The resurrected Christ, his flesh marked by the wounds from the crucifixion, holds the cross, arching back to counterbalance its weight. The shroud, a simple white cloth, hangs around him. The painting does not show any specific incident from gospel or legend, but rather offers Christ as an object for meditation. It is an affective image, meant to elicit devotion, to make the faithful reflect on Christ’s sacrifice and triumph. This is made all the more immediate because of his physical presence, conveyed through the strong contrast of light and dark, the compact, tightly framed image, and the sculptural treatment of form.
In the past this painting has sometimes been identified with a missing Christ, documented as part of a series of thirteen panels showing Christ and his apostles, painted by Rubens in c. 16121 (the twelve Apostles are now in the Prado). The Ottawa Christ was rediscovered in 1906 and exhibited as the missing work the following year in London. The attribution was controversial, although it quickly found influential support.2 Aware of the controversy, but trusting its advocates, the Gallery bought the painting in 1929. This claim would continue to find supporters for several decades, but scholars would later come to unanimously see it as a copy, one of several known. In the Corpus Rubenianum, the standard catalogue of the painter’s work, it was published as the best surviving example after the missing original.3 Due to its apparent quality, the possibility of Rubens’ participation had some credence, and the Gallery accepted it as by the master and his workshop.
Many copies and variants of the Prado Apostles survive: only one set is certainly from within the workshop, this is now in the collection of the Pallavicini, Rome.4 Others were likely made outside the workshop, which attests to the series’ popularity. In addition, there are several single panels of saints – unlikely to be part of any complete set – which derive from the Prado series. The Roman Apostles include a Christ that is closely related to the type seen in the Ottawa panel. It is likely that the Prado Apostles had a similarly posed Christ.5
A good candidate for the model for the Ottawa panel is a painting now in Vienna (illus. 1)6 This panel is clearly superior in quality, and arguably autograph. The relationship of the two works has never been closely explored, but the Ottawa Christ appears to be a copy after it.
The Vienna work shows Rubens’ characteristic means of painting flesh: the warm coloured imprimatura is left visible as a mid-tone in places, or partially visible and modifying the tone of an overlying scumble, for instance in the shadow cast on his left bicep. The flesh is painted in layers – the warm and cool hued underpainting then scumbled over with thin, tinted greys – to produce luminous, translucent skin. In the final phase of painting, the artist added cool, light touches, broadly painted, to highlight and animate the surface and strengthen the anatomy, for example at the throat and collar bone. By contrast, the flesh of the Ottawa Christ is painted directly, potentially in one session. It does not rely on layering, but goes directly to the modelled form with mixed opaque tints, and the highlighting is not a separate stage of painting. There is a consequent diminishment of luminosity, and a shift in hue resulting from the use of a mixed tint.
Given its fidelity, it is clear that the painter of the Ottawa Christ had access to the Vienna panel as he worked. It is also likely that the Ottawa painting was begun before the other was finished: a pentiment in the Vienna panel finds an echo in the Ottawa copy. The underpainting of the drapery in the Vienna Christ – broad, horizontal strokes made when the pattern of drapery folds was not yet fully established – was initially followed in the Ottawa panel, before being painted over. As noted, the flesh on the Vienna painting was complete – including the retouching – before the copyist began that area. However, the retouches that darken shadows on the drapery of the Vienna Christ were added as final touches to the Ottawa copy. All this points to the Ottawa copy being made simultaneously and as its model developed.
Scaled overlays of the two paintings make it clear that the artist of the Ottawa panel did not employ any of the usual means – such as tracing or griding – to ensure accuracy (such methods would have been precluded if the model was still incomplete when the copy was begun, as we argue: its paint, still wet, would have been easily marked.). The Ottawa version would instead seem to have been copied by sight, leading to subtle but significant discrepancies in scale and placement. Aligning with this observation, its minimal underdrawing, done in a fluid medium, is for the main, loose and freehand, searching for form. Only in areas where greater precision was needed – the expressive hands and face – is the underdrawing finer and more deliberate, possibly made in a dry medium; it was followed carefully in the painting.
Making a copy involves judgement – what to reproduce, how to do so – and reveals something of the copyist’s visual acumen, and how he saw and interpreted his model. In this instance, the more direct technique of the copy has necessarily made Christ swarthier, with a more robust musculature. In places, the copyist followed the original pedantically and unintelligently – particularly in trivial details. For instance, the copyist felt obliged to reproduce even the small circular marks which blemish the bark of the cross, but rendered them as simple, non-descriptive dots, betraying Rubens’ painterly, yet naturalistic, vision. In another instance, the copyist was required to adapt and invent to solve an unforeseen problem. Late in the course of painting an extra fold of cloth was added to Christ’s drapery over the already painted flesh of his thigh. Due to the taller format of the Ottawa panel, more flesh was visible, and this must have been perceived as unseemly. A layer of pale yellow was added, which was then followed by strokes of cooler white lead. The intention was to mimic the effect of the warm buff-coloured imprimatura seen through the thinly painted cloth, as visible elsewhere in the drapery. The change will likely have been made with Rubens’ consent.
While the paint of this addition was still wet, someone drew into it using the butt-end of the brush. It is difficult to fully understand the intention in making such a mark, which seems unmotivated. It has sometimes been interpreted as a kind of paraph or identifying mark, which seems improbable; it has no obvious explanation.
Although the Ottawa panel is a copy, it is impossible to avoid acknowledging that the two paintings would have looked very different – a result of the differing panel dimensions, deviations in the drawing, as well as tonal differences – primarily due to different techniques of painting. It would have been possible to create a much more faithful copy, and these variations may have been welcomed. They must have had Rubens’ approval, even if only tacit. In effect, the Ottawa painting is not so much a replica as a secondary version, made by the workshop. It translates, rather than replicates, the autograph model, and introduces a minor variant.
The technical evidence shows that the Vienna and Ottawa panels are closely connected in time, yet look very different. This shows that significantly different means of painting could be in use simultaneously by the workshop and its master, the result being that dating by style alone is unreliable. Variations in handling, technique and style may have been seen as less consequential than we may view them today. They were an unavoidable by-product of the different hands involved. Yet to viewer attuned to these differences – Rubens’ fellow painters, for instance, or a sophisticated collector – they reinforced the hierarchy between autograph and workshop. The case here is unusually clear. In the case of other variants, the assistant painter would more closely emulate Rubens’ method of working.
The Vienna panel, as a direct model for the Ottawa painting, and clearly influential for that in Rome, is a potential candidate for the missing Christ from the Prado series. According to dendrochronology, the panel for the Ottawa Christ was available for painting by c. 1610.7 The Vienna painting can be dated to roughly c. 1611–15 by its style and the Prado series is thought to date to c. 1612. The Vienna panel nearly matches the dimensions of the Prado panels (107 × 82 cm vs. 108 × 84 cm), and is of an appropriately high quality, given its role as a centrepiece for the set.
Whatever the case, there is a connection between the production of the Ottawa Christ, the Prado series, and the set of variants in Rome. At the very least, they are all the product of a shared climate in which the workshop was tasked with copying models in demand, the results sometimes showing a remarkable degree of fidelity, yet, in others, adaptations and deliberate changes.
The support for the painting is composed of three oak boards, the grain running vertically; butt-joined and doweled. The oak is from the Netherlands or western Germany, and is slightly off radial cut (between 12 and 20 degrees). The boards measure, from left to right at the bottom edge: 27, 26.5, and 29 cm; neither join is parallel with the edges of the painting, or with each other. The original panel has been unevenly thinned and is presently around 7 mm thick at the maximum. This likely happened in 1929, when the support was put onto a wooden core and cradled by H.E. Thompson of the Museum of Fine Arts, Boston (Rooses 1910, p. 284, saw the work when with L.L. Maillard (1845–1914), a restorer in Antwerp, but no documentation survives.). The painting was put into a micro-climate box after its 1929 treatment, but flaking was noted in 1947. The picture was cleaned and restored in 1996, when at least two varnishes were removed; at this time a new micro-climate box was constructed for it.
Jean-Gilles-Marie-Joseph Schamp d' Aveschoot (1765–1839), his sale, Ghent, 14 & ff. September 1840, lot no. 156, Rubens, “Le Christ montant au Ciel … Le Christ est en proportion de grande nature, et se trouve vu plus qu’à mi-corps. Il entoure de son bras gauche la croix sur laquelle sa main droite vient s’appuyer … gravé par Ryckmans,” on wood, 42 × 31 pouces, employing the pied du rois français, i.e. 114.0 × 84.0 cm; bought in at 750 frs. The English Convent, Bruges. In 1906 bought by Henry S. Roche, London (dies 1929); in 1929 bought £4,500 by the Gallery from Lucy Roche, his widow.
See Rooses 1910, p. 284. Vlieghe 1972, p. 38, no. 6, copy 1, tentatively identifies this with the work owned by Léopold-Joseph Cocquereau (dies 1806), his sale, Brussels, 25 & ff. August 1806, lot no. 85, Rubens, “Le Christ tenant sa croix, figure proportion naturelle, vue à mi-corps …” on wood, 105 × 76 cm; bought by François Xavier de Burtin (1743–1818) at 105 frs. Not listed in any of his sales, nor in his Traité théorique et pratique des connoissances. v. 2 (Brussels, 1808), which includes a catalogue of his collection.
I – Niedersächsisches Landesmuseum,
Hannover, inv. KM 1912, 398.
Oil on canvas. 126.3 × 87 cm.
Art trade, Hannover; gift of Hofbaurat Mackensen, 1912.
Vlieghe 1972, copy no. 4; U. Wegener ed., Die holländischen und flämischen Gemälden des 17. Jahrhunderts. Niedersächsisches Landesmuseum Hannover, Landesgalerie. (Hannover, 2000), no. 156, as after Rubens, The Man of Sorrows with the Cross, 17th century.
1See Mettingen 2010–11 on the problems surrounding this material, and especially J.J. Pérez Preciado, pp. 40–47; see also Díaz Padrón 1995, vol. 2, nos. 1646–57 on the Prado set.
2It was exhibited at the Royal Academy winter exhibition in 1907 and caused some controversy in the press. Claude Phillips wrote in The Daily Telegraph (London), Wednesday 23 January, that the owner considered it autograph, and the missing Christ from the Prado series. As part of an attack on the Royal Academy’s uncritical acceptance of lenders’ attributions, Phillips claimed it was a workshop or school piece. The owner seems to have then enlisted Wilhelm Bode to support his claims. Phillips capitulated, publishing Bode’s letters in the Telegraph, 18 March 1907. Tatlock 1927, facing renewed doubts as to its status, championed the work, citing Bode’s authority. The Gallery bought it on the advice of Charles Ricketts, its London advisor.
3 Vlieghe 1972, no. 6, as copy no. 1.
4 Zeri 1959, under Rubens, nos. 431–443.
5 The situation is complicated by the existence of two different types of Christ recorded in copies and variants of the Prado Apostles, one of which is seen here and another in which Christ holds the cross before him. Regardless of their priority, Rubens chose the type depicted in the Ottawa painting when he had the Apostles series reproduced in print some time before 1619; the prints follow the Rome set.
6 See Vienna 1930, no. 5 and Czernin 2009, p. 66. It was in the Imperial collections in Vienna by 1720, recorded in one of the inventories of Ferdinand Storffer (compiled 1720–33). Vlieghe 1972, no. 6, copy no. 2.
7 The youngest heartwood ring dates to 1591; assuming a median of 17 sapwood rings were removed during the preparation of the wood, and that it was seasoned for two years, the panel was available for use from 1610 upwards. Our thanks to Peter Klein for the dendrochronology.