Jan van Eyck’s frames

by The Frame Blog

 Jan van Eyck (fl.1422- d.1441), his eight works with original frames shown in their relative dimensions

 Introduction

The article on the frames of Jan van Eyck was first published in The Burlington Magazine (July 2026, pp. 653-67). The present text is a slightly revised version, based on the original, unedited manuscript. The format of The Frame Blog is particularly well suited to a study which relies so heavily on visual evidence, allowing many of the observations to be presented in a way which should make them more readily accessible and comprehensible to the reader.

Almost all the ‘discoveries’ recorded here have not previously been documented. The Frame Blog article will be revised and supplemented from time to time as further observations emerge. Many of the sectional drawings are based on images from the Closer to Van Eyck website; these will be checked and verified when direct examination of the actual objects becomes possible.

This is essentially a visual survey, with an emphasis on looking closely and recording what can be observed. The conclusions are therefore provisional and remain open to revision.

The exhibition of Jan van Eyck’s portraits at the National Gallery (21 November 2026-11 April 2027) demonstrates strikingly how differently each likeness is approached, as though the artist were continually searching for a more successful solution. There is no evidence of a fixed formula or routine (with the sole exception that all the sitters are depicted looking to the right). Seen in this light, Jan van Eyck’s motto, Als Ich Can, might be understood not as the self-assured declaration ‘only I can do this’, but rather as an acknowledgement that, at any given moment, the work was as good as he, Jan van Eyck, could make it. The same spirit of continual exploration and inventiveness is evident in every aspect of the surviving original frames.

Almost a third of Jan van Eyck’s paintings have survived with well-preserved original frames: a group of eight works, dating from 1433 to 1439, is discussed in this article[1]. Their frames are highly inventive, exploring a variety of shapes and surfaces. They form a nucleus of patterns recurring in Northern European picture-framing going into the early 16th century. The profiles and proportions of Van Eyck’s designs have been so thoroughly absorbed by his contemporaries and by subsequent generations that they seem almost unremarkable to us. Art historians have focused on the inscriptions which seven of the eight frames bear, and which are undoubtedly the main reason for their survival: the meaning, style and depicted materiality of these scripts are most intriguing and have been widely studied[2], whilst the morphology of the frames and details of their making have largely escaped notice.

Apart from the Madonna & Child with Canon van der Paele, the works in question are very small. Each of the eight frames is uniquely crafted to support a particular illusion of space or material. Unlike the surviving frames of other artists, these are not repetitions which a workshop of cabinetmakers might have produced; indeed, at least four of the eight frames are hand-carved, something which suggests invention and exploration during the shaping of the profiles, possibly even by the hand of the artist, but most certainly under his close supervision.

************

All the drawings for the profile diagrams are by Peter Schade unless stated, and the digital conversions of his drawings are by Bart Cornelis. They are reproduced at the sizes in which they stand in relation to each other, which explains why the profile diagram for the Madonna & Child with Canon Joris van der Paele takes up all the available width of the image, whilst the profiles of the smaller works seem to be isolated on fields of white space.

************

Portrait of a man (Self-portrait?), 1433

 

Jan van Eyck (fl.1422- d.1441), Portrait of a man (Self-portrait?), 1433,  o/oak (?) panel, 26 x 19 cm., National Gallery, NG 222

Van Eyck, Self-portrait, drawing of section of frame, with lost fillet at the back edge shown as dotted line

The earliest painting of this group is the Portrait of a man (Self-portrait?) of 1433. It is the only entirely gilded frame surviving on a work by van Eyck. The surface is of exceptional quality –  most probably double-gilded, which would explain the absence of lay lines, even in the rubbed areas. The lateral elements of the frame are made from the same piece of wood as the panel (‘integral’), while the horizontal mouldings are attached with dowels (‘engaged’). The same construction is used for the Dresden Triptych and the Portrait of Jan de Leeuw, and the profiles of those frames are also similar. Only the Dresden frame has retained its raised outer lip (the dotted line in the drawing above); this contour moulding has been removed from the Self-portrait, which – it seems likely – was originally about 1 centimetre larger on all four sides, making it similar in width to Jan de Leeuw.

 Van Eyck, Self-portrait: (top) the remains of holes drilled through the top rail of the frame; (below) illustrative holes and hanging cord on another frame

 Evidence for the original size are two holes in the back edge at the top of the frame, now largely shaved off, which were probably once used to hang the panel from a loop of cord or leather, a common device throughout the 16th century.  

Van Eyck, Self-portrait, measurements of the pictorial plane

The shape of the pictorial surface is most remarkable. Van Eyck has created his panel as a slight trapezoid, with parallel vertical sides and tapering horizontals, the right side being 5 millimetres taller than the left. The diagonals are equal, suggesting that this perfect geometrical shape is deliberate. It is just on the verge of being noticeable: the Eyckian conceit is that this diversion from the expected is so slight that it has to be noticed to be seen. It seems possible that the trapezoid was created to manipulate the foreshortening of the painted area. Since the nearer edge of a work always appears taller when seen from the side, whilst the further edge appears shorter, then – if the left side is physically 5 mm shorter than the right – the foreshortening effect will be counteracted when seen from the left and exaggerated when seen from the right.

            Van Eyck, Self-portrait, viewed from slightly to the left and slightly to the right

The apparent width of the painted space is visually squeezed when viewed from the left and elongated when seen from the right; the head appears further forward in the apparently narrower space and recedes in the wider one. The photographs above were taken at the same distance from the centre and at the same height; it appears as if the head moves within the interior space. The margins of the composition are very carefully calculated, so that the flamboyant swirl of the turban comes within a hair’s breadth of the top edge; the only exception is the way that the inscription at the bottom apparently runs out of space – this must also be deliberate and could be a calculated sleight of hand to suggest that this is not the work of careful craft, but sublime genius.

Van Eyck, Self-portrait, variation of mouldings in the individual rails of the frame

The reason for the construction of the vertical framing elements out of the same piece of wood as the pictorial panel, with the horizontals attached by three dowels, is seemingly to enable the use of planes to smooth the area of the image and to run the fine mouldings. The flat surface of the painting was most certainly made in this way, but closer study of the frame profiles reveals that they are entirely hand-carved. The steps, hollows and half rounds are shaped with gouges; this was not done with the anonymous ease of the professional carpenter, but with the hesitant exactness of the competent amateur. For its size the moulding undulates wildly (as above), differently from the frames on Jan de Leeuw and the Dresden Triptych which are based on the same pattern. It is the earliest surviving frame of this type, and the unusual portrait might have prompted its invention.

From the interior sight edge the profile consists of a small sharp hollow, followed by a fractionally wider hollow which ends in a narrow-slanted shoulder; then a convex moulding, which is of a similar radius to the preceding hollow and deeply incised behind the groove, finishing in a flat top section of similar width to the carved mouldings. The frame was originally slightly wider, with the additional raised edge mentioned above, and shown as a dotted line. The main purpose of this complex shape is the illusion of depth it creates: in slanting light the mouldings have nine distinct shades, each claiming its own increment of spatial depth.

This device is so convincing that even the most careful observer attempting to draw the profile is easily misled; the profile drawing above is from Lorne Campbell’s notes in the dossier at the National Gallery[3].

 

The Annunciation Diptych, 1435

 

Jan van Eyck (fl.1422- d.1441), Annunciation Diptych, 1435, o/panel, 38.8 x 23.2, and 39 x 24 cm., Museo Thyssen-Bornemisza, Madrid https://www.museothyssen.org/en/collection/artists/eyck-jan-van/annunciation-diptych

Van Eyck, Annunciation Diptych, drawing of section of frame

In the Annunciation Diptych, the trompe l’oeil white marble niche is set within a physical frame having an ogee profile with a chamfered inner sight edge and a flat top section (above). The lateral elements of the frame are made from the same piece of wood as the panels; the horizontal elements are attached with three dowels.

Van Eyck, Annunciation Diptych, details in which the real and fictive framing elements collide

This narrow three-dimensional frame was painted with red faux marbling at the same time as the panel itself. We can observe the red paint overlapping the white trompe l’oeil niche (detail, top), whilst the white socles of the figures extend over the three-dimensional red frame (detail, below). The interplay of the actual red frame and the fictive painted niche is striking. Jan van Eyck’s forensic observations of ageing materials is demonstrated by the hardly visible painted gap between the polished stone background of the niche and the fictive ledge on which the socles stand [4].

Van Eyck, Annunciation Diptych, details showing the trompe l’oeil painted mason’s joints at the top right of each panel

He seems to be developing aspects of the illusion during the painting process itself: on the right side of the Angel’s niche (detail, top) the painted stonemason’s joint was at first drawn straight, as if seen at eye level, and then corrected to a dramatic upward slant in the centre (the initial level line might have re-emerged over time). The upward-slanting fictive joint fixes the viewing point very close to the image (about 30 cm.). The niche holding the Virgin seems only to have been painted with the upward angled joints (detail, below).

Every aspect of Van Eyck’s work is deliberate: marbled frames are always ‘constructed’ out of fictive segments of marble. It is therefore remarkable that the red frame here is not partitioned into segments – it is the only marbled Van Eyck frame which is in one continuous piece. The artist must have realized that painted joints on the three-dimensional frame would come into conflict with the perspectival joints of the painted white marble niche.

 

Portrait of Jan de Leeuw, 1436

 

Jan van Eyck (fl.1422- d.1441), The goldsmith Jan de Leeuw, 1436, o/panel, portrait 24.6 x 19.2 cm., overall with painted frame 33.5 x 27.5 cm., Kunsthistorisches Museum, Vienna

Van Eyck, Portrait of Jan de Leeuw, drawing of section of frame

The frame on the Portrait of Jan de Leeuw is closest to that on the Self-portrait; the construction is the same and the proportions of the profile are almost identical (above). The top and bottom mouldings have been removed and re-attached, and the outer lip (which is completely missing on the Self-portrait) has been shaved off [5]. The undulations of the complex inner mouldings are softer than those of the Self -portrait. The innermost moulding has traces of gold, which might have been glazed to suggest an alloy rather than the double-gilded solidity of the Self-portrait frame. Traces of paint seem to accentuate the modulation of the profile. The inscription is one of the most complex and identifies the sitter explicitly, without the ambiguity of that on the 1433 portrait [6].

 

Madonna & Child with Canon van der Paele, 1436

 

Jan van Eyck (fl.1422- d.1441), Madonna & Child with Canon Joris van der Paele, 1436, 122 x 157 cm., Groeningemuseum, Bruges

Van Eyck, Madonna & Child with Canon van der Paele, drawing of section of frame

A separate frame is attached on all four sides to this much larger panel; it has the same inner profile as the frames of Jan de Leeuw and the Self-portrait, although the second hollow is slightly wider and deeper (above). The gilding of the inner edge appears to have been added later and has subsequently been covered in a toning layer of greenish-yellow ochre; the outer moulding also seems to have been gilded at a later stage and then over-painted.  It is likely that the whole frame was originally intended to have imitated brass. The inscription is very elaborate [7]. To my knowledge this is the earliest continuous moulding frame which deploys the device of the canted ‘rainsill’ at the bottom edge [8]. It is possible that the frame was made with the same profile on all four sides, and the artist changed his mind and introduced the sill at a later stage.

Van Eyck, Madonna & Canon van der Paele, details revealing later construction of the canted rainsill

To test this idea, I replicated the profile (detail above; top) and converted the inner moulding into a sill after joining a corner (detail; centre). This seems to be the only explanation for the straight mitre-cut connection, which is unsatisfactory and very clumsy for Jan van Eyck (detail; bottom). If the sill were intended to be used from the beginning, it would have continued into the corner, with the profile of the mouldings merging into the chamfer vertically, as in the top detail.

Memorial to the brothers Huart and Jacques Walois, c.1425-35, 174 x 130 cm., detail with rainsill at base; from the cemetery of Saint-Nicaise, Arras; now in the Musée des Beaux-Arts, Arras

The Van der Paele Madonna was intended as a wall memorial, which would have been placed near relief epitaphs in brass and stone in the church of St Donatian in Bruges. The idea of changing the profile into a rainsill might have emerged from the proximity to such three-dimensional relief panels, which tend to have a canted sill-like ledge at the bottom (above). This sill is a very effective optical device for deepening the depicted space of Van Eyck’s painting; the angle of the frame continues seamlessly into the tiled floor, a combination of real and represented spatial recession which became almost obligatory during the following decades.

 

St Barbara, 1437

 

Jan van Eyck (fl.1422- d.1441), St Barbara, 1437, 34 x 18.5 cm, Koninklijk Museum voor Schone Kunsten, Antwerp

The thin panel on which the saint is painted is set into the slot of an engaged frame. The two vertical sections span the entire height of the object, whilst the horizontal elements butt against them, in line with the inner sight edge.

Author’s replica of a corner of the St Barbara frame, showing the tenon, the hole in the vertical convex moulding and the dowel which fits into it

A concealed tenon is fixed with a dowel in each corner (above). This construction is unusual, as it necessitates the lateral mouldings being hand-carved because the corners are cut into the vertical pieces.

Van Eyck, St Barbara, drawing of section of double-sided frame; rainsill indicated with dotted lines

The bottom element is 4 millimetres narrower than the other three, thus visually losing the angled chamfer between the marbled half round and the outer – or top – black section (above). This flat dark section and the red-marbled inner moulding retain the same visual width all round.

Van Eyck, St Barbara, showing green faux marble finish on reverse of frame. Photo: Closer to Van Eyck

The black paint on the top edge and the sides is later and appears to be covering the original greenish marbling which continues in its original state on the verso.  The marbling is made up of segments with painted joints. The actual construction is only just visible in hairline cracks; Van Eyck’s faux painted joints on the verso, which run in the opposite direction from the actual construction, have succeeded in baffling even the most careful modern observer. These trompe l’oeil joints would have been the expected, logical way of constructing the frame, because they would have allowed the sill at the front to continue in one piece into the corner, but both its real construction and that of the narrower width of the bottom side have been missed in Hélène Verougstraete’s detailed survey of early Flemish frames, which shows how persuasive Van Eyck’s illusions have remained over the centuries [9].

This is the first surviving prototype of the now familiar sill frame. Unlike the Van der Paele frame, in which the sill was probably an afterthought with the unintended mitred corner, the horizontal sill here connects to the vertical moulding in a way which is almost immediately copied by other artists. The carved profile might have been executed by Van Eyck’s own hand. The sill manipulates the perspective of this unusual work: St Barbara is placed on a rise, with the building site unfolding below her, and the tilt of the sill directs the view down into the composition.

Lucas Cranach the Elder (1472-1553), Diptych: two Electors of Saxony, 1509, o/panel, National Gallery, NG 6538 & NG 6539

This type of canted sill frame became widespread in Northern Europe, with little change until the first quarter of the 16th century; it can be seen, for instance, in the bottom right-hand corner of Cranach the Elder’s Diptych (above), 1509.

 

The Dresden Triptych, 1437

 

Jan van Eyck (fl.1422- d.1441), The Dresden Triptych, 1437, o/panel, centre panel: 33 x 27.5 cm; wings both: 33 x 13.5 cm., Gemäldegalerie, Dresden

Van Eyck, Dresden Triptych, drawing of section of frame

In the Dresden Triptych, the upright elements and mouldings of the central frame are made from the same piece of wood as the pictorial panel, and the top and bottom elements are attached with three dowels. This is a miniaturised version of the Self-portrait profile, and it has retained the raised outer lip or contour which that frame has lost (above). The central panel is almost exactly the same size as that of the Self-portrait, but reducing the width of the frame has given the architectural setting greater monumentality.

Van Eyck, Dresden Triptych, outside of right wing: detail showing original grey marbling at sight edge

The panels of the wings are fitted into the slots of double-sided engaged frames. All the separate framing elements have been removed and re-attached at some point; there is damage on all corners, with some visible restoration attempts. The surface of the outer edges has been removed. The tortoiseshell decoration now on the reverse of the wings most probably dates to the late 16th – early 17th century; a thin inner moulding with original grey marbling, similar to that on the Madonna at the fountain, is still visible.

On the inside frame of the right wing, a tiny blue spot from the painting of St Catherine’s robe has landed on the finished faux marble frame, indicating the concurrent painting of frame and picture (above).

 

The Madonna at the fountain, 1439

 

Jan van Eyck (fl.1422- d.1441), The Madonna at the fountain, 1439, o/panel, 19 x 12 cm., Koninklijk Museum voor Schone Kunsten, Antwerp

Van Eyck, Madonna at the fountain, drawing of section of frame

Once again, for the Madonna at the fountain, the lateral elements have been made from the same piece of wood as the pictorial panel, whilst the horizontal elements are separate, and have been attached with three dowels. The profile is a deeper version of that on the Self-portrait, and is closest to the frame of the St Barbara: the innermost hollow is smaller and less sharply defined; the second hollow is deeper and wider; the step before the half round profile is flat, and lacks the shoulder of the earlier frames (above). The mouldings are hand-carved with a pronounced degree of unevenness, possibly to assist with the illusion of a frame carved from stone. On the lateral elements, mason’s mitres are painted crosswise to suggest that the horizontal elements span the entire width of the frame.

However, the top and bottom elements have at a later date been removed and re-glued, so that the mitres are no longer close-fitting, and the illusion of the stonemason’s joints has been lost (above). The marbling was finished after most of the painting, since it can be seen to overlap the image at the edges.

 As in the example of the Dresden Triptych, the blue of the saint’s robe has found its way onto the frame in a spot at the right-hand side (above, top) and one deliberate brush stroke amongst the marbling on the left (above, lower image).

 

Portrait of Margareta van Eyck, 1439

 

Jan van Eyck (fl.1422- d.1441), Portrait of Margareta van Eyck, 1439, portrait 32.6 x 25.8 cm., overall with painted frame 41.5 x 35 cm., Groeningemuseum, Bruges

Van Eyck, Margareta van Eyck, drawing of section of double-sided frame

Van Eyck’s portrait of his wife is painted on a thin panel set into the slot of an engaged frame. The profile is a double-sided ogee with a chamfered inner edge and a wider flat top section (above), which is similar to that of the Annunciation Diptych.  The two horizontal rails span the entire width of the object, while the vertical elements butt against them in line with the inner edge. This construction would have made it impossible to use planes to shape the mouldings, but the uneven hand-carved technique might have been chosen to assist with the illusion of a carved marble frame.

Van Eyck, Margareta van Eyck, bottom left-hand corner of frame

The undulations of the profile, especially on the lower rail, are so pronounced that it seems highly unlikely that a professional maker of mouldings could have made this. It is much more probable that it is from the unpractised hand of Jan van Eyck himself, who is inventing the profile in the knowledge that the marbled stone finish will make sense of the shape.

Replica frame on Van Eyck, Portrait of Cardinal Albergati, bottom left-hand corner

When we copied this frame for the Portrait of Cardinal Albergati in time for the exhibition at the National Gallery, the frame was carved by hand in exactly the same way out of solid oak, but the result is clearly less uneven than Van Eyck’s version, in the same ways in which a practiced woodcarver of the 15th century would have made Margareta‘s frame.

On the latter, mason’s mitres are painted horizontally and on top of the actual mason’s joints; unlike with the St Barbara frame, here actual construction and illusion support the same narrative (above).The decoration of the frame was finished before most of the painting, since the edge of the painting overlaps the marbling of the frame.

 

Conclusion

In summary, it is evident that Jan van Eyck worked on his frames as a concurrent part of the painting process. It is mistaken to consider these frames as in a separate category from the pictures. One consistent feature of the five semi- integral frames (the Self-portrait, the Annunciation Diptych, the Portrait of Jan de Leeuw, the central panel of the Dresden Diptych, and the Madonna at the Fountain) is that the separate horizontal sections are attached with three dowels. This applies to the narrower frames, like those of the tiny Madonna at the Fountain and the Annunciation Diptych, in the same way as the longer and relatively thinner profile of the central panel of the Dresden Triptych. The preparation of the semi-integral frames is the one consistent element which points to an outsourced commission.

The making of panels and joining of frames were strictly regulated in 15th century Bruges [10], but Van Eyck – as a court official and a valet de chambre for Philip the Good of Burgundy – was legally exempt from the regulations of the local painter’s guild, and might have not been compelled to comply with the exact letter of the rules when it came to small and private works. Maybe the commissioning of the panels with attached semi-integral blanks would have sufficed to appease the local regulations. This would explain why each frame was constructed in the same way, but the profiles were executed differently, the majority with clear evidence for the use of gouges rather than planes. It seems unlikely that a competent carpenter commissioned by a person of the status of Van Eyck would have produced a moulding as wildly undulating as that of the Self-portrait – or of Margareta van Eyck. It seems more possible that Van Eyck invented and carved both profiles himself, or had the carving executed by a collaborator in his own workshop, the pronounced unevenness aiding the illusion of metal work – supporting the appearance of a solid gold surround – in one case[11], and of marble in the other.

Self-portrait panel, bottom, with author’s thumb

Van Eyck was deeply interested in materials and material illusions. Even the quality of the oak on which the Self-portrait is painted is worth noticing. It is by some margin the finest, most densely grown board of oak I have ever observed (above). Slowly-grown oak has always been sought after for its superior qualities (softness, evenness, and stability). This particular and exceptional piece would have been put aside and highly prized in an age when hand tools were in common use[12]. We may even wonder whether the purchase of this prime specimen of oak for the creation of a Self-portrait by Jan of Oak (Van Eyck) is yet another of the artist’s enigmatic jests.

*********************

Peter Schade is Head of Framing at the National Gallery, London.

With thanks to Emma Capron for support and encouragement, Bart Cornelis for transforming images and diagrams and Lynn Roberts and Karl Schade for the critical reading of the text. Many of these observations have been facilitated by the excellent ‘Closer to Van Eyck’ website.

*********************

[1] The only other original van Eyck framing is that of the Ghent Altarpiece, where the frame has been damaged and extensively restored and the original framing concept was much more generic; it does not fall into the category of these smaller works executed solely by the hand of Jan van Eyck

[2] See M. Smeyers, ‘Jan van Eyck, archaeologist? Reflections on Eyckian epigraphy’, in M. Lodewijckx, ed., Archaeological and Historical Aspects of West-European Societies: Album Amicorum André van Doorselaer, Leuven 1996, pp. 403-10; R.W. Scheller, ‘ALS ICH CAN’, Oud Holland, vol. 83, no 2, 1968, pp. 135-39; D. de Vos, ‘ALS ICH CAN’, Oud Holland, vol. 97, no 1, 1983, pp. 1-4; S.F. Jones, ‘Jan van Eyck’s Greek, Hebrew and trilingual inscriptions’, in C. Currie et.al., eds., Van Eyck Studies: Papers presented at the eighteenth Symposium for the study of underdrawing and technology in painting, Brussels 19-21 September 2012, Leuven 2017, pp. 291-308; S.F. Jones, ‘Reconsidering Jan van Eyck: texts, objects and a regional revival’, in S. Kemperdick, J. Rössler & J. Corin Heyder, Der Genter Altar/The Ghent Altarpiece: Reproduktion, Deutungen, Forschungskontroversen/Reproductions, Interpretations, Scholarly Debates, Berlin 2017, pp. 129-48

[3] Illustrated with Lorne Campbell’s kind permission. In The fifteenth century Netherlandish paintings (National Gallery Catalogues), 1998, p. 214, Campbell notes that, in the case of the Portrait of Jan de Leeuw, the Dresden Triptych, the Madonna at the fountain, and the Annunciation Diptych, ‘… all four frames are constructed in the same way as the frame of [the Self-portrait] with the lateral sections carved from the same piece of wood as the panel, and the horizontal parts pegged into the panel. The mouldings of all four are similar to the mouldings of [the Self-portrait]’

[4] An observation shared by Elke Oberthaler

[5] See the catalogue of the exhibition, Jan van Eyck – >Als Ich Can<, eds S. Haag, S. Pénot, K. Schmitz von Lederbur, KHM Wien, 2019: Sabine Pénot, pp. 32 -33 and Elke Oberthaler, pp. 65-68

[6] Hugh Hudson, The chronograms in the inscription of Jan van Eyck’s Portrait of Jan de Leeuw, Oud Holland, 2003, vol. 116, pp. 96-99

[7] See Douglas Brine, ‘Jan van Eyck, Cannon Joris van der Paele, and the art of commemoration’, The Art Bulletin, no 96, 2014, pp. 265-87

[8] Canted sills can be found on earlier Gothic altarpieces (for example, the Carrand Diptych, c.1390, Museo del Bargello, Florence, and the Tower Altarpiece, c.1395-1400, Mayer van den Bergh Museum, Antwerp), but only in association with the ‘church façade’ form of frame, with arched and peaked panels, columns, spires and finials

[9] Hélène Verougstraete and Roger Schoute, Cadres et supports dans la peinture flamande aux 15e et 16s siècles, 1989

[10] Guild regulations in Bruges and elsewhere banned painters from carving wood, and their frames had to be made by specialist woodworkers. For the making of frames and the division of labour between joiners and carpenters, see Verougstrate, ibid., pp. 1-6

[11] I am grateful to Susie Nash for sharing her then unpublished manuscript for the 2026 National Gallery exhibition catalogue

[12] Slowly-grown oak, in German Milde Eiche, has yearly growth of close to 1 millimetre; the most densely grown coming close to 1 centimetre in 10 years. The board of the Self -portrait exceeds this margin for the entire width