Sebastiano del Piombo’s ‘Raising of Lazarus’: A History of Change
Jill Dunkerton and Helen Howard
Commission and early history
The huge panel painting showing ‘The Raising of Lazarus’ from the dead, as recounted in the Gospel of Saint John (plate 1), was commissioned by Cardinal Giulio de’ Medici from Sebastiano del Piombo, probably towards the end of 1516 (note 1). Shortly before, it appears that the cardinal had ordered from Raphael, Sebastiano’s great rival, a painting of the Transfiguration, which was to be on a panel of the same dimensions. Both were destined to be set beneath the tall gothic arches of the cathedral of St Just in Giulio’s bishopric of Narbonne.
Letters sent to Michelangelo (who had left Rome in December 1516 for Florence and Carrara) by his friend and assistant Leonardo Sellaio, as well as some from Sebastiano himself, are an important source of information for the competition between the two painters (at least from the point of view of the Sebastiano party). The letters also include occasional references to practical details of the making of the altarpiece. We learn from a letter from Sellaio dated 19 January 1517 that Sebastiano was responsible for arranging for the carpentry of the panel and that he had received funds for that purpose (note 2). It seems that the panel was constructed from very long boards joined vertically, like those for Raphael’s ‘Transfiguration’ (see plate 38), rather than in the horizontal arrangement that would be expected in a Venetian panel of comparable dimensions. Whether the boards were of the usual poplar or a more unusual species, such as the cherry which has been reported as the wood for ‘The Transfiguration’ (note 3), can no longer be known.
In the same letter Sellaio told Michelangelo that he believed that Raphael was holding back on the execution of his altarpiece in order to avoid direct comparison with Sebastiano, and, by implication, for fear of the borrowing of his inventions. Sebastiano did indeed take inspiration from recent projects of Raphael’s, for example the Tapestry Cartoons and ‘The Way to Calvary’ (‘Lo Spasimo di Sicilia’) (Madrid, Prado), but in designing the ‘Raising of Lazarus’ he showed himself ready for the challenge of arranging a grandscale narrative with attendant crowds while constrained by the vertical format of the altarpiece. In this he was famously aided by Michelangelo who supplied drawings for Lazarus – still extant – and probably also for the figure of Christ (note 4).
The progress of the altarpiece can be tracked from Sellaio’s and Sebastiano’s letters (note 5). On 26 September 1517 Sellaio informed Michelangelo that Sebastiano had stopped work on other commissions to concentrate on the altarpiece. In January 1518 Michelangelo himself was briefly in Rome and saw the ‘Lazarus’. In July Sebastiano reported to Michelangelo that he was delaying completion because he did not want Raphael to see it until Raphael had finished his own painting, apparently not even begun at this stage. In addition there was an issue about the frame, which Sebastiano wished to have made in Rome, whereas Raphael was trying to influence the cardinal to have it made in Narbonne, perhaps to avoid a public confrontation between the altarpieces. Nevertheless, in the spring of 1519 the ‘Raising of Lazarus’ was placed on view, probably in Sebastiano’s workshop; according to Sellaio ‘everyone was stunned’ (‘ogni uomo resta balordo’). By December the painting had been varnished (note 6) and the huge panel transported to the Vatican for a more formal presentation, where, according to the Venetian diarist, Marcantonio Michiel, it was much praised by all, including the Pope (note 7). This was followed by wrangling over Sebastiano’s large bill. Meanwhile Raphael had at last made progress on ‘The Transfiguration’, the monumental scale and colours of the foreground figures clearly influenced by the work of his Venetian rival (note 8). Eventually, in April 1520, little more than a week after Raphael’s death, both panels were brought to the Vatican (we know from Sebastiano himself that his panel had to be transported again, and had not simply remained there [note 9]) and exhibited together. According to Vasari the two works received equal praise (note 10).
‘The Transfiguration’ remained in Rome, set up on the high altar of San Pietro in Montorio, and eventually, following a spell in Paris between 1797 and 1816, where it was restored but mercifully with minimal intervention to the panel (note 11), it was placed in the Pinacoteca of the Vatican. ‘The Raising of Lazarus’, on the other hand, followed a more precarious course over the centuries. It is not known exactly when it left Rome; nor are any details known about its transport to Narbonne, which was presumably by sea, since the French city was still a port at the time. It seems that Sebastiano had his way over the making of the frame in Rome for the lower section of a frame, of highly sophisticated design with gilded ornament against a blue ground, still survives on the altar in St Just which now holds the copy of ‘The Raising of Lazarus’ made in the eighteenth century by Carl van Loo (note 12). The presence of Sebastiano’s altarpiece in Narbonne in the sixteenth century is confirmed by reflections of the composition in French painting of the period and in 1599 it was the subject of an appreciative description by a Swiss physician, Félix Platter, who mentioned its great value and much-copied status (note 13).
The painting remained on the itinerary for other visitors to the city until 1722 when it was acquired by Philippe, Duc D’Orléans, regent of France, in exchange for a grant for the repair of the cathedral and the copy by van Loo still on the altar (note 14). The original panel was moved to Paris, where it joined the duke’s magnificent collection in the Palais-Royal (note 15).
Transfer and re-transfer
‘The Raising of Lazarus’ is described as ‘peint sur bois’ in Du Bois de Saint Gelais’s 1727 catalogue of the pictures at Palais-Royal; he also records the colours of several draperies, among them the ‘jaune clair’ of the kneeling Magdalen. Since this is a colour that would be indistinguishable from a darker yellow, or indeed from white, if the painting were covered with a heavily discoloured varnish, it has to be assumed that the painting was reasonably clean and visible at the time (note 16). In the 1770s, when the palace and its collection were in the hands of Louis-Philippe D’Orléans, grandson of Philippe, a systematic programme of restoration of the paintings was undertaken. This included the transfer to canvas of many of the paintings on panel, including ‘The Raising of Lazarus’. According to National Gallery records this was carried out, or at least begun, by ‘Haquin’ in 1771 (note 17). The treatment is likely to have been protracted. The Haquin referred to must have been Jean-Louis Hacquin (before 1726–1783) rather than his son, François-Toussaint (1756–1832) (note 18). Particularly after 1775 Jean-Louis was also responsible for the transfer and lining of a great many paintings from the French Royal Collection, now in the Louvre (note 19). It was believed that by transferring the paint layers from unstable and perhaps worm-eaten wooden panels to new canvas supports the future preservation of the works would be ensured. In practice, a great deal of damage was caused to the paintings, involving at best a complete alteration to the paint texture and at worst, the loss of large areas of the picture surface. Hacquin, originally a cabinet maker (and, it would seem, the inventor of the cradle), was probably better than most, carving away the wood of the panels from the paint layers using planes and chisels (still the preferred method in those exceptional cases when transfer is unavoidable [note 20]). This technique was certainly safer than that of his near contemporary Robert Picault who had a ‘secret’ method, which seems to have involved separation of the paint from the panel by breaking down the ground layer through prolonged exposure to nitric acid vapours; this allowed him to display intact the original wooden support alongside the transferred painting (note 21). The showmanship that was part of the process of transfer meant that its consequences were recognised by connoisseurs of painting such as Richard Payne Knight, who seems to have known ‘The Raising of Lazarus’ before its transfer. He claimed many years later that ‘those who have only seen it since that fatal operation of cutting away the pannel [sic] on which it was painted, and gluing cloth to the back of the colour in its place, can form but very imperfect notions of what it was before’ (note 22).
There can be little doubt that this drastic and dangerous intervention was totally unnecessary. The distribution of damage to the painting shows that, given its great size, the panel had remained remarkably stable, with evidence for the opening up of only one of the vertical joins, that running through the standing figure of Martha and the right leg of Lazarus. As well as fashion, a reason for its transfer might well have been the size and weight of the panel – a report of 1749 concerning the proposed transfer of the two most famous Raphaels from the French Royal Collection, ‘Saint Michael’ and ‘The Holy Family of François I’, which were regularly moved between the royal apartments at Versailles and the picture stores, observed that ‘sont peints sur bois, ce qui, joint à leurs cadres, les rend d’un poids prodigeux et par consequent très difficiles à manier, ou les transporter’ (note 23). Given the readiness of Sebastiano and the cardinal to have the panel moved back and forth from the Vatican in the early sixteenth century, it would be a strange irony if this were one of the reasons for the transfer three centuries later.
Payne Knight was not alone in his criticism of the procedure and consequences of transfer (note 24) but the fact that a painting had been transferred from its original support seems to have had little effect on its value. In 1793, following a sequence of sales and changes of ownership, ‘The Raising of Lazarus’ came to London with other Italian paintings from the Orléans collection, eventually being put up for sale in 1798, when it was bought by the insurance underwriter John Julius Angerstein for the considerable sum of 3,500 guineas; this was a higher valuation than that of many now celebrated paintings by Titian and Raphael from the same collection and was surpassed only by Annibale Carracci’s ‘The Dead Christ Mourned’ (‘The Three Maries’) (NG 2923) (note 25). In part because of the association with Michelangelo, the altarpiece caused a sensation and much debate among artists; its most vocal admirer was Benjamin West (note 26), who in about 1820 is supposed to have been responsible for the restoration and repainting of parts of Lazarus’s damaged right leg (note 27). West’s intervention was recorded in the Manuscript Catalogue of the newly founded National Gallery, centred on the core collection of 38 pictures from the Angerstein collection which were acquired for the nation in 1824. The importance of Sebastiano’s altarpiece was recognised by its being assigned the first number in the new catalogue.
Over the next few decades several entries were made in the Manuscript Catalogue relating to the painting’s condition. On only one occasion is the support mentioned, in 1837, when it needed treatment for infestation by insects; these apparently fed on the glue of the lining adhesives and were probably either flour or biscuit beetles – this outbreak at the National Gallery occasioned a short report in the ‘Observer’ of 19 September 1841. In general, there was greater preoccupation with the surface appearance and especially the varnish layers. Already by 1798 it was observed that many of the Orléans pictures appeared to have become ‘dirty, or more sunk in their colours’ (note 28), although the Sebastiano appears surprisingly bright and richly coloured in the watercolour by Frederick Mackenzie of Angerstein’s pictures hanging in his Pall Mall house (plate 2). Nevertheless, it needed varnishing in 1834, 1852 and 1867. Two letters sent in 1865 by Sir Charles Eastlake to the Keeper, Ralph Wornum, express concern about the sunk and opaque condition of the varnish, but Eastlake was emphatic that ‘no cleaning, in the picture cleaner’s sense of the term, should on any account take place’; instead ‘Pinti [Raffaelle Pinti, the London-based Italian restorer most trusted by Eastlake] should endeavour to tone down what is prominent and crude, and in short to harmonize the whole’ (note 29). In 1881 following an enquiry among the Trustees, ‘assisted by artists and others’ who included the restorers Bentley, Dyer and Pinti (note 30), it was agreed that the painting should be cleaned and revarnished by Dyer, although the extent of the cleaning is not known. Following this cleaning the frame was fitted with an enormous sheet of plate glass in order to protect the paint surface from the dirt and pollution of nineteenth-century London.
In the twentieth century it was the structural condition of the painting that caused the greater concern. In 1929 it was treated using a mixture of glue and rye flour for ‘a large number of small blisters’. Some of the old varnish was removed (suggesting that the previous cleaning was no more than a partial thinning), followed by a light revarnishing. At this point the restorers involved, Morrill and Holder, decided that the altarpiece showed no signs of having been painted on panel and that it must always have been on a canvas (note 31). In November 1939, when the painting had been evacuated to Penrhyn Castle in Wales, a ‘sizeable’ flake loss was noted and the following year areas of the surface were covered with facing paper in order to secure loose paint. Further attempts to secure the flaking paint were made in 1941 and again in 1951 but with little success.
Following the establishment of a Conservation Department at the Gallery in 1946 the stabilisation of ‘The Raising of Lazarus’ became a priority. Eventually in 1958 it was decided that it should undergo a radical structural intervention with the aim of reducing the large amount of glue that was present as a result of the application of several canvases to the reverse following the transfer. Raking-light photographs taken at the time illustrate the alarming extent to which the contraction of the glue was causing compression and lifting of the paint film (FIG. 1). When treatment began, the intention was to remove three of the lining canvases, leaving a last canvas in place. The plan was that this would be stretched out and the paint flakes secured once there was sufficient space to reattach them. Unfortunately, once the first three canvases had been removed, it was discovered that the fourth layer was not a canvas textile, but instead consisted of sheets of paper. These were badly decayed and in many areas had separated from the paint and ground.
It could not have been known at the time, but the use of paper to back a transferred paint film is characteristic of a transfer by Hacquin. In his transfers of paintings by Domenichino and Eustache Le Sueur, now in the Louvre (note 32), Hacquin used sheets of paper cut from old printed books or manuscript texts. In the report on the treatment of ‘The Raising of Lazarus’ there is no mention of any text on the paper layer, although it may have been so stained and decayed that this was overlooked. In the other transfers by Hacquin to have been investigated, two layers of fine silk (sometimes printed with a pattern) have been found embedded in the mixture of glue and flour paste which lies immediately behind the paint film, together with a residue of the original ground; he did not add a new ground or ‘enduit de transposition’ – usually lead white and oil – as used in later transfers, including those by his son, François-Toussaint. In the case of ‘The Raising of Lazarus’, however, there is no mention of there having been any silk, but there is a new pinkish-brown ground, present in all the paint samples that include the full layer structure and consisting mainly of red earth, lead white and a carbon black pigment (note 33). The gesso ground was removed completely but in some of the samples a thin layer of an unpigmented material, probably glue, can be seen between the new ground and the original paint layers. This was probably applied as part of the transfer rather than being the remains of an application of glue to seal the original gesso ground. At present too few of Hacquin’s transfers have been examined to know whether he had a standard practice. The works by Domenichino and Le Sueur referred to above already had red-brown oil-based grounds (which were not removed), and so Hacquin may not have seen the need to supplement them with a second ground. With the Sebastiano, on the other hand, the removal of the gesso necessitated a new ground. An alternative and less likely possibility (which would contradict Payne Knight’s admittedly distant memories of the operation) is that Hacquin’s transfer was not the first, and that the painting had already been transferred by Picault who worked on Orléans pictures in the 1750s (note 34). He did sometimes use a red-brown ‘enduit de transposition’ (note 35) and the premature failure of some of his transfers meant that they had to be reattached to new supports only a few years later (note 36). Some record of such a major undertaking might be expected, however, and the survival of a few splinters of the wood of Sebastiano’s panel (FIG. 2) is more indicative of a panel removal by mechanical means.
The alarming discovery made in 1958 that the first transfer layer was paper, and not canvas as expected, meant that the surviving original layers were held together only by a tissue paper facing, applied using a mastic and wax adhesive. As the treatment report (note 37) candidly records, this was adequate for the planned operation but not for a total re-transfer. There was no possibility of turning the painting over in order to attach a more substantial facing and the original layers were found to be sensitive to water-based adhesives (‘wrinkling and breaking into minute fragments’) and to the amount of heat needed to melt a solid wax-based adhesive; eventually it was decided to brush on multiple thin layers of warm wax-resin dissolved in white spirit, embedding a layer of inert terylene net fabric within the layers as they solidified. Although the discoloured varnishes had yet to be removed from the paint surface, the appearance of the picture was considered to be darker than intended as a result of lack of reflectance from the ground and so titanium white (titanium dioxide) was added to the wax cement. This bright white layer appears in some of the paint samples (for example plate 3). These were taken only after completion of the retransfer of the paint film and so the presence of the red-brown ‘enduit de transposition’, which negates any reflective properties of the new white ground, could not have been known – presumably its brown colour meant that previously it was taken to be a discoloured old glue layer.
Once the wax and titanium white layers had been built up to a sufficient thickness the paint film could then be mounted on a new solid support. This had previously been coated with wax-resin allowing a bond to be achieved by ironing with a thermostatically controlled iron to soften the wax-resin layers which then fused as they cooled. Although it is unlikely that these methods and materials would be used nowadays, the treatment can be judged a success in that there has been no further flaking of the paint layers. Unfortunately, the work took place before the introduction of lightweight and stable panels made from glass fibre with aluminium honeycomb cores (note 38). The painting is mounted, therefore, on a support constructed with ‘sundeala’ composite board outer faces and a core of paper honeycomb (note 39). In spite of its wooden edges and an internal wooden framework this panel is now showing signs of instability, with a tendency to flex and twist when the painting is moved, an operation which is therefore avoided as far as possible.
With the paint film secure, removal of the old layers of varnish could proceed. Judging by the extent of the discoloration visible in the patches of varnish that still remained on Martha’s dress and in the area of Lazarus’s shroud and the forearm of the man supporting him when the painting was photographed before retouching in 1967 (FIG. 3), several layers of the notorious ‘gallery varnish’ (a mixture of mastic and drying oil) were present; Dyer’s cleaning in 1881 can therefore have involved no more than a partial varnish removal. Six months of retouching then followed, but, considering all that the painting has been through, the amount of loss is less than might be expected. The many small scattered losses can be attributed to flaking, while certain patterns of damage atypical of panel paintings, for example the jagged lines through the group of bystanders in the background on the left, can be attributed to accidents in the original transfer process. The delicate condition of the painting meant that some of the older and relatively insoluble restoration, including the repainting down the join that ran through the figures of Lazarus and Martha, was not removed in the most recent cleaning; where it was very discoloured it was covered by new retouching. The paint layers of some of the figures are damaged by abrasion, particularly in the lower part of the picture (for instance the figures supporting Lazarus). In common with other paintings of its age, it is likely to have been cleaned wholly or partially on several occasions before its recorded conservation history; the distribution of the damage suggests that these first cleanings (including the removal of the varnish that we know from the documents was applied by Sebastiano) are likely to have taken place while it was still on the altar in Narbonne.