• 1 Vandivere and Pottasch 2021, p. 6–7.

1The Lamentation of Christ (c. 1460) by Rogier van der Weyden and his workshop is the earliest painting in the Mauritshuis collection, and the only painting by Rogier in the Netherlands (fig. 1a–1b). During the exhibition Master of Passions (De Passie van de Meester, Museum M, Leuven, September-December 2009), on the occasion of which the Lamentation was displayed alongside other paintings related to this artist, differences in aesthetic condition became evident. The yellowed varnish layers and discoloured retouchings on the Lamentation were regarded as visually disturbing enough to justify a primarily aesthetic restoration. Between June and September 2018, the first phase of treatment – the removal of the varnish – was carried out in front of museum visitors at the Mauritshuis, as part of the temporary exhibition Rogier van der Weyden Unveiled.1 Afterwards, the treatment continued in the museum’s conservation department.

• 2 Pottasch et al. 2023.

2The treatment of the Lamentation offered an opportunity for a comprehensive scientific investigation. A combination of non-invasive and micro-sampling techniques provided insights into the original materials and methods used by Rogier van der Weyden and his workshop. Materials from earlier interventions were also analysed.2 Multi-scale optical coherence tomography (MS-OCT) proved particularly effective for imaging the stratigraphy of semi-transparent layers (see section 2). This non-invasive technology supplied information about non-original varnishes and original glazes, especially when combined with data from cross-sections. It also supported the restoration process by indicating the number of layers likely to be encountered during varnish and overpaint removal.

• 3 Hall 1987, p. 99.
• 4 Everaarts 2020.

4The scene of the Lamentation is set in a stark landscape at Golgotha (Calvary), but it also incorporates European alpine elements, such as castles and snow-covered mountains. Christ’s body has just been taken down from the cross and is surrounded by mourning figures, supported by Joseph of Arimathea at his head and Nicodemus at his feet.3 In the centre, the Virgin Mary is comforted by Saint John. On the left, the three Marys are depicted weeping. Standing up, Mary Magdalene can be recognised by her ointment jar. On the right are Saints Peter and Paul, identified by a key and a sword, and an unidentified kneeling donor.4 The depiction of grief and emotion, including realistic tears and religious devotion, is quite striking.

• 5 Dhanens and Dijkstra 1999, p. 13.
• 6 Dhanens and Dijkstra 1999, p. 163; Steyaert 2012, p. 112–113.
• 7 Dhanens and Dijkstra 1999, p. 157.

5Rogier van der Weyden was born in Tournai in 1399 or 1400, where he possibly trained and worked alongside Robert Campin.⁵ In 1432, he became a member of the Guild of Saint Luke in Tournai. Shortly afterwards, he moved to Brussels, where, in or before 1443, he bought two large properties that he lived in and likely used as his workshop.⁶ In 1436, he was referred to as the town painter, and it was decided that after his death, he would not be replaced.⁷ Although none of his works are signed or dated, Rogier van der Weyden’s artistic output can be divided into three periods. His early works, created before and around 1440, include the Descent from the Cross (before 1443, Madrid) and the Miraflores Altarpiece (c. 1440–45, Berlin). In his mid-career, he probably painted the Last Judgement (c. 1445–50, Beaune) and the Braque Triptych (c. 1452, Paris). Late works, such as those produced in the last five years of his life, likely include the Medici Madonna (c. 1460, Frankfurt) and the Deposition (c. 1460-64, Florence). Rogier died in Brussels in 1464.

• 8 Delsaerdt 2023, p. 150 ‘The complex included a chapel, in which a panel with a Lamentation, probabl (...)
• 9 In 1805, the painting belonged to P.J. Geedts (see Broos 1987, p. 403 and p. 409, n. 10). In 1817, (...)
• 10 Broos 1993, p. 331, n. 9.
• 11 Waagen 1847, p. 177; Broos 1993, p. 327 and p. 331, n. 9.
• 12 Waagen 1847, translated from German by the authors.
• 13 Friedländer 1924, p. 106.
• 14 Van Suchtelen 2018, p. 8–9; Kemperdick 2009, p. 121. Kerber 1972 placed it early in the artist’s oe (...)
• 15 Campbell 2009, p. 124 and Steyaert 2013, p. 101 connected the Lamentation to a workshop assistant: (...)

6For at least the last two centuries, there has been no consensus regarding the authorship and date of the Lamentation. The painting is likely identical to the work described in a 1609 inventory of the Atrecht College in Leuven.8 It does not reappear until 1805 in the collection of Joseph Pierre Geedts (1770–1834), where it was attributed to Jan van Eyck. In 1817, it was acquired by Baron Charles-Louis van Keverberg van Kessel (1768–1829), but in his book, he changed the attribution to Hans Memling (1430/40–1494).9 It arrived at the Mauritshuis when it was acquired by King Willem I in 1827.10 Gustav Waagen was the first to attribute it to Rogier van der Weyden shortly thereafter.¹¹ Based on the ‘less warm colouring’ (1847) and ‘a rich composition with moving expression of the faces and excellent execution, though somewhat cooler in the flesh tone’ (1862), Waagen dated the Lamentation to the ‘somewhat later period of the master’.12 Friedländer (1924) regarded the Lamentation as a ‘well-preserved [late] masterpiece’ painted around 1460.¹³ Since then, a debate has developed about when in Rogier’s career the Lamentation was painted; suggested dates range from his early to his late career, between 1440 and 1464, or possibly finished posthumously by his workshop.14 Furthermore, there remains uncertainty regarding the extent of Rogier’s involvement in creating the Lamentation, or whether his assistants wholly or partly executed it.15

• 16 Pottasch and Derks 2024, p. 17–23; Dyballa and Kemperdick 2019; Pottasch and Vandivere, Many Hands (...)
• 17 Panofsky 1951, p. 39; Panofsky 1953, p. 284.
• 18 Châtelet 1999, p. 236–237 assumed Rogier’s son Pierre finished it after the master’s death.
• 19 Panofsky 1951, p. 37; Kemperdick and Eising 2023, p. 111; Châtelet 1999, p. 236–237.
• 20 Kemperdick and Eising 2023, p. 110–113.
• 21 See, for example, Kerber 1972, p. 297 and Broos 1993, p. 328.
• 22 Mund 2009, p. 512; De Vos 1999, p. 403, who also states that the composition is ‘not one of the bet (...)
• 23 Pottasch and Derks 2024, p. 17–23: Pottasch and Vandivere 2019 ‘Many hands make light work: Evidenc (...)

7As was common practice in late medieval workshops, several different hands may have contributed to the Lamentation. Even within Rogier van der Weyden’s core oeuvre, the work of assistants or pupils has been identified in the stages of preparation, underdrawing, and paint application.16 Panofsky (1951–53) was found it difficult to discern Rogier’s hand in the ‘somewhat mannered in postures and gestures’ of the Lamentation.17 He identified the donor in the lower right corner as Pierre de Ranchicourt (bishop in 1463, died 1499), which would have placed the date of production in the last years of Rogier’s career, or even after he died in 1464.18 Panofsky thus proposed that the painting might have been begun by Rogier and finished by Memling.19 Recently, Kemperdick (2024) expanded on this idea, albeit proposing the involvement of Hugo van der Goes, as a similar depiction of Mary Magdalene appears in his Portinari Altarpiece (1472–48).20 Conversely, other scholars such as Kerber (1972) and Broos (1983) attributed the flaws in the Lamentation to the inexperience of the young Rogier himself, placing it early in his career.21 In recent scholarship, Rogier van der Weyden’s involvement is generally acknowledged, though it is also accepted that some parts of the Lamentation were executed by assistants in his workshop.22 Technical examinations suggest that several workshop assistants contributed to different phases of the underdrawing and paint layers.23

• 24 De Vos 1999, p. 404–405. He compared it to the figures on the side-panels of the Sacraments Altar ( (...)
• 25 Everaarts 2020, p. 40.
• 26 Everaarts 2020, p. 48–61.

8Regarding the dating of the Lamentation, the identity of the kneeling donor figure in the lower right corner was also reconsidered. As mentioned earlier, Panofsky identified him as Pierre de Ranchicourt (bishop in 1463, died 1499), thereby dating the painting as a late work, possibly completed after Rogier’s death. De Vos (1999) proposed Jean Chevrot (1395–1460) as the donor.24 More recently, Everaarts (2020) argued that the figure wearing a mitre, dressed in a gold-brocaded cope and holding an ornate crystal and gold crozier, can be identified as a bishop.25 Her suggestions – either Jean de Bourgogne (1414/17–1480, bishop in 1440) or David de Bourgogne (c. 1427–96, bishop in 1456) – are convincing. Their consecration dates support an earlier dating of the Lamentation, possibly to the 1440s or 1450s.26

• 27 Klein 1992, Table III, p. 44–45, and Klein 2009, p. 164
• 28 The support of the Lamentation was made from Baltic oak that was radially sawn. It comprises six ve (...)
• 29 Van Asperen de Boer et al. 1992, p. 180.
• 30 Billinge et. al. 1997, p. 18.

9Scientific evidence from technical examinations can influence the debate about attribution and dating. Dendrochronology (tree-ring dating of the wood panel) suggests a felling date of 1423, indicating that the Lamentation could have been painted from 1438 onwards.27 Recent reconsideration of the seasoning time (shortened from 10 years to 2 years, depending on where in the Baltic area the trees grew) suggests the panel could have been used earlier: from 1430 onwards. The commission for the Lamentation must have been significant, as high-quality wood was used for the panel, which remains in pristine condition.28 Van Asperen de Boer (1992) noted that it was remarkable that the wooden planks were arranged vertically, especially for a composition wider than it is tall.29 He also proposed that, despite the early felling date, this unusual format implied a later use of the wood. However, technical research by Spring (1997) identified several examples of large paintings where planks were arranged vertically.30 The horizontal format of the Lamentation, combined with the vertical arrangement of the planks and the absence of wings, is comparable to an early work by Rogier: the Prado Descent from the Cross (before 1443, Madrid), although the latter has a larger format.

• 31 For comparative studies of artworks from The National Gallery, London, see Spring 1997; Spring 2012
• 32 Specifically, the underdrawing of Mary Magdalene in the Lamentation closely resembles that in the D (...)
• 33 De Vos 1999, p. 403.

10Other results from the technical examination of the Lamentation align with findings from paintings created early in Rogier’s career.31 Using infrared reflectography, Pottasch and Derks (2024) identified notable similarities in the underdrawings of several early works, including the Descent from the Cross, the Miraflores Altarpiece (c. 1440–45, Berlin), and the Middelburg/Bladelin Altarpiece (c. 1445–50, Berlin).32 Some painted details were also similar; for instance, the gentle features of Mary Magdalene in the Lamentation resemble those in earlier paintings, as does her hairstyle, which is similar to that in the Descent from the Cross. Additionally, scholars such as De Vos (1999) have pointed out stylistic parallels with other early works, including the Exhumation of St Hubert (late 1430s, London) and the Last Judgement (1445–50, Beaune).33

• 34 Pottasch and Derks 2024, p. 23.
• 35 Pottasch et al. 2024. MA-XRF was carried out using a Bruker M6 Jetstream scanner.
• 36 Pottasch et al. 2024.
• 37 Koelemij 1984.
• 38 Pottasch et. al. 2024.
• 39 Vandivere, Pottasch and Kneepkens 2021.

13The restoration treatment began with a visual examination and scientific investigation using non-invasive imaging techniques such as raking light, ultraviolet radiation (UV), and X-radiography. These technologies provided insights into the painting techniques and materials used to create the Lamentation. Infrared reflectography (IRR) revealed that the underdrawing was applied in several stages, likely by different hands or with corrections to earlier lines.34 Further non-invasive imaging was conducted using macro-X-ray fluorescence scanning (MA-XRF), which mapped the distribution of chemical elements across the composition.35 Fiber optic reflectance spectroscopy (FORS) and reflectance imaging spectroscopy (RIS) offered additional information about specific pigments and mixtures.36 Micro-samples taken in 1983, along with additional samples taken during the recent treatment, were examined as cross-sections using light microscopy.37 A selection of these cross-sections was analysed with scanning electron microscopy/ electron dispersive X-ray spectroscopy (SEM-EDX).38 Based on this scientific research, reconstructions of select areas were created using historically appropriate materials, enhancing understanding of the properties and techniques employed in making the Lamentation.39

• 40 MS-OCT was carried out using a Thorlabs Ganymede-II-HR spectral-domain MS-OCT system. OCT indicated (...)
• 41 Cheung, Spring and Liangl 2015; Callewaert et al. 2020; Vandivere et al. 2019.
• 42 Raised paint was mainly located in the green-brown paint of the landscape, Mary Magdalene’s blue cl (...)
• 43 Evolon® is a highly absorbent microfilament textile containing polyester and polyamide. It is produ (...)

14Prior to the treatment, the non-destructive scanning method of multi-scale optical coherence tomography (MS-OCT) was particularly useful. It provided insight into the number and thickness of varnish layers and uncovered undocumented treatments from the past.40 OCT is used in the medical field to measure thin layers in the human eye, where dark lines in the acquired image represent boundaries between layers. In paintings, it reveals the application of translucent and semi-translucent layers, such as varnishes and glazes, and differentiates them from more opaque layers like retouching.41 This enabled conservators to proceed as safely as possible with varnish removal, equipped with knowledge about the number of varnish and retouching layers they were likely to encounter. They found that the thickness and number of varnish layers varied; some cross-sections showed two, while others had up to five. In certain areas, additional varnish had been applied to saturate porous or underbound paint and – intentionally or unintentionally – to consolidate raised paint. After testing various mixtures of mildly polar organic solvents, the Mauritshuis conservators chose an approach that minimised mechanical contact with the original paint surface, especially where brittle and raised paint was present.42 The solvent mixture was gently applied to the surface with a cotton swab rolled on small pieces of Evolon® tissue.43 This technique proved highly effective, as the tissue absorbed the varnish evenly and reduced the mechanical action caused by swab-rolling. Notably, the tissue absorbed dark varnish within the cracks through capillary action, making the craquelure pattern less dark and visually less intrusive. Using this method, the varnish layers and most of the discoloured retouchings on the Lamentation were removed simultaneously. Generally, the retouching had been applied between old varnish layers – a phenomenon observable using MS-OCT and cross-sections. Scans within cleaning tests confirmed that only non-original varnish layers and retouchings were being removed by the chosen solvents, without affecting the original glazes. In Saint Paul’s robe (described in the following section), the combination of MS-OCT and information from samples mounted as cross-sections proved valuable in distinguishing original glazes from layers of later-applied varnish and retouching.

• 44 In all cross-sections taken from red draperies containing red lake, colourless particles were obser (...)

16Before treatment, Saint Paul’s red mantle (on the far right) appeared slightly patchy and flat, with relatively undefined folds (fig. 2a). When examined under UV light, the surface of his clothing showed varying fluorescence, which suggested uneven layers of varnish, as well as dark areas that indicated remnants of retouching (fig. 2b). An MS-OCT scan along a line confirmed that multiple varnish layers and opaque retouching had been applied over a layer of translucent original glaze (fig. 3a–b). The uppermost dark line in the OCT cross-section marks the surface/air interface; beneath it, a white layer indicates varnish, followed by several light grey layers containing retouching pigments. The original glaze layers are distinguishable from these non-original materials and appear darker. A sample taken along the same line confirmed this complex stratigraphy (fig. 3c–d). On top of three original red layers containing red lake and vermilion, the cross-section reveals multiple layers of varnish.44 Dirt and retouching are present between the varnish layers, showing that time had passed between applications. The pigments in the retouching layers – vermilion and red lake – were very similar in composition to the original pigments and might have been applied as early as the sixteenth century.

• 45 For information about drying cracks, see Brury and Bratasz 2024. A similar phenomenon was present t (...)
• 46 In other areas of the painting, retouching was applied directly on top of original paint: for examp (...)

17Once the varnish had been carefully removed from the surface of Saint Paul’s mantle, it became clear why it had been overpainted in the past. The upper layer of original paint showed pronounced drying cracks, probably caused by an excess of medium or slow-drying paint in the upper layers.45 Additionally, the red lake glazes on the surface had almost entirely faded, especially affecting the shadows, which were created from thin layers of red lake applied over a red underlayer. The fading of these glazes reduced the three-dimensional modelling of Saint Paul’s garment, making it look relatively flat. Furthermore, once the varnish was removed, older retouchings became more visible. The ageing of the binding medium in the red retouchings caused them to darken over time; unfortunately, because these very old retouchings had been applied directly onto the vulnerable original paint without an intermediary varnish, they could not be removed safely during the current treatment (fig. 2c).46

• 47 The Hague, Netherlands Institute for Art History (RKD), M.J. Friedländer Archive, 0615: 264 Rogier! (...)

18Another area where historical retouchings affected the appearance of a garment was in the lower left corner: the skirt (part of the dress) of the seated Mary. Before the recent treatment, the seated Mary wore what appeared to be a blue dress; however, a substantial area in the middle exhibited a divergent crack pattern and an uneven surface (fig. 4a). This appearance was not recent; when Friedländer saw the Lamentation in the early 1900s, he noted an area of retouching in his notebook: ‘264 Rogier! Condition of half of the garment seated Magdalena wrong.’47

• 48 Two methods were used to imitate the crack pattern on the skirt to conceal the areas of retouching; (...)
• 49 The materials in the fills and paint layers were identified using a combination of MA-XRF, X-radiog (...)

19Scientific technologies utilised to examine the painting before the recent treatment offer further insights into the original and non-original materials in the lower part of Mary’s dress. Under raking light, part of the skirt exhibited an uneven surface, with cracks differing from the surrounding paint (fig. 5a-b).48 A significant portion of the skirt (approximately 7 by 7 cm) was identified as non-original. Although the painting had been X-rayed in the 1970s, this retouching was not documented in the literature. The X-ray revealed that damaged areas of the skirt had been filled; some of the older fills contain chalk, whilst later fills, which appear white in the X-ray (fig. 4b) and brown under UV light (fig. 4d), contain lead.49 On top of both fills, the skirt was overpainted with a layer of paint containing lead white and smalt. Since smalt was not used in the original paint layers but was most common as a pigment in sixteenth- and seventeenth-century easel paintings, the later intervention on Mary’s skirt probably dates from this period. Additionally, a layer of grey retouching applied directly over the chalk fills appears to be even earlier.

• 50 Winkler 1913, p. 132 (translated from German by the authors).

20The discoloured and uneven appearance of this overpaint – and the fact that X-radiographs showed it covered a significant amount of original paint – led to the decision to attempt removing the retouching during the recent treatment. After removing the overpaint and filling material that extended over the original paint, a waxy substance and black spots became visible on the surface (fig. 4c, 5c). This discovery marked an important moment in the painting’s history, related to its function in its early centuries. A similar area of damaged paint was found at the same height and position on the right side of the painting: in the foot of Nicodemus (fig. 6). His foot was also overpainted; Winkler (1913) wrote ‘only when you look closely do you discover weakness, as in the man holding a corner of the shroud: his feet are narrow, and he is quite unsteady. He is in danger of tipping forwards.’50

• 51 An illustration of the proposed positions of the candles and painting on the altar is included in P (...)

21The location of these two damages strongly indicates that the dark material contains soot and wax: the result of damage caused by candles placed on an altar in front of the painting. The original paint had become heated and slightly singed, leading to cracking and flaking. These localised damages from candles, in an otherwise well-preserved painting, confirm that it was originally used for devotional purposes or as an epitaph.⁵¹ The tradition of burning candles in front of the Lamentation continued for a considerable period, as shown by two campaigns of filling and retouching. This example demonstrates that damages uncovered during treatment can offer important insights into the painting’s function and history.

• 52 Similar folds can be found in Rogier van der Weyden’s Triptych with the Crucifixion (c. 1440, Kunst (...)

22After removing the retouching and overpaint from Mary’s skirt, the surface of the original paint was examined using a 3D digital Hirox microscope. This enabled the conservators to observe the original pigments and revealed that the overpaint had covered parts of the original paint, protecting them from light over the centuries in specific areas (fig. 5d). The protected regions appeared more purple and pinkish compared to the rest of the skirt. The presence of lead white, azurite, vermilion, and faded red lake in two paint layers suggests that the skirt originally had a somewhat more purple hue. Additionally, remnants of highlights, midtones, and shadows indicated that the folds were once painted more linearly, with sharp creases and higher tonal contrasts, emphasising the weight of the fabric. Rogier van der Weyden painted skirts with similar folds in other works, and later copies of the Lamentation also align more closely with these original forms (fig. 4f).52 These served as valuable visual references during the retouching phase of the restoration.

• 53 For an overview of all pigments, see Pottasch et al. 2023.

24Once the overpaint had been removed, delicately painted details were revealed in the original paint of Nicodemus’s shoe and stockings, as well as in the gilded knob and crystal staff. Also, highlights that had been applied around his foot proved to be later retouchings; they were painted with a yellow earth pigment instead of the lead-tin yellow that was originally used for highlights in the rest of the Lamentation.53 The findings and discoveries made prior to and during the conservation treatment showed evidence of the early history of the painting. Damage from candles on an altar were filled and retouched twice an early stage in the painting’s life, after which multiple layers of varnish were applied.

• 54 The original paint layers of the Virgin Mary’s clothing are built up from a layer of azurite (and a (...)
• 55 Spring 2012, p. 97.
• 56 Gambardella et al. 2020. A sample from the Lamentation examined with micro-X-ray absorption near-ed (...)

25Other retouchings on the Lamentation might have been carried out even earlier than those in Mary’s skirt. In the blue garment of the Virgin Mary in the centre of the composition, a cross-section revealed that it had been retouched with a paint containing both ultramarine and red lake, alongside a small loss.54 This pigment combination was commonly used in fifteenth-century Netherlandish art.55 In this case, it was probably employed to match the rich blue tone of the surrounding original paint, which contained large particles of ultramarine.56 Since the choice of retouching pigments is similar to this fifteenth-century practice, it could have been applied early in the painting’s history. Furthermore, the refinement of the retouching and the utilisation of high-quality materials like natural ultramarine suggest that the painting was held in high regard. This new information bridges a gap in the timeline of the Lamentation, as these early restorations were not previously documented.

• 57 In 1841, the surface dirt was removed, the varnish was thinned mechanically, and an additional laye (...)
• 58 The 1957 yearly report (jaarverslag) of the Mauritshuis records that Traas ‘cleaned and restored’ t (...)
• 59 Kisner et al. 2023.
• 60 Susijn 1973.

26The first recorded restoration of the Lamentation occurred shortly after its acquisition in 1817 by Baron Charles-Louis van Keverberg van Kessel (1768–1841). It was restored by Mattheus Ignatius van Bree (1773–1839), who was also a painter. After the Lamentation was acquired for the Mauritshuis by King Willem I in 1827, annual reports indicate that Nicolaas Hopman (1794–1870) carried out minor treatments in 1841 and 1852, which involved removing surface dirt and applying additional varnish.57 In 1957, a restoration was undertaken by Jan Cornelis Traas (1898–1984), who removed surface grime and selectively thinned the old varnish layers to prevent over-cleaning, consistent with his generally careful approach.58 He also improved the painting’s appearance by toning older discoloured retouchings to match the surrounding original paint. It is highly probable that the painting still retained a warm ‘gallery tone’: a tinted varnish that would diminish minor contrasts or colour differences.59 In 1973, J.J. Susijn (1934–2018) performed a structural intervention, reinforcing the joint between the top of the two right planks of the panel with glue and consolidating raised paint.60 It is likely that he adjusted the colour of the retouching along the treated joint using an oil-bound paint.

• 61 Ainsworth 2008.
• 62 Campbell 2009, p. 124 (p. 122 in the English edition).
• 63 Borchert 1995, p. 171.
• 64 Ainsworth 2008.

27At some point in the painting’s history, gilded round haloes were added around the heads of the saints. Yet, haloes of this kind rarely appear in fifteenth-century Netherlandish paintings.61 In 2009, Campbell suggested that ‘the circular haloes sit rather uneasily on the heads of the apostles John, Peter and Paul.’62 Microscopic examination of the saints’ haloes revealed that they are not original; they differ in colour and character from the original gilded details (described below) and more closely resemble the material used for retouching (fig. 7a–c). A notable aspect of these haloes is their precise size; the diameter of Peter’s halo is 9.5 cm, while the other three are 10 cm. Their exact conformity to the metric system suggests they were probably added in the nineteenth century. Considering the history of the painting, it is likely that they were added by Mattheus Ignatius van Bree, who restored the painting at Keverberg van Kessel’s request before 1824.63 He probably believed it was by ‘Hemmelinck’ (Memling); however, such haloes are uncommon in works by Memling or Rogier. Their addition may have been linked to a change in taste or function; such modifications occurred more often in early Netherlandish paintings.64

• 65 The shape of the rays is the same as those of Christ and Mary on the Bladelin altarpiece (c. 1450, (...)
• 66 Billinge et al. 1997, p. 30 describe that gilding in early Netherlandish paintings was usually wate (...)
• 67 The non-original haloes had also been retouched with a lighter shade of gold.

28Some of the gilded details in the Lamentation are original to the painting and differ significantly in material and appearance from the round haloes described above. Both Christ and the Virgin Mary have fine golden rays surrounding their heads, symbolising their high religious status. Mary’s rays fan out around her white headscarf, while those of Christ extend from his head on three sides in the shape of a cross with seven rays each (fig. 7d).65 Microscopic examination confirmed that these are original to the painting and made of fine, sparkling gold.66 A similar fine powdered gold was found in the gilded decorations on the hems of the dresses and mantles of the three Marys (fig. 7e). Mary Magdalene’s clothing features a fine line along its contour, while small floral motifs embellish the borders of the Virgin’s dress and the seated Mary’s purple mantle (fig. 7f). The gold decorations show signs of wear. At some point, they were retouched with a warmer tone of gold, which looks more like paint than powder, and extends into the cracks of the original paint layer.67

• 68 Vogelsang 1949, p. 10.

29The recent treatment allowed for an in-depth examination of the painting, revealing many aspects of the original technique associated with Rogier van der Weyden and his workshop. After the final varnish was applied following retouching, the painting not only regained much of its original colouring but also displayed incredible details such as subtle modelling and numerous tears on the faces of the saints, heightening the scene’s dramatic tension. In 1949, Vogelsang stated that ‘the seated Mary does not weep’; perhaps, at that time, the tears were not visible through layers of yellowed and cracked varnish.68 A comparison of cleaned and uncleaned areas of her face during varnish removal shows how one of the tears had been partially obscured before treatment (fig. 8). Furthermore, subtle nuances in the faces and colour contrasts in the draperies had been affected by the overall murkiness of the varnish and dirt layers.

• 69 Panofsky 1951, p. 39, n. 40; Panofsky 1953, p. 284; Waagen 1847, p. 177.
• 70 Waagen 1847, p. 177.
• 71 Vogelsang 1940, p. 50.
• 72 Panofsky 1951, p. 39, n. 40.

30Several art historians who wrote about the Lamentation over the centuries based their judgments and conclusions on the colours they observed in the painting.69 Some might have seen it through a veil of yellowed varnish; others may have relied on black-and-white photographs. Waagen, who published an article describing the painting in 1847, would have seen the Lamentation relatively soon after a treatment that took place between 1817 and 1827. He wrote, ‘The picture, with its less warm colouring, probably belongs to a somewhat later time of the master.’70 Vogelsang reiterated the link between the perceived colour of a work and its dating in 1940, associating a particular brown hue (described below) with a later date within Rogier van der Weyden’s oeuvre.71 In a 1951 publication, Panofsky mentioned the ‘pleasing colouring’ but concluded that the painting was ‘grandly conceived but un-Rogerian in colour.’72

• 73 Plesters 1993; Pottasch et al. 2023, p. 472.
• 74 Keverberg van Kessel 1818, p. 164.

31Rogier van der Weyden used two types of blue pigments for painting draperies: azurite, a copper-based blue mineral, and ultramarine, derived from the semi-precious stone lapis lazuli. Unfortunately, both pigments have deteriorated and changed colour over time, affecting the three-dimensionality and modelling of the blue fabrics. After varnish removal, it became evident that the draperies primarily painted with azurite – including Mary Magdalene’s robe and Saint Peter’s mantle – had undergone irreversible alteration with age. The paint has darkened to an almost black hue, resulting in a loss of modelling.73 Keverberg van Kessel, who owned the painting in the early nineteenth century, described Mary Magdalene as wearing a ‘dark-coloured garment, which nevertheless has a great elegance and much richness’74 (fig. 9a).

• 75 Vandivere, Pottasch and Kneepkens 2021, section 3a. Reconstructions made using historically appropr (...)

32Examination of cross-sections revealed that Mary Magdalene’s robe was originally a deep, intense dark blue (fig. 9b–c). In the shadow, it was painted with an undermodelling composed of large azurite particles covered by a thin glaze containing fine ultramarine particles. The highlights included a small amount of lead white, but the primary modelling was achieved by varying the thickness of the blue glaze on top.75 Saint Peter’s original blue cloak has also darkened with age. In a cross-section from the shadow, the top layer contains coarse dark blue azurite particles, while the finer azurite particles in the underlayer have a greenish tinge. Lead white was added to the underlayer to create highlights; initially, it would have been partially visible through the semi-transparent upper layers, but this effect has been lost over time. The painted shadows are also less discernible; they were once painted with thick, dark lines, but these now have a similar colour to the surrounding darkened paint. Unfortunately, this colour change is irreversible. During recent treatment, damages were retouched with a mixture of blue and black pigments to match the surrounding paint.

• 76 Spring 2012, p. 95. This layering is typical for Early Netherlandish paintings, including those by (...)
• 77 Pottasch et al. 2023 p. 469; Vandivere, Pottasch and Kneepkens 2021, section 3a.
• 78 Pottasch et. al. 2023.

33Other blue areas of the Lamentation have also experienced colour changes, including the Virgin Mary’s robe and cloak. Before and during the restoration, her clothing looked quite flat, as if the modelling of highlights, midtones, and shadows had been damaged (fig. 10a). Examination of a sample mounted as a cross-section revealed it was painted with an underlayer of azurite, topped with a costly natural ultramarine.76 The shadowed areas, in particular, appeared light and lacking saturation. A glaze containing large particles – up to 20 µm in diameter, indicating high-quality ultramarine – was applied on the surface. These glazes have been most impacted by the phenomenon known as ‘ultramarine sickness’.77 Prior to degradation, they would have contributed to a greater sense of depth and colour intensity. The surface now scatters light and therefore appears blanched. This diminishes the contrast between dark and light blue areas mixed with lead white, which are better preserved. This is evident in the MA-XRF map for lead (fig. 10b).78 During the recent treatment, a very thin, translucent glaze was locally added to the most affected areas to restore three-dimensionality (fig. 10c).

• 79 Pottasch, Vandivere and Kneepkens 2021, section 3b.
• 80 Vogelsang 1940, p. 50. Translation from Dutch by the authors.
• 81 Vogelsang 1949, p. 14
• 82 Vogelsang 1949, p. 14.
• 83 Kemperdick 2013, p. 121.

34On the left in the background, the standing Mary at the cross is wringing her hands in sorrow (fig. 11a). On top of a kirtle with a warm orangish colour,79 she wears a brown fur-lined garment. The brown garment is unusual in Rogier van der Weyden’s oeuvre. In 1940, Vogelsang wrote about the ‘somewhat strange colour here: that sombre, ochre-like brown in the robe of one of the women. This terracotta does not appear elsewhere on Rogier’s palette. But based on such a remark, one would only have to proceed to a late dating, not to a deletion of the piece from Rogier’s oeuvre.’80 In 1949, Vogelsang seems to have changed his mind and associated the colour with Hans Memling, using this as one of the arguments for why the painting might have been made through reciprocal influence or collaboration between Memling and van der Weyden.81 He described her clothing: ‘in which a colour occurs that we never find in such an important place in Rogier’s palette: that remarkable distinguished brown of the fur-lined overgarment. That is a tint and a nuance that, on the contrary, often occurs in much of Memling’s work.’82 In 1997, Kemperdick remarked: ‘nor do the brown and orange tones in the robe of the saint wringing her hands behind the Magdalene resemble Rogier’s typically cooler colouring.’83 These observations were made through yellowed varnish, since the painting had not recently been treated.

• 84 Pottasch et. al. 2023, p. 19.
• 85 Other brown areas of the painting, such as the cross, the ladder and parts of the landscape, were p (...)
• 86 Vandivere, Pottasch and Kneepkens 2021, section 3b.

35A recent examination of paint samples mounted as cross-sections showed that this brown colour was probably not what the artist intended (fig. 11b). The paint is built up in three layers, as is typical for most of the draperies. The bottom layer contains lead white, black, with small additions of red lake, vermilion, and blue particles. The middle layer appears translucent reddish-brown in the cross-section and contains mostly red lake and some bluish particles. Examination with FORS suggests the bluish particles may be verdigris, which could have been added as a siccative for the slow-drying red lake.84 The top layer includes vermilion, black, lead white, and a somewhat discoloured red lake. There is no brown pigment present.85 Given the pigments used, it seems unlikely that the dress was meant to be brown but rather more burgundy coloured. A reconstruction using historically appropriate materials based on the pigments resulted in a deep reddish-brown colour.86 Nicodemus’s stockings also contain a similar pigment mixture that now appears brownish (fig. 6).

• 87 Spring 2012, p. 97.
• 88 For example, Portrait of a Lady attributed to the workshop of Rogier van der Weyden (c. 1460, The N (...)

36The knowledge that the ‘brown’ garment was originally more reddish aligns with other draperies in his oeuvre. His paintings frequently feature a range of red tones, almost all created using the same pigments – vermilion, red lake, and lead white – applied systematically.87 There are, however, other ‘brown’ dresses in works more closely linked to his workshop.88

• 89 To saturate the colours of the unsaturated paint surface after varnish removal, an intermediary var (...)

37Following the recent restoration at the Mauritshuis (2016–21), the dazzling colours, crisp details, and impeccable technique of Rogier van der Weyden can be appreciated once again (fig. 1b).89 The revelation of convincing three-dimensional modelling makes the figures appear more robust, and the visible tears have given them more emotional impact. New information from dendrochronology and the analysis of paint layers supports an earlier date in Rogier’s oeuvre. Thoughts on the identity of the bishop donor were recently expanded to include two bishops consecrated in 1440 and 1456, allowing for a possible earlier date. One would expect an important altarpiece commissioned by a bishop to involve the master. The Lamentation was most likely painted by Rogier van der Weyden with assistance from his workshop, around 1440–50. Art historians and scholars are encouraged to look at the Lamentation anew, now that the painting has been treated.

• 90 Van Asperen de Boer et al. 1992 p. 179; Billinge et al. 1997; Spring 2012; Pottasch et al. 2023.
• 91 Panofsky 1951, p. 39 and n. 40; Waagen 1847, p. 174.

39The scientific examination confirmed that all of the pigments in the Lamentation are consistent with Early Netherlandish painting, and their use and layering – usually in three layers, except for the blue colours – align with the painting techniques of Rogier van der Weyden and his workshop.90 The high quality of the blue pigments – such as ultramarine and azurite with large particles – demonstrates that the Lamentation was an important commission. It also led to a better understanding of the original appearance of the Lamentation: specifically, how the colours have changed over time. An unusual brown colour in the clothing of the standing Mary and Nicodemus was found to contain red lake, which has since lost most of its colour intensity. Especially in the blue draperies, colour change has affected the illusion of three-dimensionality and modelling. These changes have significant implications for art-historical interpretations of the paintings.91 In future, this could influence discussions about attribution and dating, which remain a matter of debate in art historical literature.

• 92 Waagen 1847, p. 177; Kemperdick 2013, p. 121.

40A thorough comprehension of materials and techniques also guided the restoration process. During the recent treatment, old varnish layers and discoloured retouchings were removed. Damages were retouched illusionistically, with the understanding that some colours had changed over time. After the treatment, the cool colours that art historians associate with Rogier van der Weyden are now more visible, bringing it closer to the artist’s original intention.92 The recently restored Lamentation now offers viewers a more accurate representation of Rogier van der Weyden and his workshop’s artistry and craftsmanship.