After spirited bidding between Mr. Lesser of New Bond Street and the Parisian dealer F. Kleinberger on July 2, 1909, Kleinberger paid 7,800 guineas for The Descent from the Cross. Although the painting had not been cited in the literature and was unknown to Rembrandt scholars until just before the sale, the price was a record for a Rembrandt painting sold in a London auction house. Aside from the excitement surrounding the discovery of a new Rembrandt, the high price was undoubtedly influenced as well by the positive opinion given about its authenticity the previous week by the leading Rembrandt authority of the day, Dr. Wilhelm von Bode. ^[1]

The painting, then signed and dated 1651, was recognized by Bode and, subsequently, by other scholars as a free variant of Rembrandt’s earlier representations of The Descent from the Cross, his 1633 painting for the Passion series that was intended for Prince Frederik Hendrik ^[fig1], and, more specifically, his large-scale depiction of this subject, signed and dated 1634, in the Hermitage, Saint Petersburg ^[fig2]. Hofstede de Groot, Stechow, and Rosenberg, among others, noted how Rembrandt had in this work reduced the number of compositional elements that had appeared in the earlier examples. They also wrote movingly about how the changes had enhanced the scene’s pictorial expression and emotional content. ^[2] Scholars also identified the 1651 Descent from the Cross as one of the two paintings of this subject listed in the inventory of Rembrandt’s possessions in 1656. ^[3]

The positive assessment given to the painting for the sixty years after it appeared at auction in London in 1909, however, came abruptly to an end in 1969 when Horst Gerson wrote that the painting was the work of a pupil. As far as he was concerned, “the gestures are lame, the expression sentimental and the composition as a whole lacks concentration.” He suggested that the painting may have been executed by a “pupil like B. Fabritius or S. van Hoogstraten.” ^[4] Since Gerson’s publication no Rembrandt authority has accepted the work as autograph. In 1984 the attribution of the painting at the National Gallery of Art was changed to “After Rembrandt van Rijn.” The Rembrandt Research Project (RRP) later listed The Descent from the Cross as a copy of the Hermitage painting, the attribution of which they also reject. ^[5] The RRP suggested that the Gallery’s painting “may very well have been produced in his circle.” It further allowed that Rembrandt may have permitted “variants done by pupils to be included in the 1656 inventory of his belongings as being his own work.” ^[6] Finally, Sumowski placed the painting among those executed by anonymous members of the Rembrandt school. ^[7]

Heavily discolored varnish and extensive repainting ^[fig3] have profoundly affected earlier assessments of the emotional content of the work and even its attribution, including that of the RRP. ^[8] To try to come to a clearer understanding of the place of this work within Rembrandt’s workshop, removal of the overpaint ^[9] was undertaken in 1991–1992. ^[10] Although this conservation treatment helped resolve some of the questions about the complex genesis of this work that were first raised when the painting underwent technical examination in 1978, many questions still remain. ^[11] The following text examines the painting’s original appearance and the character of the extensive revisions that were made to it in the mid-seventeenth century. It will also discuss the probability that the painting was treated again prior to being sold in London in 1909. This text will then examine the attribution of the work and its place within Rembrandt’s workshop.

The compositional connections between the painting and The Descent from the Cross in the Hermitage, which have been noted ever since the time of Bode, are even closer than one would assume from looking at the surface. X-radiographs [see X-radiography ^[12]] ^[fig4], which are admittedly difficult to read because of the unusual striations across the image caused by an uneven application of ground ^[13], ^[14] but they do reveal much about changes both in design and in the shape of the canvas. Although slight distortions in thread patterns in the canvas at the top and right indicate that these edges may have been trimmed a small amount, no such scalloping patterns are evident along the left or the bottom, an indication that the canvas has most likely been reduced substantially along these edges. Also evident in the X-radiographs is a vertical seam to the left of center where two canvases have been joined. ^[15] If one hypothesizes that this unusually situated seam originally marked the central axis of the painting, then it is clear that two-thirds of the left half of the original image have been eliminated. The original format thus would have been comparable to that of the Hermitage painting in that Christ was centrally placed in the composition. It is also probable that the proportions of the two compositions would have been comparable. The Washington painting would then have been substantially larger than the Hermitage version (the Hermitage painting measures 159.3 x 116.4 cm; the proposed width of the Gallery painting would have been approximately 160 cm, or the double of 80 cm, with a proportional height of about 220 cm). ^[16]

The hypothesis that the Washington painting was once a larger-scale version of the Hermitage painting is reinforced by the character of the design changes evident on the X-radiographs. The clearest of these is the change in the position of the man on the ladder who holds the torch that illuminates the scene. This middle-aged man ^[fig5] has been painted over a younger figure whose head, in a position identical to that in the Hermitage painting, can be seen in the X-radiographs at about the man’s chest level ^[fig6]. Although the image of Christ is difficult to read because of the density of the lead white paint, his legs were initially bent back in a position comparable to that in the 1634 composition. Also vaguely visible in the X-radiographs are the profiles of two figures found in the Hermitage painting that were subsequently eliminated from the Washington version—that of the bearded man standing just below the youth with the candle and that of one of the male onlookers crowded to his right. Finally, the arm of the Virgin was originally illuminated as it is in the Hermitage painting.

Cross-sections have provided corroborating evidence that the paint layers are quite complex and that the colors underlying the surface paint are similar to those in the Hermitage painting. The most striking instance is a bright orange that can be found in exactly the same area one finds the bright orange costume to the right of the turbaned man in the foreground of the Hermitage painting. The cross-sections also suggest that the extensive changes in the National Gallery of Art painting were made only after the first composition had been blocked out with a thin, dark layer of paint. This layer has been found in every cross-section with the exception of the neck of the Virgin. Indeed, the head of the Virgin does seem to be the only part of a figure in the painting not extensively reworked, although the broadly executed highlights on her face may have been added to the preexisting form to tie in to the handling of the other figures. Interestingly, associated with this pervasive layer of dark paint is an unpigmented layer. While this layer generally appears to lie on top of the dark layer, sometimes it seems to pass through it and sometimes to lie below it. ^[17] The layer is probably varnish, which may indicate that a short lapse of time existed between the execution of the underlying image and the final composition.

Evidence indicates that The Descent from the Cross must have undergone further treatment prior to its sale in July 1909. A letter from June 26, 1909, states, “the Rembrandt has been very badly treated, having apparently been hung against a hot flue, which has blistered the picture all up the left side.” No evidence of such blisters [see blistering ^[18]] is currently found on the painting, thus it seems probable that the painting was trimmed along the left and bottom to remove them. Reinforcing this hypothesis is the fact that when the painting was sold in July 1909 (55 x 42 inches) its dimensions were smaller than they had been when it was sold in April 1840 (5’ x 4’5”, which equals 60” x 53”). Just when this reduction took place is unknown, but it was probably not long before the July sale. ^[19]

The individual who restored the painting prior to the 1909 sale did his best to solidify the Rembrandt attribution. It was probably at that time that the signature and date were added, and that the head, shoulders, and turban of the figure in the foreground were overpainted. When these later additions were removed during the treatment of 1991–1992, it was revealed that the paint surface below that figure had been badly abraded. Although various underlying layers of paint are difficult to interpret with certainty, it appears that the repainted turbaned figure in the foreground covered a comparable figure that had been scraped down. Beneath that earlier figure, however, was yet another one: remnants of his black, flat-shaped hat still exist in an underlying paint layer. It may be that the earlier restorer scraped away the head and shoulders of the turbaned man because he saw evidence of a different figure beneath it. Eventually, it seems, he concluded that so few remnants of the earlier head with the flat-shaped hat existed that it behooved him to repaint the turbaned figure he had just removed. ^[20]

One further compositional change (probably made by a different restorer) was the shape of Joseph of Arimathea’s red coat. This restorer, who made the addition of the tempera paint, must have felt that this aged man’s body should have been more massive than it had been painted, and he added significantly to its bulk. With this repaint removed, the rigid angularity of the seventeenth-century image is now visible.

With the removal of the discolored varnish and later overpaint, it has become apparent that the seventeenth-century changes to The Descent from the Cross were undertaken with a great deal of sensitivity. Indeed, a comparison of the Hermitage painting and this work demonstrates that profound differences exist between the two works despite their apparent similarities. The Gallery’s composition is far more focused than is the Hermitage version. Not only is the cross brought forward and the figures given greater prominence, but light is concentrated on two major areas of activity: that surrounding the lowering of Christ’s body by the aged Joseph of Arimathea and the swooning figure of the Virgin. Because the figure holding the torch stands higher on the ladder and Christ’s legs have been brought forward, light focuses more broadly on the central figure group. These changes furthermore reduce the diagonal thrust seen in the Hermitage composition: the disposition of forms is more balanced, and gestures, including the arm holding Christ’s waist, have a predominantly horizontal emphasis.

The feeling evoked by the Washington painting is more reverential than that in the Hermitage version. Joseph of Arimathea seems to present Christ to the viewer, while the figures below wait quietly to assist. In the Hermitage painting, on the other hand, Joseph of Arimathea struggles with the weight of Christ’s body as others labor to pull out the nail that secures Christ’s left hand to the cross. The emphasis there on the physical activity of removing Christ’s body from the cross is reinforced by the angular gestures, the strong diagonal shadows on the white shroud, the ungainly position of Christ’s body, and the press of the crowd around the foot of the cross.

The total rethinking of the composition in the National Gallery of Art painting speaks strongly for the participation of Rembrandt in the process, particularly because the emotional content of the work is so sympathetic with his approach to religious imagery during the 1650s. Nevertheless, while Rembrandtesque, the execution is not that of the master. Heavy impastos on the face of the man holding the torch, for example, are coarsely applied, while the white sheet wrapped around Christ is painted in flat planes of color that only superficially suggest folds in the material. Many questions thus remain: what was the date of the original composition of the Washington Descent from the Cross; who painted it; how and why were the later reworkings undertaken; and who was the artist responsible?

Dating the underlying image is quite difficult. Technical evidence gained from examinations of the canvas or paints used has not yet provided precise correlations with other works. ^[21] On the basis of the compositional similarities with the Hermitage Descent, it would seem logical to assume that the Washington version would have been executed at approximately the same period of time. Just when that was, however, is a matter of some dispute. Although the Hermitage Descent is signed and dated 1634 and apparently was painted on the same type of canvas as that used for Rembrandt’s Flora, 1634, also in the Hermitage, the RRP has not only rejected the attribution to Rembrandt but has also argued on stylistic grounds that the painting was executed in Rembrandt’s workshop around 1640. Nevertheless, the existence of the date, the evidence of the canvas weave, and the close compositional similarity to the 1633 Descent from the Cross from the Passion series makes it seem most probable that the Hermitage Descent was conceived in the mid-1630s. During these years Rembrandt was particularly fascinated with the drama and emotional intensity of this story, something he explored in comparable ways in oil sketches, prints, and drawings. It is unlikely that in the 1640s he would have entrusted a student in his workshop to recreate a composition that resonated so distinctly with a compositional idea first developed in 1633.

If the Hermitage Descent were, indeed, first conceived in the mid-1630s, it then is probable that the Washington painting was as well. Perhaps both large-scale works were made in anticipation of forthcoming commissions that never materialized. An added incentive may have been Rembrandt’s desire to compete with Rubens, Peter Paul, Sir. Rembrandt, who based his 1633 Descent from the Cross for the Passion series on Lucas Vorsterman’s reproductive engraving after Rubens’ altarpiece The Descent from the Cross, now in the Antwerp Cathedral, may have decided to emulate not only Rubens’ composition but also the large scale in which Rubens worked. Although the degree to which Rembrandt was involved in the execution of either of these works cannot be determined, he may well have blocked in the composition for at least one of them. The final execution may then have been left to a student or students. In any event, there seems to be no stylistic or technical evidence by which to conclude that one of these works was a direct copy of the other.

Although the signature and date 1651 on the Washington Descent were determined to be later additions and removed during treatment in 1991–1992, the date is not inconsistent with the style of the figures painted over the earlier composition. As mentioned above, moreover, the artistic concept is also consistent with Rembrandt’s work from this period. Although no documents provide information to explain why such extensive reworkings were undertaken at this time, the reason may once again have been the hope that a commission for such a scene would materialize. Rembrandt may have decided that the very large size of the original composition made the work particularly difficult to sell. He may also have felt that the original composition provided the foundation for a particularly fascinating challenge, and thus he reconceived a dramatic story by subtly changing the positions of the figures, the lighting effects, and even the moment depicted to create a painting with a different mood and emotional impact.

Rembrandt was undoubtedly closely involved in the rethinking of this composition and may well have blocked in forms to serve as a compositional guide, but no evidence of his own brushwork exists in the final image. Just who may have been responsible for the execution is difficult to judge. Nevertheless, sufficient stylistic connections can be found between this work and the paintings and drawings attributed to Renesse, Constantijn van to make a tentative attribution of The Descent from the Cross to this fascinating Rembrandt student.

Van Renesse, about whom very little is known, seems to have been with Rembrandt between 1649 and 1652. ^[22] Rembrandt must have taken a great deal of interest in his work, if one is to judge from the drawings by Van Renesse that he corrected. ^[23] Van Renesse had a preference for biblical scenes, many of which focused on the life of Christ. ^[24] Stylistically, Van Renesse’s figures compare closely to those in the Washington Descent. In Van Renesse’s drawing of Doubting Thomas (Staatliche Graphische Sammlung, Munich), Christ’s elongated proportions and anatomical structure are the same as those in the Descent. ^[25] Similar figure types also occur in paintings convincingly attributed to Van Renesse. In his The Good Samaritan ^[fig7], for example, the crossed legs of the wounded man have much the same structure as do those of Christ in the Washington painting. ^[26] Finally, Van Renesse’s painting technique combines the smooth, flat planes of color and rough impastos found in The Descent from the Cross. This combination of techniques is particularly evident in his Conviviality near the Inn in the Corcoran Gallery of Art ^[fig8]. Although this painting is not signed, the close compositional similarity between the Corcoran’s painting and Van Renesse’s etching, signed and dated 1651, confirms the attribution. ^[27] Despite the different nature of the subject, the manner in which a number of figures in this work have been executed has distinct parallels in The Descent from the Cross (see ^[fig4]).

This painting thus is a fascinating document about the complexities that sometimes exist with works produced in Rembrandt’s workshop. ^[28] The evidence suggests that it was initially larger in size, with a composition that resembled that of the Hermitage Descent from the Cross. The exact date of the first period of execution cannot be precisely determined, but it probably was during the mid-1630s. Around 1650, or shortly thereafter, it was severely cropped at the left and bottom, and virtually the entire composition was reworked. Although Rembrandt was probably involved in the rethinking of the composition, he does not seem to have had any part in the final execution. Stylistic evidence suggests that the artist responsible was Constantijn van Renesse. If this hypothesis is correct, one could then argue that Van Renesse was a far more central figure in Rembrandt’s workshop in the early 1650s than has hitherto been believed. It may well be that he was involved in a number of other large-scale religious paintings from this period that were produced by unidentified members of Rembrandt’s workshop. ^[29]

The support is a medium-weight, plain-woven fabric consisting of two pieces seamed vertically to the left of center through the Christ figure.[1] The painting has been lined with the tacking margins removed. Slight cusping is visible along the top and right edges, but not at the left or bottom. Both the seam and the figure of Christ would typically be in the center of the composition, as in The Descent from the Cross in the Hermitage Collection. Consequently, it seems apparent that the painting was significantly reduced along the left and the bottom. It is probable that the painting was cropped on two different occasions, once in the seventeenth century and a second time in the early twentieth century. Nevertheless, the extensive compositional changes noted below indicate that most of the cropping occurred in the seventeenth century.

Evidence that the canvas was trimmed in the early twentieth century is based on a letter dated 26 June 1909 that states: "the Rembrandt has been very badly treated, having apparently been hung against a hot flue, which has blistered the picture all up the left side." [2] Since no blisters are visible along the left edge today, and there is no evidence that they once existed, it seems that the painting was trimmed prior to its sale in London in July 1909. At the time of the sale the painting’s dimensions (55 x 42 inches) were approximately the same as those today. A set of tacking holes and crease marks along all four edges within the picture plane indicate that the edges were turned over a smaller stretcher at one time. The edges were subsequently returned to plane when the painting was lined.

The paint was applied over a double ground composed of a thick, light gray lower layer followed by a thin, brown gray upper layer.[3] The paint handling varies from rich opaque layers to thin glazes, with complex layering and dramatic brushmarking in light passages. The X-radiographs show artist’s changes to the figures supporting Christ’s body, the legs of which were once bent backward to a greater degree. Initially a young man stood where the older man with a torch is placed. Two profiled figures, visible just below the younger figure’s head, were also painted out. The original composition was painted out with a thin layer of dark paint.[4] Upon removal of later repaint in 1992, it was determined that the turbaned foreground figure had been painted over another figure that had been intentionally scraped down. It is unknown when and why this change was made.

The seam and creases protrude slightly. Scattered small tears are visible in the X-radiographs, notably along the top edge at center and in the background right of center. Numerous small paint losses are scattered overall, and abrasion is light, save in the turbaned figure. The painting underwent treatment in 1991–1992 to remove discolored varnish and overpaint. It was inpainted in 2000, including reconstruction of the foreground figure’s head and face. These were reconstructed on the basis of the remnants of paint left from the original head.

[1] This vertical seam is located 37.9 cm. from the left edge and 73 cm. from the right edge.

[2] The information about the condition of the painting in the letter was provided by "an artist friend" of E.W. Parker’s solicitor. It is not known when the "artist friend" saw the painting in its deteriorated condition. The text continues, "The sky also shows signs of having been tampered with. But there can be no doubt of its authenticity although it strangely resembles (and yet differs from) another Rembrandt "Descent" of a date about 20 years earlier, at St. Petersburg." Records of the Parker family of Skirwith Abbey, Warwick Hall, and Newbiggin Hall; Cumbria Record Office, Carlisle; WD PKR, box 4, bundle 18, document 12; copies in NGA curatorial files.

[3] The ground and paint were analyzed by the NGA Scientific Research department using cross-sections, Fourier Transform Infrared spectroscopy, polarized light microscopy, X-ray fluorescence spectroscopy, and gas chromatography in conjunction with mass spectrometry (see reports dated July 1978; July 8, 1991; August 10, 1991; August 14, 1991; October 8, 1991; November 15, 1991; December 2, 1991; plus undated cross-section studies, 1991, in NGA Conservation department files). The ground was further analyzed by Karin Goen using cross-sections (see Karin Groen, "Grounds in Rembrandt’s Workshop and in Paintings by His Contemporaries," in Stichting Foundation Rembrandt Research Project,　A Corpus of Rembrandt Paintings, vol. 4, Self-Portraits, ed. Ernst van de Wetering [Dordrecht, 2005], 664-665). Groen found quartz in the lower ground layer.

[4] This layer was confirmed by cross-sections taken and analyzed by the NGA Scientific Research department (see report dated October 8, 1991, in NGA Conservation department files). The cross-sections also showed an unpigmented layer, presumably varnish or oil, directly on top of the dark layer.