wholly recognizable spontaneity of Rubens' brushwork. The ex-Koetser version includes a triangular section of sky above the Virgin not dissimilar to the pre-cleaned image of the present lot, though the diagonal roof beam is missing. The figure of St. Joseph in that painting bears little resemblance to either the pre-cleaning or post-cleaning Joseph of this painting.

The painting itself most likely dates from the second half of the second decade of the 17th century and, in its lively surface and use of stark primary pigments, it has much in common with other sketches from this time. The Lyon canvas is dated by most scholars to 1617-19 and its many similarities to the painting of the same subject in Mechelen (documented commission finished in 1619) would assist the argument for such a dating. Time and again the subject provided Rubens with the opportunity to explore an extraordinarily varied cast of characters, and he returned to it more than any other subject from the life of Christ, executing at least ten projects on the theme, two of them within a year or two of this example.5  While the Lyon canvas does have a distinguished early provenance (it was acquired by the Elector of Bavaria in Antwerp in 1698 from Gijsbrecht van Ceulen, along with twelve further canvases by Rubens now in the Alte Pinacothek in Munich), its original purpose, and thus that of the sketch, remains unclear, though, as Peter Sutton has argued, the horizontal form of both canvas and sketch would suggest a secular rather than ecclesiastical commission, its shape not lending itself to the traditional form of an altarpiece.

1.  P.C. Sutton and M.E. Wieseman, with N. van Hout, Drawn by the Brush. Oil Sketches by Peter Paul Rubens, Greenwich/ Berkeley/ Cincinnati, 2004-05, p. 107, cat. no. 6, reproduced.
2.  J.S. Held, The Oil Sketches of Peter Paul Rubens, Princeton 1980, vol. I, p. 613, cat. no. 448, reproduced vol. II, color plate 5.
3.  Ibid., vol. I, pp. 607-09, cat. no. 441, reproduced vol. II, plate 428.
4.  See M. Jaffé, Rubens, Milan 1989, p. 232, no. 455, reproduced p. 456.
5.  In addition to the aforementioned example in Mechelen, another from circa 1619 is in the Musées des Beaux-Arts, Brussels; see Jaffé, op. cit., p. 247, no. 526, reproduced.