prayers and portraits

Unfolding the Netherlandish Diptych

National Gallery of Art, Washington | November 12, 2006 – February 4, 2007

the exhibition is made possible by the homeland foundation, inc.

technical appendix

276 277 the following is a record of we examined and documented two full Abbreviations individuals Cat. 3. Albrecht Bouts gold leaf here has a butter-yellow auto- dots over the gold, trailing off to the upper cal quality of a tracing. The hatching and observations made by a team of conser- diptychs (cats. 26 and 31), both uniquely ef Eugene Farrell (huam) fluorescence like that immediately under left from the body of the dots. Although crosshatching from the second stage were vation, curatorial, scientific, and other well preserved, in a single day. Peter Klein, institutions the gold on the frame (see below), possibly the dots appear to have been applied in done in a liquid medium and are very mdg Maria Dolores Gayo (mnp) ecce homo and mater dolorosa professionals who studied sixty-five of Hamburg University, performed new huam Harvard University Art a mordant or a bole layer. lines from bottom right to top left, no regular, marking zones of shadow. They mg Melanie Gifford (nga) wings: the angel gabriel and Netherlandish panel paintings that either dendrochronological analyses for us and Museums, Cambridge, ma The frame has a different buildup effort was made to standardize the pattern, do not always follow contours precisely nk Narayan Khandekar (huam) the virgin annunciate currently function or are thought to have provided findings from his earlier exami- kmska Koninklijk Museum voor of preparation layers. Over a ground layer so an overall speckling is produced rather (i.e., at the nostrils). We did not observe the pk: Peter Klein (Hamburg University) Suermondt-Ludwig-Museum Aachen, originally served as part of a diptych. This nations. Paint samples for the project were Schone Kunsten, Antwerp similar to that on the panel, two colored than regimented rows. powdery underdrawing seen in the Ecce hl Henry Lie (huam) inv. no. gk 57 research project was initiated, organized, taken by Melanie Gifford and Catherine mnp Museo Nacional del Prado, layers are present: the first is gray; the In the first paint stage the right hand Homo, and we found this underdrawing cm Catherine Metzger (nga) Examined at sral: 15 – 27 September 2003 coordinated, and administered by Ron Metzger of the National Gallery of Art Madrid second is a thin yellow-tan with a butter- was more horizontal, with the fingers generally more rigid and schematic than jm Joe Mikuliak (pma) by cm, mr, rs, cvd, av, uv Spronk, associate curator for research, and Narayan Khandekar of the Harvard mtb Museo Thyssen-Bornemisza, mp Michael Palmer (nga) yellow autofluorescence also seen in upper starting farther to the right and the little that for its mate. Straus Center for Conservation, Harvard University Art Museums, then analyzed Madrid ap Amy Powell (huam) Documentation and analyses layer on the panel under the gold leaf. The finger slanted downward. In the final Several minor changes were made University Art Museums. The other core by Gifford at the National Gallery and nga National Gallery of Art, mr Michael Rief (Suermondt-Ludwig- 2003 at sral: Phase One visible and ir top two layers in cross section are 1) dirt painted position the knuckles are higher. between underdrawing and the paint members of the research team were Khandekar at Harvard. Washington, dc Museum Aachen) as well as Inframetrics irr by cvd, and 2) traces of repaint with an orange Both the x-radiograph and irr show stage: the mouth was underdrawn slightly Catherine A. Metzger, senior paintings Research materials were processed and pma Philadelphia Museum of Art cs Christoph Schmidt (smbg) av; binocular microscopy by cm, rs; autofluorescence like that on the top layer increased density of paint in this hand. open with a dip in the center, but it was conservator, National Gallery of Art; archived at Harvard’s Straus Center, largely rte Rijksmuseum Twenthe, gs Gerhard Schultz (smbg) x-radiography by at; dendrochronology of the composite frame. The rope was painted over the robe rather painted straight across; and the fold in the Catharina Van Daalen, Theodore Rousseau by Catharina Van Daalen.2 X-radiographs Enschede rs Ron Spronk (huam) by pk; cross sections by cm than left in reserve, as were the hair and veil at the top center is painted slightly preparatory design: Two stages of under­ Intern, Harvard University Art Museums were digitized at high resolution and smb Stedelijke Musea Brugge at Arnold Truyen (sral) 2005 at nga: cross section analysis by mg hands, implying that it was a late addition. narrower than it was underdrawn. drawing are readily visible in irr; a prior (2003 – 2005); and Adriaan Verburg, Founda- assembled into composites where neces- smbg Staatlichen Museen zu Berlin, mt Mark Tucker (pma) stage, possibly brushed away, is implied paint layers: The paint was so thinly tion Arcobaleno, Antwerp. We typically had sary.3 Assemblies were produced from Gemäldegalerie cvd Catharina Van Daalen (huam) ecce homo (db no. 23.45) by residues of a powdery material in mater dolorosa (db no. 23.46) brushed that the underdrawing is vis- infrared documentation and x-radiographs digital infrared reflectograms and macro­ sral Stichting Restauratie Atelier av Adriaan Verburg (Foundation striations of the intermediate layer. The ible, seeming to lie atop the intermediate at hand when studying the paint surface photographs in the visible and infrared Limburg, Maastricht Arcobaleno, Antwerp) frame: Integral. Regilded. Exterior pro- frame: Integral. Regilded. The flat outer medium of the revealed underdrawing layer and directly beneath the paint. The with the microscope. Large numbers of ranges of the spectrum. Files were super- uv Ulrike Villwock (Suermondt- files were probably cropped, as the inner plane is the same width throughout, appears to be liquid, with both thin and paint follows the underdrawing fairly photographs, macrographs, and micro- imposed, allowing for close comparison Ludwig-Museum Aachen) profile is wider in places than the exterior, unlike that on the Ecce Homo. X-radiogra- bolder washlike lines. Thin pale gray lines closely, and hatching shows through graphs were digitally captured for later between technical documents. Our and the width of the flat outer profile is phy revealed a metal element at the top, visible in irr were probably traced over to create midtones. An unusual system reference. We traveled with digital cameras, observations and documentation were slightly wider at the top of the arch. The probably a remnant of a hanging device. frame on the Mater Dolorosa is generally now-removed material of an initial under- of shading, using clusters of tiny thin infrared equipment, and computers owned entered in a relational database; and docu- support: Oak, 45.3  31.1. Single plank wider, and its flat outer profile has an even drawing to define contours. The final strokes of brownish red (i.e., at the bot- by Harvard University’s Straus Center ments (about 75,000 files totaling around with vertical grain and shallow bevels on width throughout, but the overall widths underdrawing was apparently executed in tom of the chin and nose), adds color to but used microscopes provided by local 200 gigabytes) were renamed to facilitate the reverse, wider at the top. The reverse 1 4 and heights of the two framed panels a heavier, more washlike material to pre- shadows. Forms are painted with slight institutions. We are deeply grateful to archiving and later accessibility. It is our is covered with reddish brown paint that are similar. This suggests that the frame pare areas of shading. gaps between the color areas so that the colleagues who allowed us unparalleled hope that all these materials will at some is thinner than the paint on the reverse of the Ecce Homo was trimmed to match Compared to the underdrawing in effect is flat and lacks the vitality of the access to research facilities and works point become accessible on the Internet. of the Ecce Homo, and it has no dark under- that for its companion panel. Several holes the Mater Dolorosa, this was more organic, more richly developed technique in the of art in their care, sometimes for weeks at layer. This paint lies over paper scraps in the top may be traces of former hanging fluid, assured, and free. Paler, thinner Ecce Homo. a time (in Antwerp, Berlin, and Maastricht). securely dated to 1804 (see entry), indicat- devices. A large insert at the back of the contours were applied with confidence, In other locales we worked under greater ing that the coating and inscription on the top edge reinforces this area of damage. often using several lines for a single con- wings: the angel gabriel and time pressure; in Bruges, for example, reverse were added after 1804. Grooves in the sides allow the panel to be tour. Darker lines added shading with the virgin annunciate fit into the secondary frame. crisply parallel hatching. In some places dendrochronology: Earliest possible (db nos. 23.118 – 119) (i.e., the back of the left hand where a felling date: 1509; earliest possible creation support: Oak, 45.5  31. Two planks, verti- frame: None. Illusionistic painted frames. tangle of thinner, lighter lines goes in date: 1511; statistically more plausible pro- cal grain. Left plank is 27.1 / 27.3 cm wide; One edge of each panel is hinged directly different directions) the drawing is dif- duction date: after 1517 (assuming 2 years right plank is 4.3 cm wide at the bottom. to the composite frame enclosing the Ecce ficult to interpret. The contours of the left for seasoning and transportation). Reverse is painted reddish brown over a hand were redrawn in the darker line used Homo and Mater Dolorosa. The wings have dark underlayer and over paper scraps elsewhere for hatching. preparation of support: White, coated frames later added by the museum at the securely dated to 1804 (see entry), indicat- The crown of thorns is completely with an intermediate layer that is difficult top, bottom, and nonhinged edges. ing that the coating and inscription on the to see in the x-radiograph but is also pres- drawn but not carefully followed in the support: Oak, 52  38.2; and 51.9  38. reverse were added after 1804. ent under the gold leaf. A slight indication paint stage (thorns were painted where The Angel Gabriel comprises two planks, of striated lines across the nose and at the dendrochronology: Earliest possible none were drawn, and vice versa). The 1.4 cm thick, with vertical grain. The left chin is visible in ir as well as in surface tex- felling date: 1483; earliest possible creation fingers on the left hand were painted plank is 26.7 cm wide; and the right is ture at the chin. A cross section shows that date: 1485; statistically more plausible pro- longer than they were underdrawn so that 11.4 cm wide at the bottom, 11.6 at the top. the layer immediately beneath the gold is duction date: after 1499 (assuming 10 years they appear to rest on the frame. The right The Virgin Annunciate comprises two planks, a dense orange gilding preparation, which for seasoning and transportation). hand was changed in both underdrawing also 1.4 cm thick. The left plank is 10.7 cm This technical appendix is excerpted from Prayers and Portraits: differs from that in the Ecce Homo. and paint stage: it appears to have been wide at the top, 11.3 cm at the bottom; the preparation of support: Chalk, with an The preparation layers on the frame drawn with the thumb higher, but the right is 27.2 cm wide at the top, 26.3 cm at Unfolding the Netherlandish Diptych, Copyright © 2006 Board of Trustees, intermediate layer that has little pigment are not the same as those on the panel. hand and fingers placed farther to the the bottom. The wings were originally a in cross section. It does not register in A black layer lies over a thin light-colored National Gallery of Art, Washington, available November 2006. right and more horizontal. single painting that was cut in two. the x-radiograph (striations visible in the layer that may be the ground. Over these x-radiograph register the paint layer on paint layers: The paint was applied with is a thick yellow-tan gilding preparation dendrochronology: Earliest possible the reverse). Traces of the intermediate confidence and economy, working up to (with whitish autofluorescence), then felling date: 1480; earliest possible creation layer are visible in irr, notably in the face, the lightest and generally thickest areas. gold leaf. The layers above this appear date: 1482; statistically more probable near the eyes and nose, as thin dark diago- Red lake glaze was used in the shadows, to be restoration and consist of (from creation date: after 1496 (assuming 10 years nals. These may be residues from brushing especially in the crown of thorns. The bottom to top): a dark layer, a gray layer, for seasoning and transportation). Narrow away a powdery dry underdrawing mate- painting technique is comparable but less a fragment of gold leaf, and a dark trans- boards of each originally formed a single rial that was trapped in a ridged surface. elaborated than in the Cambridge pair parent heterogeneous layer with orange plank. The x-radiograph is generally less (cat. 4), resulting in an effect of less volume autofluorescence similar to topmost layer preparation of support: White. No evi- opaque than that of the Mater Dolorosa, and lower quality. on frame of the Ecce Homo and on the com- dence of an intermediate layer in cross though the paint layers appear thicker in The gilding of the background is posite frame. section, ir, irr, or visible light, though the Ecce Homo. This contradiction may be original (unlike that of the frame). A trans- preparatory design: Two stages. The some brush strokes not related to the explained by a difference in the prepara- lucent red glaze was applied in rounded initial underdrawing was executed in a image are visible in x-radiography (prob- tion layers on the two paintings. In cross combination of thicker and thinner lines, section the layer immediately under the with the thin lines having the mechani-

278 279 ably related to the painted underlayer of Not original to the paintings. Cat. 4. Albrecht Bouts volume was indicated are the bags under paint layers: The paint was applied with Cat. 5. Robert Campin paint layers: Haloes were incised in the the inscriptions on the reverses). the eyes, which were underdrawn with a mater dolorosa (db no. 25.50) finesse and skill. The red dots on the gold, and the x-radiograph shows the dendrochronology: Earliest possible broader line of gray material resembling a gilded background were completed first. interior contours as white lines and the preparatory design: Multiple campaigns. felling date: 1592; earliest possible creation man of sorrows and frame: Not original christ blessing with the virgin wash. Slight changes in the hands include These dots are very different from those outer contours as dark; the latter must Perspective lines for the architecture date: 1594; statistically more plausible mater dolorosa in pr ayer (db no. 1.1) especially the thumb and the tips of the support: Oak, 37.5  26.7. Single plank, on the Man of Sorrows, with each carefully have been done after paint was applied. were defined before the rest of the com- production date: after 1604. Harvard University Art Museums, Philadelphia Museum of Art, inv. no. 332 fingers of the left hand; the fingers of the trimmed at top and possibly the right. executed so that all are round (or rounded) Paint application was direct and sparing, positions. Parallel lines of hatching and Fogg Art Museum, inv. nos. 2001.170, 171 Examined at pma: 20 – 24 October 2003 by Cross sections taken from two locations right hand were lengthened in paint and Reverse is thinned to 5 – 6 mm, flattened, and regularly spaced in neat vertical lines. and the technique was straightforward. crosshatching in a liquid material indicate Examined at huam: 3 – 7 March 2003 by cm, rs, cvd show a layered structure corresponding to shifted slightly to the right. and cradled. The cradle, with broad soft- The flesh was painted next. A cross section Colors were mixed on the palette and zones of shadow. The style and method cm, ap, rs; and 9 – 13 May 2005 by cm, rs diverse decorative schemes. At the upper wood members, is analogous to that on the shows that the flesh tones were applied on Documentation and analyses applied adjacent to one another, blend- of underdrawing are not comparable with paint layers: In the initial paint stage surface, over a layer of dirt, a dark trans- Documentation and analyses Man of Sorrows. a translucent brown layer lying directly unknown date at pma: x-radiography ing the edges. The final details, such as that of the Ecce Homo or Mater Dolorosa. The red “dots” were applied over gold leaf in lucent heterogeneous layer with orange 1998 at huam: visible light by rs; Phase on the ground, with an overall medium by jm eyelashes, tiny touches of red paint on the top and bottom frame elements were the background in a distinct left-to-right dendrochronology: Earliest possible autofluorescence is similar to the topmost One ir by hl, rs; Inframetrics irr by ap; tone over which fine strokes of lighter and 1993 at pma: dendrochronology by pk upper eyelid, and highlights on eyes and drawn farther into the image than painted. diagonal orientation. The dots vary in felling date 1478; earliest possible creation layer on the frames of the Ecce Homo and binocular microscopy by rs; x-radiography darker paint built up the form of the face. 2003 at pma: Phase One visible and ir by fingernails, were meticulously added with A single line marks center of each bench leg. size, density, and degree of completion; date: 1480; statistically more plausible Mater Dolorosa. by ef, rs; analytical microscopy by ef One unusual aspect of the technique is cvd; binocular microscopy by cm, rs a fine brush. The paint is thin enough that they are linear, teardrop-shaped, or production date: after 1494. The plank paint layers: The paint is opaque, with 1999 at huam: dendrochronology by pk the use of a thin line of red paint atop the underdrawing can be seen through it, yet amorphous; they are not organized on a came from the same tree as that for the frame: Not original. textured brush marks. The application 2003 at huam: cross sections by cm flesh tones to outline the shadowed areas it gives an impression of solid volumes. regular grid. This differs markedly from Man of Sorrows. was direct and simple, with little blending Summary 2005 at huam: Phase One visible and ir (seen in a cross section from the proper support: Oak, 28.4 / 28.6  45.4 / 45.3. One The jewels adorning the haloes received the companion panel. Next, the flesh or glazing. The technique in the landscape Altarpiece was assembled long after the by cvd; binocular microscopy by cm, rs preparation of support: The whitish left side of the Virgin’s face). The hair and plank, with horizontal grain. Small piece the most elaborate treatment, worked tones were laid in, using an overall warm used the striated red brown intermedi- death of Albrecht Bouts, using paintings 2005 at nga: cross section analysis by mg calcium carbonate ground was applied in garment were painted after the flesh tones of conifer added at the lower edge is not wet-in-wet with minute flecks of color. beige (visible near Christ’s hairline), over ate layer, more typical of the sixteenth or that may have originated in the Bouts two layers and continues to the edge of the and lie over a translucent isolating layer original. The panel is trimmed at the top Another example of incised line is seen in which fine strokes of lighter and darker seventeenth century than of the fifteenth. workshop. Dendrochronology suggests panel. The particles are coarser in the ini- that fluoresces brightly. A studio assistant and slightly at the bottom and is thinned the hair above Christ’s right hand, where man of sorrows (db no. 25.49) paint were blended. The painting of the Trees in the landscape were summarily that the central panels were painted some tial layer than in the final one. A red brown may have been assigned to paint hair and and cradled. the brown paint was scratched through skin tones was meticulous, with red paint brushed with swirled strokes of color to twenty-five years apart, while parts of the frame: Not original. isolation layer was then applied, contain- drapery after the master had painted the while wet to create separate strands of hair. defining shaded areas. Surface details like dendrochronology: Earliest possible form a general mass. The bench at the back secondary frame can be dated to the seven- ing lead white, chalk, and red ocher. This face, veil, and hands. Finally, details such support: Oak, 37.5  26.7. Single plank, bruises, blood, and thorns were applied felling date: 1408; earliest possible creation of The Virgin Annunciate had a lobed design teenth century. The wings were originally panel does not have a ridge in the ground as tears, individual strands of hair, trim c. 5 – 6 mm thick, with vertical grain. with similar care. Cross sections show an date: 1410; statistically more plausible leading to the central leg, taken into the a single panel that was cut in half. These at the bottom edge. on the garment, and cool scumbles on the Summary Trimmed at top and possibly on left edge. unpigmented layer between the basic flesh production date: after 1424. first paint stage but changed to simple disparate parts were combined within the flesh tone were added. The band of red This painting is not a diptych but was Reverse was thinned, flattened, and tone and subsequent layers. For the eyes, preparatory design: This panel has three arches. Studied under the microscope, the painted and gilded secondary frame, per- and yellow paint at the edges, which gives preparation of support: White ground part of our study because it influenced cradled. The cradle, with broad softwood the lids were first modeled with warm skin stages of underdrawing. The first shows far right edge of the window opening at haps as late as 1804, the terminus post quem the impression of a shadow, was a last with no isolation layer visible in either the development of the Netherlandish members, is analogous to that for the Mater colors, then the whites with a bit of blue as frontally oriented eyes, a nose shaded on the right side of The Angel Gabriel shows a for application of paint on the reverses touch, here as on the companion panel. x-radiograph or visible light. diptych tradition. Campin’s technique Dolorosa (see below). well as strokes of bloodshot red, and then the right side, two reference lines that fine sliver of light-colored paint, probably of the panels. The Virgin’s head covering was suggests the possibility of collaboration the irises and pupils. The glistening high- cross at a dot on the Virgin’s chin, and preparatory design: The drawing from the landscape at the left edge of The Evidence that the Mater Dolorosa and dendrochronology: Earliest possible changed after the underdrawing, as between the master and workshop assis- lights and the stroke of shadow marking broad straight lines at the bottom and appears to include at least three stages: Virgin Annunciate, which was originally Ecce Homo have different restoration his- felling date: 1476; earliest possible creation described above. In addition, it was first tants. The first two stages of underdraw- the bottom of the upper lid completed the both sides that appear as a gray mate- first the contours, then the shading in part of the same panel. The edges of the tories further implies such a late pairing. date: 1478; statistically more plausible painted in white and black with a diagonal ing are confident, while the third is almost eyes, and only after this were eyelashes and rial. The initial underdrawing is highly robust descriptive liquid lines, and lastly illusionistic frames are incised, with lines The sample from the background of the production date: after 1492. Analysis section over her forehead, then repainted overly explicit. The paint closely follows tears added (with fully described volume, schematic (for instance, the shading of the a fine, dense network of diagonal hatch- of incised paint in the framing elements Mater Dolorosa shows a structure with two produced these dates for this panel, but in white and blue extending farther over the third stage of drawing wherever it is spilling over the bottom lid). The base nose is achieved with evenly spaced hatch- ing. The initial shading included zigzags continuing across the cut that divided layers of different varnishes. The sample because it came from the same tree as that her forehead in its present configuration. present, while changes are common in the color of the hair and the garment were ing). The second stage of underdrawing and described the general contours and the panels. from the background of the Ecce Homo does for the Mater Dolorosa, the slightly later areas described only by the first two stages not include these layers. Likewise, the fin- dates for that plank must be taken into painted next. Finally, a border at the edge has been largely removed but remains volumes of the forms. The final stage of underdrawing. It is possible that the reverses (db nos. 23.122 – 123): No isola- ish of the engaged frame of the Mater Dolo­ account here. of the panel was executed in red lead and visible as scattered dark specks in the irr Summary was more uniform and indicated the fall third underdrawing was intended to pro- tion layer is evident. The preparatory rosa was revised repeatedly: over the black lead-tin yellow over vermilion, identical to and occasionally where the final drawing These high-quality panels were painted of light through increased line density, vide unambiguous direction to a studio preparation of support: The whitish design consists of a simple contour line, paint layer lies a yellow-tan mordant and the border on the Mater Dolorosa. does not cover it, such as at the top of the around the same time by the same hand, creating a tonal field, not a wash. The assistant in the application of the paint on calcium carbonate ground was applied in possibly executed in a liquid medium, gilding layer, followed by a thin dark layer, The artist increased the size of the left eyelid and on the forehead under but the markedly different method and modeling of the pouch beneath the Vir- the faces. multiple layers and continues to the edge that describes the architecture surround, a thin gray layer, and a fragment of addi­ fingers on Christ’s proper left hand and the headcloth. We infer that the second style of the underdrawing point to the gin’s eyes ignored an earlier underdrawn of the panel. The translucent uppermost including scrollwork on the flat faces, tional gold leaf. The frame of the Ecce shifted them slightly to the right, except stage was done in a dry, powdery material, possibility that they were created in an placement of the eyes. layer seems to be a single glue-rich layer profile heads in the roundels, and swags Homo shows little evidence of changing for the index finger, which was extended which was traced in a dark liquid at the active workshop with various hands The underdrawing is more visible with a well-defined border rather than with hanging rings. The drawing was decorative schemes. Only in the top layer to the left. The entire thumb was also third stage of drawing, then brushed away, participating in production. The initial in the hands than in the faces, perhaps irregular staining caused by medium that followed only approximately in paint, with on the engaged frames do the samples enlarged, from its tip to the base of the leaving only faint traces in the porous underdrawn composition of a frontally because less lead white was used in the has soaked down from the paint layers. many deviations especially in the scroll- show a similar, dark translucent coating, palm. The middle finger on Christ’s proper surface of the ground. Too little of the oriented face on the Virgin panel was hands, or because only the first two A slight ridge in the preparation occurs work. The paint was thinly and efficiently but this was applied over the several layers right hand was made larger, and that second underdrawing is visible to describe abandoned before painting began, and the drawing stages, which appear darker in along the bottom edge, but a true barbe applied. Major color areas (brown, gray, on the frame of the Mater Dolorosa and over index finger smaller, while the little finger its style accurately. The final underdraw- final underdrawing appears to have been irr, were used there. The third draw- is not present. There is a thin whitish and black) were quickly brushed in a thin, dirt and damage on the frame of the Ecce was shifted to the right in a second paint ing, made with a heavily loaded brush, is traced. It is possible that the Man of Sorrows ing campaign codified the shading and isolating layer of irregular thickness. The striated layer. Slightly more opaque paint Homo. The same coating is present as the stage, with final touches of light flesh clumsy and insecure, with contours bro- was adapted from an autonomous image. modulation of flesh tones in the faces. ground has an unusual pattern of cupped, was then summarily applied to indicate final layer in a sample taken from the tones added over the paint of the garment. ken into multiple short strokes and hatch- This might explain the discrepancies in This drawing was followed carefully in diagonal craquelure, which runs from the smaller features such as the highlights, secondary frame, suggesting that the The position and size of pupils changed, ing that is irregular in scale and direction. organization of the background dots. the paint, while there are deviations from top right to the bottom left. berries, leaves, etc. Mordant under the coating was applied only after the panels possibly in relation to the pairing with the Several small changes occurred after Pairings such as these might never have the first and second drawing campaigns. lettering is x-ray opaque. were assembled as an altarpiece. This preparatory design: The scarce under- Mater Dolorosa. the final underdrawing. The Virgin’s veil functioned as a folding diptych, as none Christ’s eyes were drawn lower than they sparsely pigmented layer has an orange drawing that appears in irr defines the had a more sharply pointed fold at the has a painted verso. Instead, they were are painted, with dark irises. The Virgin’s Secondary Frame (db no. 23.138) autofluorescence, which is typical of a eyes, the hollow on the upper lip, the forehead in the second stage drawing probably displayed as pendant paintings. eyes were initially drawn lower, but com- toned shellac coating and was probably contours and hollows in the neck, and the than in the painting. Her thumbs were pletely redrawn in the third stage. The ring Documentation and analyses applied to unify the appearance of the contours of the fingers. Some features underdrawn much longer than painted. on the Virgin’s finger was painted lower 2003 at sral: Phase One visible by cvd, av; present triptych. are underdrawn with thin, grainy lines. A long diagonal line went through the than underdrawn, and the fingers of both binocular microscopy by cm, rs; x-radi- Contours do not describe volume, and thumbs, parallel to the index finger, but figures were slightly altered. A curl at the ography by at; dendrochronology by pk; there is no hatching to establish light it is not clear what it delimited. The blue top center of Christ’s forehead was under- cross sections by cm and shade. The only areas where some headcloth was painted larger at the crown drawn but not painted. Changes were also 2005 at nga: cross section analysis by mg of the head and at the right than indicated made in the Virgin’s hair and in the attire in the drawing. of both figures.

280 281 Cat. 8. Jan van Eyck The drawing of the pedestal base in Cat. 14. Quentin Massys and/or painted after the underlayer of hair, then paint layers: Oil paint. The paint was Cat. 15. Quentin Massys paint layers: The paint stops at the The Virgin Annunciate changed in the same final strands of hair were added as a fin- applied efficiently in broad masses, begin- barbe, showing that the panel was framed Summary Jan Massys way as in the Gabriel, though there is no ishing stage. This working back and forth ning with the background and leaving prior to paint application. The paint was the angel gabriel and This diptych shows evidence of thought- virgin at pr ayer and double line at the panel’s upper corners is seen elsewhere, such as at the thumb a reserve for the figure. Blending and smoothly applied and blended, with the virgin annunciate ful planning for its use. Changes made in christ as savior where the lintel meets the supporting virgin at pr ayer and and finger, where a reserve was left in the surface details were handled with great texture evident only in the crown, the (db nos. 3.4 – 5) the reflection behind Gabriel and in the Koninklijk Museum voor Schone Kunsten, columns. The Virgin’s cape originally had christ as savior, 1529 background, then the flesh was painted, care. In the painting of the inscription, mordant-gilded halo, and the gold Museo Thyssen-Bornemisza, Madrid, angle of the lintel join above him prob- Antwerp, inv. nos. 241 – 242 a round clasp at the tie end on the left side. Museo Nacional del Prado, Madrid, and finally the background was brought for example, a first underlayer of yellow embroidery on the Virgin’s cloak. The inv. nos. 1993.11.1 – 2 ably indicate that this panel was painted Examined at kmska: 11 – 22 November Binding cords on her book were under- inv. nos. 1562 and 1561 back over the flesh. Final touches in the ocher was followed by maroon, then (tiny) major forms were quickly laid in with Examined at mtb: 13 – 21 September 2004 first. These changes reflect Van Eyck’s 2003 by cm, rs, cvd, av drawn but not painted. Examined at mnp: 13 – 21 September 2004 hair (in black and in white) were applied pink strokes perpendicular to the maroon, broad brushwork, as seen in the ir of the by cm, rs, cvd, av active concern with the construction of a by cm, rs, cvd, av in a very liquid medium, breaking up then lead-tin yellow. A dash of blue/green, Documentation and analyses back of the Virgin’s head. Subsequent paint layers: The paint was expertly viewing angle for the diptych (see entry). Documentation and analyses into tiny beads. This technique was also perpendicular to the underlying mauve, 2003 at kmska: Phase One visible and ir paint layers were added with care and applied in a painterly manner, incorporat- The jaspering of the sides and reverses Documentation and analyses 2004 at mtb: Phase One visible and ir as observed in cat. 15. was painted in the yellow at the top and as well as Inframetrics irr by cvd, av; attention to detail. Short strokes of black ing or departing from the underdrawing indicates that these elements were meant 2004 at mnp: Phase One visible and ir as well as Inframetrics irr by cvd, av; binoc- The Virgin’s little fingers had contour bottom of the inscription. binocular microscopy by cm, rs paint over a midtone foundation were used as desired. The handling is masterful, to be seen, and together with the lack of a well as Inframetrics irr by cvd, av; ular microscopy by cm, rs; x-radiography changes visible in ir but painted out, with Finishing touches in a very liquid 2004 at kmska: dendrochronology by pk to model the flesh, with brighter flesh with confident strokes of thick white used frame on the reverse, this creates a strong binocular microscopy by cm, rs; by RayXArt Madrid a reserve left in the underpaint and the black and white, which break up into colors brushed over it so that the black to pull out highlights and adroit strokes association with books of the time. The dendrochronology by pk finger in front painted over the one in tiny beads, were applied with a fine brush established shading. Fine points include frames: Semi-integral; top and bottom of black to push the deepest shadows change in the bases of the pedestals so that 2005 at mnp: cross sections and cross virgin at prayer (db no. 35.69) back. More of the pupil originally showed to adjust contours and add individual the way the white of the eye was brought pieces are secured to the fronts of the back over a structure partly modeled by they appear to overhang the frame and the section analyses by mdg in the right eye, and the eyelid was painted strands of hair. As in the companion panel, frame: Not original. Frans Francken ii’s over the lower part of the iris and the panels with three pins each. The pins on the densely hatched underdrawing seen painting of the lintels as if viewed from over the top of the iris. The white veil the edges of the fingers are redefined with Banquet in the House of Burgomaster Rockox eyelashes were painted with intermingled the bottoms are evenly spaced but not through the thin paint. beneath intimates that the viewer should virgin at prayer (db no. 48.95) behind the Virgin’s head was made more background paint. shows the painting as a diptych in a red- black and flesh tones, layered over each exactly centered, while the center pin on Subtle but significant changes in see these elements as though enclosed voluminous than the reserve left for it, and dish frame. Traces of red paint on the other. Black strokes of the bottom eye- the tops is offset to the right. The frames the architectural elements at the top and by them. The diptych seems to have been frame: Not original. the shadow in the blue robe behind her barbes under the surface paint may be lashes bead over the flesh paint. Similar are jaspered red on the front and halfway bottom affected the perspective, bringing designed for use as a book. support: Oak, 44.8 / 44.6  35.4 / 35.3. head is completely transparent in irr. Summary residues of the original frame paint. beading is seen in the strands of Christ’s around the sides. There are no frames viewers closer to the picture plane. A thin Single plank, slightly thinner at the edges These two paintings are by the same hair. The use of what appears to be an on the reverses. Traces of nail holes in horizontal brown line beneath the present support: Oak, 40.9  30.6 / 30.4. Single than the center. Trimmed on all sides. The hand and were planned and executed as aqueous material over the oleaginous base the right side of the Gabriel panel align join between the lintel and uprights in christ as savior (db no. 48.96) plank, with slight bevels on the reverse. left edge (seen from the back) is cut into a pair. Although they are missing their for final touches can also be observed in with similar holes on the left side of the the Gabriel (partly under the white paint, The original red and brown marbling on a tongue the length of the panel. The frame: Not original. original frames, the identical treatment cat. 14. Virgin panel, marking the former location partly over it) shows that the point of view the reverse is overpainted with lead white reverse is painted with a predominantly of the reverses shows that they probably The halo was completed first in yel- of hinges. The x-radiograph also reveals shifted from eye level to below the lintel. support: Oak, 44.6 / 44.8  35.6 / 35.4. (fig. 1). yellow marbling. functioned as a diptych. The artist seems low paint, possibly lead-tin yellow, with nail holes where hanging devices were Similarly, the pedestal had been flush with Single plank. Top, bottom, and right sides to have been working from a familiar pro- dendrochronology: Earliest possible mordant and gold leaf applied afterward. attached at the tops and a closing device the frame and now appears to overhang dendrochronology: Earliest possible fell- have been trimmed. The reverse is painted totype, as very little preparatory drawing felling date: 1496; earliest possible creation Additional lines of gilding, usually shorter on the left side of the Gabriel and right side it slightly (the latter was also observed in ing date: 1519; plausible creation date: after with a predominantly yellow marbling, was required. The paintings exhibit a high date: 1498; statistically more plausible and with less rounded topography, are of the Virgin. The Virgin Annunciate). These two changes 1521. Has twenty sapwood rings. except for a thin margin along the right degree of finish. production date: after 1504. Has four sap- interspersed amid the longer rays. A whit- ensure that viewers focus on the subject side as seen from the reverse. supports: Wood not identified. Each preparation of support: Chalk/glue wood rings. ish gray paint in the crown, scumbled over of the panel, looking up at the lintels and ground continues to the edges of the dendrochronology: Earliest possible panel comprises a single plank, the Gabriel preparation of support: the underlying paint, is similar to whitish down at the pedestals. Short horizontal panel. The x-radiograph and cross sections White with measuring 38.8 / 38.7  23.3 / 23.4, and the felling date: 1507; earliest possible creation paint in the part in Christ’s hair and may lines across the pedestal reflection and show a thin intermediate layer of lead a barbe on all four sides. The right Virgin, 38.8 / 38.7  23.5 / 23.4. The reverses date: 1509; plausible production date: after represent a last refinement of the images. in the white paint of the pedestal itself unpainted margin has a barbe, and an and halfway around the sides are painted white with a small proportion of calcium 1515. This panel was planned and executed The contours of the cheek and chin were incised through wet paint and are incised line at the bottom has a ridge with a primarily greenish black marbling. carbonate applied overall. Striations vis- with the Virgin at Prayer, and its later dendro­ were first painted wider following an discernible on the surface, but their func- of ground beyond it, then unpainted ible in x-radiograph are diffuse, and the chronological dating should be taken into underdrawn line, and the collar of the preparation of the supports: White. tion is unclear. wood. For other reports of ground applied intermediate layer does not texture the account. Virgin’s original robe reached that pre- No intermediate layer was observed. The reflection on the black marble to prior to framing and manipulated while surface paint. liminary line. But the final flesh tones the right of Gabriel was painted narrower preparation of support: Chalk/glue. still wet, see Stroo et al. 1999, 145, 162 preparatory designs: Densely spaced describe a slightly smaller contour, and than first planned (the change is visible preparatory design: A simple contour There is a barbe on the left side and traces n. 12. No intermediate layer was observed, fine lines model the forms, with hatch- adjustments were made in the background in both irr and x-radiograph). The initial line marks the edges of forms in the hands, of one at the bottom right edge. On other but the lead white paint on the reverse ing and crosshatching used to show the elements as well. Several details illustrate paint stage followed the underdrawing, such as the fingers and palms. The white edges the (presumed) barbe is trimmed off. impedes x-radiography. fall of light on each surface and to create the care taken with the ultimate appear- but the form of the pedestal in the reflec- cloth at the neckline was also delimited. Cross section analysis finds a thin inter- nearly sculptural volumes. The border on preparatory design: A spare but broad ance of the painting: the ring finger on tion also changed to the stepped-in profile. There is little underdrawing in the facial mediate layer of lead white in oil with an Gabriel’s sleeve was drawn with two bands liquid contour defines larger forms such the Virgin’s proper left hand was painted The fold at the lower right was blocked features, and no clear evidence of a line admixture of calcium carbonate. on either side of a series of circles. The as the right cheek and chin. The face slightly narrower than the reserve left for in as free hanging, then painted over the between the lips. The little underdrawing pedestal’s base was drawn with a straight preparatory design: Scant and hard to and other details are not underdrawn it, and the fingers in front were painted underpaint to connect it to main body that exists is followed closely in the paint. rather than a stepped-in profile and seen determine. The most readily visible line in a medium that registers in irr, but longer than the reserve left for them. The of the garment. from eye level rather than from above. The paint layers: Oil paint. The paint was of underdrawing is at the separation microscopy suggests the use of a red pupil of Virgin’s proper right eye was In The Virgin Annunciate the dove seems brooch at Gabriel’s chest was sketched smoothly blended with little surface of the lips. The broad gray drawing mate- drawing material in these features, which underpainted slightly to the right of the to have an ocher underlayer (not seen with fewer lobes than were painted. His texture. The analysis shows that the blue rial can be easily confused with paint, as recalls the painted contours in cat. 17. final position, then repainted with black under the Virgin), which was used with a proper left thumb was bent in the draw- was mixed with calcium carbonate. The many finishing touches in paint are seen The ir and irr show the Virgin wearing a finishing touches. scumble of white at the back edge of the ing. The loose fold of his robe was changed composition was laid out with reserves in irr. One small line slightly left of the large cloak over her robe and a white head wings to create a soft outline appropriate left in the underlying paint for the figure, nose is discernible, and there may be a covering wrapped around behind her neck between drawing and the paint stage. christ as savior (db no. 35.70) to the depiction of feathers. The Virgin’s which overlaps the background slightly. thick line at the left side of the painting. and falling onto the far shoulder. The blue robe was finished after the background, The paint was used with fluid blending cloak was painted with a large collar or frame: Not original. No traces of red overlapping it in places, and the contour for large areas and liquid flourishes in the hood behind the Virgin’s neck. The white paint on the barbes of the panel (as of her book was changed from curved to details. The blue of the robe at the left was head covering was later finally changed to observed in the Virgin at Prayer). straight by painting the top left side over a translucent fabric with her hair showing support: Oak, 40.8 / 40.9  30.8 / 30.7. the background. through. Single plank, cropped at an angle on the left edge, possibly reduced by c. 2 cm on that side. Original green and black mar- bling on the reverse is overpainted with lead white.

282 283 The pupils were enlarged late in pro- Cat. 16. Quentin Massys paint layers: The paint is thin and was Cat. 17. Quentin Massys drawing is restricted to a line placing the Cat. 18. Master of the duction, using a paint that absorbs more used in a liquid manner, with texture thighs and knees as well as two lines that ir than does the original paint. The new created by feathered brush strokes rather mark the top and bottom of the fingers Benson Portraits desiderius er asmus and saint mary magdalen and saint paint also appears to have been used to than a buildup of material. The contours (in a slightly lower position). The briefly peter gillis mary of egypt (db nos. 39.77 – 78) add strands of hair and eyelashes, and it of the right hand and paper were painted sketched thigh shows it wider and may portrait of a man and Philadelphia Museum of Art, inv. nos. often beaded up, like aqueous material on over the surrounding dark paint, while the depict the figure in profile. portr ait of a woman 366 – 367 an oleaginous surface (a peculiarity also desiderius er asmus fingers on the left hand were restruck with (db nos. 47.93 and 47.94) Examined at pma: 20 – 24 October 2003 paint layers: The painter worked rela- seen in cat. 14). The additional hair had Galleria Nazionale d’Arte Antica, dark lines over the lighter flesh colors. Museum Mayer van den Bergh, Antwerp, by cm, rs, cvd tively freely but accurately, capturing the the effect of increasing the height and Palazzo Barberini, Rome, inv. no. 1529 inv. no. 368 facial features in only a few brush strokes. moving the part and the side of the head Not examined. Documentation and analyses Examined at kmska: 11 – 22 November He used zones of umber-colored paint to to the right. The new part in the hair was Summary unknown date at pma: x-radiograph by jm 2003 by cm, rs, cvd, av establish shading in the flesh, followed later painted over with a scumble of white peter gillis (db no. 107.132) Although several copies of the original 2003 at pma: Phase One visible and ir as by scumbles of lighter colors to create Documentation and analyses that may correspond to the whitish gray Koninklijk Museum voor Schone Kunsten, pairing commissioned by Erasmus and well as Inframetrics irr by cvd; binocular midtones. The final shadows were defined 2003 at kmska: Phase One visible and in the Virgin’s crown. The halo was first Antwerp, inv. no. 198 Gillis for Thomas More exist, the similar microscopy by cm, rs; cross sections by mt with more umber. The reddish outlines ir as well as Inframetrics irr by cvd, painted with lead-tin yellow rays, after Examined at kmska: 11 – 22 November dimensions of these paintings — both 2005 at nga: cross section analysis by mg around the figures, umber as well, recall av; binocular microscopy by cm, rs; which the additional strands of hair were 2003 by cm, rs, cvd, av replicas — suggest that they were made as a frames: Not original. the underdrawn red contours in cat. 15. x-radiography by at completed; a rounded mordant was then pair. The slight deviations from the under- Documentation and analyses The panels are extensively restored, not applied, carefully following the painted drawing in the Gillis and the substitution supports: Veneer of oak (est.). Mary Mag­ frames: Integral. This diptych is made in 2003 at kmska: Phase One visible and ir just the backgrounds but also the scum- rays and avoiding the new hair, and these of a rolled-up rather than a flat sheet dalen is 31.1 / 31.2  21 (total width, includ- the form of a book. Four raised bands are 1 as well as Inframetrics irr by cvd, av; bles in the flesh. lines of mordant were gilded. of paper suggest that the artist felt free to ing the lateral added strips; the width carved on the outside right edge of Portrait Cross sections were taken from two binocular microscopy by cm, rs make creative changes, which may imply from barbe to barbe, measured from the of a Man and the outside left edge of Portrait 2004 at kmska: dendrochronology by pk points in the foreground landscape, where Massys’ involvement in this copy. None x-radiograph, is 19. 2 cm). Mary of Egypt of a Woman, so they resemble the spine of a 2 a passage of light green seems to extend Summary frame: Not original. among the group of paintings is known to is 31.2 / 31.1  21.2 / 21.1. Top and bottom book when the diptych is closed. The outer from the right side of the Mary Magdalen These two panels are depicted as a diptych have original paint on the reverse, which edges were cut, and both panels were part of the face of each frame is painted support:  onto the left side of the Mary of Egypt. Cross in the background of Francken’s Banquet Oak, 61.3 / 61.4 47. Two boards, suggests they probably functioned as pen- cradled. Lateral margins are obscured by black, while the inner elements are covered sections analyses determined that the in the House of the Burgomaster Rockox. But it the left measuring 24.7 cm at the top, 25.9 dants rather than as folding diptychs. overpaint, so it is not possible to deter- with gold leaf and strokes of red and green layer structure in these two samples is appears that the image of Christ might cm at the bottom; the right measuring 22.3 mine if they were originally unpainted paint to imitate gilt and marble-edged very similar. Though definite conclusions have been begun as a Holy Face, which at the top, 21.1 at the bottom. The edges and the original panels veneered. The book pages. Two holes in the flat black cannot be drawn based on a single sample was trimmed at the left, altered to create are cut into a tongue to slide into the extensive losses suggest a complete trans- margin at the left side of Portrait of a Man from each painting, both cross sections a Christ as Savior by the addition of hands frame. The reverse was coated with a mod- fer during which any original unpainted match holes on the right side of Portrait of included two green paint layers above a and a scepter, and framed with the panel ern layer of lead white but appears to have edges were lost. Exploratory cleaning a Woman and were likely intended for tying buff layer with almost identical pigment of the Virgin. Original paint was observed been left untreated originally. would be required to resolve this question. the diptych closed. Three rectangles of 1 mixtures and a yellow green highlight beyond the barbes in some locations, indi- dendrochronology: Earliest possible The poor condition of the paintings may paper and parchment now serve as hinges, layer over a midtoned green layer. In Mary dendrochronology: Earliest possible edge of the painting. Lines of red glaze cating that painting occurred both before felling date: 1506; earliest possible creation reflect multiple structural treatments. but these are not strong and probably do of Egypt these pigments were mixed in a felling date: 1450; earliest possible creation define the hands of Christ, with lighter and after framing. The Virgin appears to date: 1508; statistically more plausible not represent the original situation. This dendrochronology: Not possible with proportion that yielded slightly darker date: 1452; statistically more plausible flesh tones applied following their guide. have begun as a Mater Dolorosa but was production date: after 1514. diptych may have had a leather binding production date: after 1458. These red lines may be comparable to the transformed into a Queen of Heaven dur- no exposed edges. and less yellowish paint layers. that held the wings securely together, with preparation of support: White. In Mary Magdalen the reserves in the red drawing material used in the features ing the painting process. Final touches preparation of supports: White. A ran- the present panels as bookplates. preparation of support: White. In con- underlying paint were generalized and of the Virgin (see description above). The of paint, in a watery black and a grayed Unpainted margin with a barbe on all domly brushed intermediate layer seen trast to the Virgin at Prayer, the ground layer sides. The striated brown intermediate did not include the jar. During the supports: Unidentified wood. Portrait of a underdrawing for the scepter lies over white, plus the gilded haloes, were added in the surface of the Mary Magdalen across here was evidently applied prior to fram- painting process the proper right thigh Man is 11.2  7.35 / 7.4. Portrait of a Woman is the paint of the robe and was carefully to unify the images. layer over the ground was used with a the chin and folded hands is also visible in ing, as it extends to the edge of the panel was lengthened (lowering the knee) and 11.2 / 11.35  7.3 / 7.4. followed in paint. The underdrawing in glaze to create the background color. x-radiograph and ir. Cross section analysis at the top, right, and bottom; the left side extended to the right. Contours of the the brooch is higher than seen in the paint. preparatory design: A fine, lightly shows a similar layer in Mary of Egypt. The preparation of supports: White (est.), was cropped at an angle, forming a beveled left leg in x-radiography match the paint sketched contour was redrawn in some ir suggests that it is a lightly pigmented with a striated white intermediate layer area that resembles an unpainted margin. paint layers: The painting was started surface exactly, although the reserve was areas with a coarser, heavier line (i.e., in the layer. Although originally interpreted as a visible in the x-radiograph. The top may also have been trimmed before it was framed, then finished follow- lower and straighter than the final render- proper right hand). The underdrawing was paint layer corresponding to the landscape slightly at the perimeter. The first ground ing framing. The hands and scepter were ing. In Mary of Egypt the loaves were not left preparatory designs: The composition not followed closely in paint, with small (because of the presence of a few particles appears to have dried before a barbe-like added later, only after the red glaze was in reserve. for Portrait of a Man was underdrawn with ridge was created (perhaps only out put on the robe and strands of hair were adjustments frequently made in the out- with properties of malachite), the buff- thin red lines, which did not register in of paint). An incised line runs through the defined. They were conceived together, lines (i.e., the outside of the left hand and colored layer immediately above the ir or irr but are visible through the paint ground at the limit of the painted image. for the staff of the scepter is encircled by the right edge of the book). ground in the samples could be this layer. Summary under the microscope in the fur collar, the the fingers but not painted under them. Samples from Mary of Egypt reveal large The continuing landscape elements, espe- eyes, the hands, and the cuff of the right preparatory design: The scant under- The irr shows some suggestion of eyes individual black particles. It is unclear cially the diagonal formed across the two hand. The red drawing material can also drawing includes a line near the lower just to the left of the painted ones, which if these particles are integral to the ground panels, point to an original function in be seen on the surface at the back of the left pearl on the medallion, another at seem to have been completed, including or perhaps to a layer of black between two contiguous format, either as a diptych or ear and was used for shading in the final the right side of the opening of the robe, the whites. For these eyes to be centered, layers of ground. as the exterior wings of a triptych. Yet the image. The man’s costume was drawn with and a horizontal line along the bottom the panel would have to be two 2 cm preparatory designs: The scant under- structural treatment(s) make it impossible pleated folds radiating from his neckline wider at the left — where it has clearly drawing for Mary Magdalen consists of a few to describe the format with certainty. The but is painted with a flat front. His eyes been trimmed. contour lines in the face near the figure’s cut edges at top and bottom indicate that were slightly lower and looked to the right proper right eye and through her left foot the pictures were once taller, though it in the underdrawing. The red line under- and her thighs, indicating only the general is not clear by how much. The similarity drawing in the Portrait of a Woman does not layout. Contours appear around the figure of paint type and structure in the cross show in ir or irr but can be glimpsed in the ir and irr, but they seem to be sections suggests synchronous creation. beneath the paint with a microscope in underpainting rather than underdrawing, the sitter’s eyes. as they define shadow rather than form. paint layers: The paint was handled in The figure may have been planned more broad masses. In Portrait of a Man reserves in profile, but the drawing is too cursory were left for the sitter’s head and hands, to be certain. For Mary of Egypt the under- the ruffle at his cuffs, and his fur collar (the reserve for the collar had a somewhat different shape on the left side). The edges

fig. 1 Micrograph from the reverse of the Virgin at Prayer, showing 1) the original marbling and 2) white overpaint

284 285 of forms were clarified at a late stage by Cat. 19. Master of the preparation of supports: White. Cat. 20. Master of 1499 preparatory designs: Spare liquid con- not followed precisely in the application flower vase, the ewers to right of the Cat. 21. Master of 1499 repainting adjacent contours. Details No intermediate layer evident. tours describe the compositions of both of subsequent highlights and shadows on Virgin’s head, the pillow, and the prayer were applied in linear fashion, with thin Female Half-Lengths panels and their most important features. the tiles. book. Only the smallest or most elaborated preparatory designs: A thin spare line the angel gabriel and virgin in the church strokes of paint used to embellish the The underdrawing for Gabriel uses hooked forms were added over previous layers in the Saint Peter describes the contours the virgin annunciate paint layers: The paint is full-bodied with reverse: salvator mundi, 1499, fur, eyebrows, and hair. The x-radiograph saint peter and saint paul lines to signal folds in his robe. The draw- of paint, including the orange and the of the flesh tones, while those for the (db nos.102.107 and 102.106) so that brush strokes are textured. It is and abbot christiaan de hondt shows the original position of the irises, (db nos. 29.57 – 58) ing for the red drapery over his arm is tumbler on the windowsill, the lilies, the drapery and attribute are heavier. Some Staatliche Museen zu Berlin, economical in application. The lower paint with reverse: robrecht de clercq carried into paint, which were directed Museum Mayer van den Bergh, Antwerp, readily visible in irr, with hatching that paper scroll hanging off the edge of the of the thicker lines may be in a liquid Gemäldegalerie, inv. no. 548 layer describes form with adjacent strokes Koninklijk Museum voor Schone toward the female sitter. It also shows lines inv. no. 369 denotes the shadow in the fold below cabinet, the candle, and the lock and material. The key was drawn in a more Examined at smbg: 22 July – of highlights and darks. Over this the art- Kunsten, Antwerp, inv. nos. 255 – 256 of small dots following the neckline and Examined at kmska: 11 – 22 November his hands. The underdrawing differs in handle. The x-radiograph also shows that vertical position than it was painted. Very 1 August 2003 by cm, cs, rs ist used glazes, often with the same brush (reverses 530 – 531) center front opening of the man’s jacket, 2003 by cm, rs, cvd, av style and execution from that in cat. 21. the line for the curtain rod across the nave little underdrawing was revealed in the stroke enhancing the depth of shadow Examined at kmska: 11 – 22 November which correspond to the black trim. The Documentation and analyses The painted draperies usually follow the was first painted slightly below its present Documentation and analyses Saint Paul, although the sword was drawn at adjacent to a middle tone. This is true in 2003 by cm, rs, cvd, av highlighted areas of pleats were blocked 2003 at smbg: Phase One visible and ir as underdrawing closely, though the con- position. The ewer to right of the Virgin’s 2003 at kmska: Phase One visible and more of an angle than it was painted, with The Angel Gabriel not only of the green and in with paint following the initial under- well as Hamamatsu irr by cs; binocular tours of the robe are often softened in the head was initially shorter and had straight Documentation and analyses ir as well as Inframetrics irr by cvd, its point closer to the saint. red garments but also of the blue robe, drawing. Uneven densities in the lower microscopy by cm, rs; x-radiography by gs paint compared to the drawing. Gabriel’s sides, then was heightened and given the 2003 at kmska: Phase One visible and ir av; binocular microscopy by cm, rs; which is underpainted using white with portion of the x-radiograph may result paint layers: The paint is thin and was face is painted larger than it was drawn, present profile. as well as Inframetrics irr by cvd, x-radiography by at frames: Integral. The thickness of the a small admixture of blue and completed from the first jacket being painted out. applied in opaque blocks, working back with his mouth placed slightly lower. av; binocular microscopy by cm, rs; closed diptych is 2.7 cm. The butterfly- with touches of pure blue. The reddish tan In the Portrait of a Woman reserves were frames: Original. Semi-integral, with and forth across boundaries. In the Saint No hatching was observed in The x-radiography by at shaped hinges, which are attached with intermediate layer is used as a midtone in left for the hands, dog, face, and collar. the vertical members carved out of the Peter the sky was painted first, then the Virgin Annunciate, but the underdrawing Summary 2004 at kmska: dendrochronology by pk modern screws and are not original, are the flesh as well as showing through the Details such as the eyebrows, hair, and the same panel as the support. Old nails vis- figure, and finally the landscape. The for her face is fully comparable with that Although the original hinges have been relatively large and fully x-ray opaque, green draperies. fur on the dog were applied in thin lines, ible in the x-radiograph show the original foot was painted over the background at for the angel. Her features are indicated lost, these two paintings were most likely obscuring the area in which evidence The dove was left in reserve from the virgin in the church (db no. 27.53) while red contours define the outlines in hinges were larger than the present the bottom, however, and the hand was with short brush lines, all underdrawn conceived as a diptych and executed by the of original hinge nails might exist. The textured paint around it. Incised lines the hands and face. The woman’s eyes, like ones. No traces of a hanging device were painted over the robe. At the end, edges somewhat higher than painted. Contours same hand. The artist made small changes frame: Integral. The gilding, which has hanging devices at the top of the reverses through the wet paint of the dove to the the man’s, were changed in paint, with observed. The nail that secures the hook were sharpened, bringing the landscape of folds, only faintly visible, are drawn in the paint stages from a design that was a bright white layer under it, is not origi- are also modern, but the hole above each warmer color of the intermediate layer her gaze shifted to the right, away from now used to close the wings appears to be over the figure or vice versa, and a black in the Virgin’s blue robe. The shoes were well understood in the drawing stage. nal, and the black line demarcating the might be a remnant of original hanging define the head and the bottom contour the man. modern, but two nail holes to its right on contour was added to delimit the flesh not underdrawn. There was a change The pinhole at the vanishing point in the interior border may not be original. devices. The closing device may be original, of the right wing. The gold leaf was the x-radiograph of the Saint Peter might tones. Small touches of bright paint give between drawing and painting in the Virgin’s chamber door and the plumb The inscription on the bottom section judging by the appearance of the nails applied over this reddish tan intermediate be remnants of an original closing device. a sense of the fall of light. In the Saint Paul chest at the lower right, but it is not clear line through it prove the artist’s concern of the frame is probably not original. in the sides opposite the hinges. The layer, as shown by the streaks in the x- Summary These holes penetrate the thickness of the the blue robe was first painted red, then whether the drawing depicted a cloth that with the proper construction of the point X-radiography shows nails and traces x-radiograph also shows the remains radiograph that continue under the gilded This pair was conceived and executed as a panel and correspond with a single hole changed to blue, and finally glazed with was to go over the chest or the front of the of view for this interior. The gold leaf on of nails at the right edge in line with simi- of two filled holes in the bottom edge areas. The gilded rays under the dove were folding diptych. The diptych was designed on the right edge of the reverse of the Saint red (now significantly abraded). Final con- Virgin’s robe. the frame is probably not original. The x- lar marks at the left edge of the adjoining of each panel. The gold leaf is not original, applied over a tan mordant. The edges to appear, in both opened and closed posi- Paul. A closing device attached in that way tours in the shaded areas were reinforced An elaborately prepared perspective radiograph suggests that a painted surface panel, where they once secured hinges. It and the x-radiograph reveals brush strokes of the composition are hatched with a red tions, as a book, with imitation gilded and would cross two single planes, as the wings with black. system was revealed in The Virgin Annunci­ lies under the present gold leaf, although also reveals the location of an original in the interior profile of the frame, which glaze, which created shading and a sense marbled page edges and with raised bands of a triptych (which close like window ate, with a single vanishing point in the samples were not taken to verify this. As closing device at the left edge, with match- suggests that it was originally painted. In of depth. The folds of red robe behind on the spine. It is possible that it was shutters), rather than lock the edges of the door to the Virgin’s chamber at the top left there is no underlying ground, it is not ing evidence at the right edge of the donor addition, a loss in the arched top of the the angel were slightly altered from a first bound in leather, or had a leather spine, panels as in a diptych (which close like Summary of the panel. X-radiography revealed a pin- clear that the paint on the sides and reverse panel. Fills at the top in both panels indi- Virgin panel is visible in the x-radiograph paint stage to the final one. The tips of the now replaced by the parchment and paper books). The gold leaf is not original. The We believe that this pair were originally hole here, probably from a small nail used is original. The diptych may have been cate that they once had hanging devices. but not in the gold-leafed surface. angel’s wings lie over the gilding. hinges. The changes in both sitters’ gazes reverses were originally painted but were the wings of a disassembled triptych, with to construct the perspective lines. irr also intended to hang from a chain attached The technique of The Virgin Annunciate support: Oak, 37.1  20.1 / 20.4 (includ- away from each other are unexplained later stripped bare. The bole for the gold the holes through the thickness of the supports: Each wing consists of a single revealed a plumb line that runs through to the apex of each arched panel, as in the resembles that of the companion panel in ing frame). Single plank, with two-sided but might suggest a possible intervention leaf lies over residues of original ground panels being traces of an original closing board with vertical grain. The two panels this point, from the top center of the background of the Christiaan de Hondt its economy of means and in the texture integral frames. Salvator Mundi is painted of the patron. The painting technique, and paint in the wood grain on the reverse, device. Additional arguments against the are the same size and format — 18  11.5, arch above the hall leading to the Virgin’s diptych attributed to the same master of the paint. The blue of the Virgin’s robe on the reverse. with a red underdrawing and opaque indicating that the gilding took place after original format being a diptych are the tall, arched at the top — and are evidently chamber through the floor in the fore- (cat. 21). In such a display, it may not have is conventionally constructed, with a layered paint used with linear detail, is the reverse was stripped. narrow dimensions of the panels and the unaltered. The sides and back have dark ground. Incised lines were used to define been necessary to paint the reverse. It is dendrochronology: Earliest possible lower quality azurite under higher quality not typically Netherlandish. fact that the saints gaze into the middle paint directly on the wood (without a the perspective in the tiled floor, all con- likewise possible that the present black felling date: 1467; earliest possible creation supports: Unidentified wood. The Saint azurite. Yet an echo of the unusual paint- distance and appear to be looking toward ground layer). verging at the vanishing point. Incisions paint may have been applied to the reverses date: 1469; statistically more plausible Peter is 18.9  7.8; the Saint Paul is 18.9 / ing technique used for the angel is seen in a missing central scene rather than toward for the chest at the lower right also follow after the removal of a deteriorated original production date: after 1475. The plank is 19.1  7.8. Single planks with vertical grain. dendrochronology: Not possible. the white curtain and the wooden vault each other. the perspectival schema. Other incisions decorative treatment. from the same tree as that for Abbot Chris­ The reverses are flat, without a bevel. behind the Virgin, where red lake serves preparation of supports: White. An mark straight lines in the architecture, tiaan de Hondt. Microscopic traces of ground and blue as the underlayer to indicate shadow. The intermediate layer is visible in the x-radio- and those on the wooden ledge of the prie- paint remain in the grain of the reverses, reddish brown intermediate layer is used preparation of support: White. No graph as a randomly applied striated layer. dieu at the lower right continue under the especially around the outer edges. as a middle tone throughout the Virgin’s intermediate layer. No such intermediate layer was observed Virgin’s robe where it folds over this ledge face, and as the shadow under the chin, in the Master of 1499’s diptych in Antwerp at the knee (such a carefully constructed preparatory design: The precise, detailed under the lips, and around the eyes. As (cat. 21). perspective system is not evident in any underdrawing in a liquid material was in the Gabriel panel, the incised lines in part of cat. 21). In contrast, the floor tiles at begun with finer drawn lines, some the floor at the foreground are not always the bottom of The Angel Gabriel were incised of which may have been in a dry material. followed in the paint. The painting tech- in wet paint, but the perspective lines do Fold lines were later reinforced and the nique does not make use of the under- not converge in a single vanishing point as drawing worked up with hatching and drawing (as seen in cat. 21). in the Virgin panel, and the incisions were crosshatching. The artist consistently In the space between the two columns drew the shadows rather than the high- to the left of the Virgin, a third column lights. Most of the drawing is secure and can be seen in the x-radiograph that was definite, but there are restruck contours in decorated with a diamond pattern (see the floor tiles and at the Child’s head. The entry). The painter left reserves for the drawing was used under thin scumbles of light-colored paint for shading in the architecture and for a midtone in the flesh. It is sometimes difficult to distinguish the underdrawing from paint strokes in red lake — the paint is indicated by the presence of clumped black particles and is generally darker in the irr than are the

286 287 the artist applied gold leaf, followed by a inscription in the background arch and Cat. 22. Master of the Lille underdrawing and generally higher finish light-colored paint scumble. The donor’s the corresponding reduction in the size of the paint surface may suggest that face has more color than the other flesh of the arch. On this image the candle is Adoration 3 the master completed this entire panel. tones, and there is a reddish brown paint glowing rather than completely snuffed 2 The finishing touches were handled with under the pewter vases, the snuffed-out out. A reddish brown underpaint defines the trinity and saint jerome the same finesse as on The Trinity. candle, and the abbot’s staff. The oranges the candle and the gold trim on the robe, (db nos. 49.112 – 113) on the mantle were left in reserve, and as on the fronts of the panels. Private Collection, on long-term loan the cross depicted in the hanging diptych Sometime after the diptych was com- to Harvard University Art Museums, Summary in the background was incised into wet pleted, the coat of arms of Robrecht de Fogg Art Museum, inv. nos. 6.2005.1 – 2 1 Despite the discrepancy in the dendro- paint — first the vertical, then the hori- Clercq was painted over the coat of arms Examined at huam: 1 – 5 March 2003 by chronological dates, these panels share zontal stroke. As in the floor tiles for the of the abbey Ter Duinen at the bottom left. cm, ap, rs; and 5 February 2005 by cm, a working method that ties them closely facing panel, cracks between foreground This was done without taking care to paint rs, cvd together. The evidence for blue clouds originally spilling from The Trinity into planks in the floor here were incised into out the abbey’s coat of arms, which is still Documentation and analyses the upper left corner of the Saint Jerome the wet paint. visible through the De Clercq arms. 2003 at huam: Phase One ir and Infra­ irrevocably links their compositions as The placement of the socle with the metrics irr by ap, rs; binocular micros- well. The panels were also prepared in the abbey’s coat of arms changed slightly from abbot robrecht de clercq copy by cm, rs; x-radiography by hl same way, with a fluorescing intermedi- underdrawing to paint. Also, De Hondt’s (db no. 27.128); reverse of abbot 2004 at huam: dendrochronology by pk 3 ate layer applied over the ground prior coat of arms has the hounds in the lower christiaan de hondt 2005 at huam: Phase One visible and ir to underdrawing. The drawings were left and upper right quadrants, whereas by cvd; cross sections by cm 2 frame: Integral. The blue-green trim done in the same material and a similar the position is reversed in the Salvator 2005 at nga: cross section analysis by with rosettes appears to be a late addition, preparatory designs: The underdrawing paint layers: The paint in The Trinity was style. The first paint stage of The Trinity Mundi. The abbot’s proper left thumb was mp and mg underdrawn lines. There are scattered tiles. The tiles in the nave at lower left were bent in the underdrawing and first paint probably dating to the time the portrait appears to have been done in a dry mate- applied efficiently in a few thin layers. may have been executed by the workshop, guidelines for arches near the altar. painted with a shallower perspective than layer, then finally painted flat against of Robrecht de Clercq was added. frames: Not original. The panels were rial, which may have been soft, for the Because the underdrawing did not articu- given its reliance on a more extensive painted outside the frames. lines are broad in places. In cross section late every detail, the reserves were only and detailed underdrawing and the less The urn with flowers at the lower was drawn. The text board on the fluted his hand. The artist painted over the por- preparation of support: White. No right was not included in the underdraw- column at the left was not left in reserve tion of the robe immediately to the right the underdrawing forms an irregular layer generally defined. The painter worked in skillful handling of the putti and the intermediate layer was observed. supports: Oak. The Trinity, 42.8  31.4 / with dispersed particles that seem to have a controlled manner, but not fastidiously. background. The master must have com- ing. The Virgin’s crown was drawn wider, (that is, the column was completed). The of the miter using a different arrangement 31.7, comprises two planks with vertical preparatory design: No underdrawing some medium incorporated. The drawing Eyes, noses, and mouths were executed in pleted the figures of Christ and of God as in Jan van Eyck’s Virgin in the Church Child’s proper left hand was changed of folds. grain, glued and doweled together. The observed. lies over an isolation layer. The contours a few brush strokes, rapidly but accurately. the Father as well as the entirety of the in Berlin. No jewels were drawn in the so that it is now more closed than it was reverse was thinned and had a cradle, for The Trinity were drawn freehand, and Although the background and second- Saint Jerome. There is no evidence of paint Virgin’s bodice, and the Christ child’s in the first painted position, where the salvator mundi (db no. 27.129); paint layers: Marbling was applied in now removed. Five blocks of wood were a number of changes between the drawn ary figures were painted in more cursory on the reverses of these panels, nor on the proper left hand was underdrawn with a fingers were spread wider and the thumb reverse of virgin in the church multiple layers of paint. Transparent and inserted across the seam from the reverse. and painted stages, especially in the hands fashion, God the Father and Christ were reverses of the other versions, which indi- wider knuckle span and the thumb lower bent lower. translucent overall layers were intermin- Thin strips of wood are attached to the than painted. frame: Integral. and in the putti, suggest that this was the reworked with additional paint layers, cates that they were probably pendants gled with splatter-applied opaque paint, lateral edges. Saint Jerome, 41.9 / 41.7  31.8 / first of several painted versions. Single including delicate highlights and shadows rather than a diptych. abbot christiaan de hondt preparation of support: White. An creating an illusion of shallow depth. 31.9, is single plank with vertical grain. paint layers: The paint application is lines define the eyelids. Parallel hatching as well as umber for shading areas of the (db no. 27.54) intermediate layer with diagonal stria- The paint for the portrait of De Clercq, Possible original adze marks appear on the highly detailed and decorative. A reddish was used to establish shading but was flesh. Scumbles of light paint over these tions is visible on the paint surface and is which is thick and buttery, with brush- reverse. Strips of wood were attached to brown first layer was used throughout the frame and dendrochronology: Same as angled in various directions, not necessar- areas created midtones. The final defining confirmed with x-radiography. This layer marks evident, registers more clearly in the left and right edges. Numerous small Virgin’s face as a middle tone and as the for Virgin in the Church. ily following the volume depicted. darks and the outlines were added with is not apparent on the front of the panel. the x-radiograph than the paint depicting blocks of wood were inserted into the shadow under her chin, below her lips, and The underdrawing provided a rough more umber (est.). Lines incised through support: Oak, 37.5  20.2 / 20.3 (including De Hondt. Nonetheless, the marbling is reverse, and the surface was thinned at around her eyes. Over this, liquid white preparatory design: See Virgin in the Church. rather than a specific guide for the painted wet paint revealed the sky beneath and frame). Single plank, with two-sided inte- visible, especially through the flesh. The the left and bottom right edges. paint was lightly scumbled to create a The underdrawing is not visible through forms, so small deviations can be seen also defined the separate feathers in the gral frames. Robrecht de Clercq is painted on painter completed only the foreground transition to the illuminated parts of the the paint layers as readily here as elsewhere, dendrochronology: For The Trinity: between drawing and painting. The putto upper left putto’s wing. the reverse. corner of the base of the prie-dieu, leav- face. The paint was applied in complex, probably because the intermediate layer earliest possible felling date: 1522; earliest on the right was drawn with a larger head At some point, presumably following ing the marbling as the shadow. On the multilayered sequences. For the Virgin’s preparation of support: White. No lies over it; the underdrawing is, however, possible creation date: 1524; statistically than finally painted, although the first the separation of the two panels, blue bottom ledge, the abbey Ter Duinen’s coat crown (fig. 2), dark brown paint came first, intermediate layer was observed. visible under the red lake and in the shad- more plausible production date: after 1530. paint stage followed the underdrawing. clouds that originally filled the upper left of arms and two sets of initials “ch” were followed by pink, jewel colors, highlights, ows of the flesh tones. Inscriptions (on the For Saint Jerome: earliest possible felling Christ was underdrawn with the left corner of the Saint Jerome, visually linking it preparatory design: See Virgin in the Church painted out but are visible because they and finally the brightest dots of light and globe and at the top of the background date: 1500; earliest possible creation date: nipple higher, and the first paint stage to The Trinity, were painted over with brown (above). The underdrawing was carefully project from the surface (see entry). darkest areas of shadow. Light from the arch) were underdrawn significantly larger 1502; statistically more plausible produc- retained that placement. The close adher- (the blue paint is still visible beneath the followed in the paint for the most part, windows was painted in two stages: first than painted (and “et” was underdrawn tion date: after 1508. ence of the underpaint to the under- brown in micrographs and cross section: although De Hondt’s hairline was shifted pink, then a lighter white. Tiny touches “e”). The foremost interior ribs with col- drawing may suggest that the workshop fig. 3). The skull was painted larger than up compared to the drawing, and there are Summary preparation of supports: White. There of blue on the mullions at the sides of the umns were not underdrawn. blocked in the underpainting. the reserve left for it. Incised lines define fewer beams in the ceiling and wider spac- The original diptych was by a single is an unpigmented oil intermediate layer windows suggest light coming through The contour drawing in The Trinity is the tabletop, both side and back, and pos- ing between them; the perspective of the paint layers: The paint was handled in a artist with access to a linear copy of Jan on top of the ground that fluoresces blue glass. The embroidered edge of the comparable to that in the Saint Jerome, but sibly the edge of one book. Horizontal beams also seems to have been changed. similar manner to that on the front of the van Eyck’s Virgin in the Church, but without brightly when examined in cross section. Virgin’s red cloak was painted with the latter has no hatching. In the Saint lines in the background architecture were panels, except that it is generally thicker. knowledge of the painting itself. The A randomly applied striated layer is visible three underlying colors and two colors paint layers: The paint application is Jerome numerous changes in the back- also incised, though these do not continue Christ’s flesh was modeled with pink paint, underdrawing on the left panel faithfully in irr. of “jeweled” dots. Such attention to sur- as considered and elaborate as in Virgin in ground landscape, such as the hills and into the upper left corner. The less explicit rather than white, and the pink paint has followed Van Eyck’s composition, although face detail distracts from the overall sense the Church. The jewels on the miter were the bridge, occurred between drawing and more body than the white used for the some aspects — especially the distant choir of light and shade and diminishes the painted using the multilayered technique painting, and the book pages were under- final touches in the Virgin and De Hondt and angels — were not fully understood, illusion of depth in the scene. seen in the jewels on the Virgin’s crown. drawn higher than they were painted, images. The thickness of the paint may and revisions were already planned in the Reserves were left for the vase at the The abbot’s hands and the architecture are again implying that this was the first be due to the anticipated wear on an exte- first paint stage. This rendering empha- bottom right and for jewels in the Virgin’s worked in the same technique as similar of several versions of the composition. rior surface. sizes detail and explicit description over bodice, neither of which was underdrawn features in the companion panel, with a Christ’s hair, the book, and the orb volume and form. After Robrecht de Clercq or part of Van Eyck’s composition. The white scumble over the underdrawing, under his feet were all left in reserve, while became abbot of Ter Duinen, his portrait Virgin’s crown was initially painted so although white paint was not used to the edge of the pedestal and divisions in and coat of arms were painted by a differ- that it extended down the back of her model the contours of the donor’s face. the floor tiles were incised in wet paint. ent artist on the reverse of the previous head, as in Van Eyck’s original. The change The fire in the fireplace was depicted with Late additions include the forward interior donor’s portrait, with rosettes added to was made before the final surface details white and red painted flames over which of the crown were completed. Incised arch and columns and the page turning in the frame to match the original decoration lines, which were used minimally, can be the center of the book. Two other changes on the Salvator Mundi, facing this portrait seen in the divisions between brown floor may be related: the smaller type for the when the diptych is opened.

fig. 2 fig. 3 Macrograph of the crown from Cross section from upper left the Virgin in the Church corner of Saint Jerome, showing 1) the ground, 2) original blue paint, and 3) brown overpaint

288 289 Cat. 23. Master of the Magdalen Cat. 24. Master of the dendrochronology: Earliest possible the present donor’s forehead); indications willem van bibaut (db no. 42.84) the five wounds of christ felling date: 1475; earliest possible creation of the bottom edge of a head veil, which Legend and unknown French (db no. 42.104); reverse of willem van Saint Ursula Legend date: 1477; statistically more probable came low over the previous donor’s fore- frame: Integral. Regilded, with outer artist bibaut production date: after 1483. head, similar to that of the older female edges painted black. The top frame mem- virgin and child and donor; and faint indications of the folds ber has at least three partly filled nail holes frame: Integral. preparation of support: Chalk in several three donors, 1486 of a different headdress. There may also be virgin and child and at the center, possible traces of a hanging layers and (possibly) a single layer of lead preparatory design: The underdrawing Koninklijk Museum voor Schone Kunsten, a suggestion of fingers below the present willem van bibaut, 1523, with device. The inscription lies over the new white upper ground, as suggested by cross consists of sparse outlines and restruck Antwerp, inv. nos. 5004 – 5004bis donor’s hand. reverse: the five wounds of christ gold and is thus not original. X-radiography section analysis. If lead white was used for contours, similar to that on the front Examined at kmska: 11 – 22 November The fur trim on both female donors’ Private Collection reveals empty nail holes with traces of an upper ground, it differs from the inter- of the panel. The heart, the inri sign, 2003 by cm, rs, cvd, av garments was a late addition. An earlier Examined at sral 15 – 27 September 2003 corrosion in the locations of the present “ ” mediate layer on the facing panel, as it is and the right wooden brace at the foot stage showed the wrists and the entire by cm, rs, cvd, av hinges as well as modern screws and nails. Documentation and analyses not evident in the x-radiograph and it does of the cross are painted smaller than backs of the hands, whereas the fur cuffs The shaft of an old corroded nail appears unknown date at kmska: cross sections not mask the underdrawing. Documentation and analyses underdrawn. Three pins were drawn in now cover the backs of the hands. The at the center of the right frame member, by Susan Farrell 2003 at sral: Phase One visible and ir as the cross, but only two are painted. preparatory design: Most of the panel young woman’s hands were part of the under the present nail securing the cur- 2003 at kmska: Phase One visible and well as Inframetrics irr by cvd, av; binoc- was drawn in a single stage, with contour original portrait, and a reserve was left rent hook. paint layers: The paint has a buttery ir as well as Inframetrics irr by cvd, ular microscopy by cm, rs; x-radiography lines delimiting large shapes, some with that corresponded to the drawing, but this texture that captures brush strokes. Colors av; binocular microscopy by cm, rs; by at; dendrochronology by pk support: Walnut, with left frame member the slightly “wavering” quality seen in the was not followed in the final painting. The are opaque and superimposed from dark x-radiography by at; cross section analysis of oak. Overall the panel is 30.4  20.2 / first stage of underdrawing in the Virgin male donor’s hair was changed after the to light. The haloes around each of the by Susan Farrell and mg virgin and child (db no. 42.83) 20.4, and the oak strip at left is 2.5 cm wide. and Child. Large shadows are indicated first stage of painting, and his chin was five wounds have a tan mordant under the 2004 at kmska: dendrochronology by pk A butt join attaches the oak and walnut with long parallel hatches, smaller ones lowered compared to the underdrawing. frame: Integral. Regilded, with outer now mostly missing gold leaf. Changes sections. Remnants of old, repaired dam- with short, dashlike hatches or with The older female donor’s head veil also edges painted black. The inscription lies between the first and final paint stages virgin and child (db no. 19.38) age appear at the left edge of the walnut 4 zigzags. The deepest shadow at the far changed at this stage, when it was painted over the new gold and is thus not original. include the horizon at the right side, the plank, especially at the top left corner. frame: right edge of the male donor’s torso is the lower over her face at the right, but not The hinges have old as well as more recent larger skull at the base of the cross, and Integral. Thickness 2.5 cm. Hinges X-radiography revealed two modern metal angels, and contour lines have a slightly only occurrence of crosshatching in the as low as indicated in the drawing. The nails. The top frame member has four the smaller braces that secure the base and nails possibly original. Corrosion staples in the walnut, one at top and one at “wavering” quality. This drawing is com- three donors (db no. 19.127) entire diptych. numerals at the top of the panel were partly filled nail holes at center, possibly of the cross. traces of an empty nail hole at the center bottom, securing a check in the wood; they parable in handling and execution to that A second stage of underdrawing was painted rather than applied in gold leaf. remnants of a hanging device, and two at of the left frame member marks the loca- frame: Integral. Both the inner and outer are not visible to the naked eye. The Five in the donor panel but has a sketchier confined to the young female donor at the the right. tion of a now-missing closing device. Two parts of the frame were originally painted Wounds of Christ is painted on the reverse. Summary partial old nails in the top of the arch may quality. The second stage of underdrawing, far right. This drawing lies over a previous, rather than gilded: the inner frame was Summary support: Oak, 30.4 / 30.5  20.6 / 20.7. These two panels were not originally part have secured a hanging device. especially notable in the angels and in different portrait. Prior to the redrawing, dendrochronology: Not possible. reddish, and the outer frame blackish. This diptych may have been produced in Single plank, c. 19 mm thick (including of the same diptych. The underdraw- Cross sections show the inner frame the Virgin’s hand but also found sporadi- a layer of paint was apparently applied to This frame has a restoration history identi- a “phased” production process. The differ- the frame). The surface of the reverse is flat preparation of support: White. No ings and painting techniques are quite was originally finished with gilding over cally throughout the composition and block out the original donor. A single line cal to those for the companion panel. See ent preparation of the frames and panels overall and painted black. intermediate layer observed. dissimilar, and the use of two support a yellow-tan layer, and the outer frame consistently visible through the paint from the underdrawing of the first donor’s the Virgin and Child (above) for the hinges, and the late alteration of the Virgin’s gaze materials in a single diptych, much less a with red and black marbling. The frame to the naked eye, is fairly free, with the portrait can possibly be seen in the present dendrochronology: Earliest possible preparatory design: Underdrawn lines closing device, and hanging device. The suggest that an existing Virgin and Child was single panel, is unique, to our knowledge. is regilded, with a restoration history image being worked out on the panel to donor’s hair, parallel to the contour of the felling date: 1474; earliest possible creation are visible in the face and in the folds bottom frame member initially included adapted to fit a commissioned donor panel. We believe that the oak section (the entire identical to that on the donor wing. A some degree. The angels were drawn on underlying black headdress. date: 1476; statistically more plausible of the foreground sleeve. The face was the ages of the donors in mordant gilding. As the position, gaze, and gesture of the left wing and the left frame member of the dark autofluorescent layer lies over the top of curtains defined in the first stage. The underdrawing was followed in production date: after 1490. drawn with restruck lines for the crease at This mordant is x-ray opaque but also vis- Christ child were not changed, the panel right wing) came from one diptych, and original finish layers, followed by an even- The Virgin’s proper right hand was first paint except for the redrawn figure of the the right of the mouth, and the lips were ible beneath the present surface in raking was clearly conceived as the left wing of a preparation of support: White. Faint the walnut (most of the right wing) from textured dark red layer, a yellow-tan layer, underdrawn with straight fingers point- young woman, whose face was under- outlined with the mouth slightly opened. light. The painted frame was later gilded, diptych. The two stages of underdrawing traces of an intermediate layer appear to another. What may be insect damage at and gold leaf. ing diagonally down to the right, then drawn higher, whose body at the right was The left edge of the cowl was underdrawn and the ages inscribed in black paint over in the Virgin panel may point to workshop be present in the x-radiograph, especially the left edge of the walnut section sug- with curved fingers and a tighter grip on thinner and more curvaceous, and whose behind the ear, to the right of its present support: Oak, 33.2  25 / 25.3. Single plank the gilding, but the male donor was mis- participation. The redrawing of the young in the head and the sleeves of the Virgin. gests a reason for this intervention. the Child, and finally painted with longer veil on the right was drawn higher, farther location. The ear was underdrawn but with vertical grain. The reverse was carved takenly identified as “30” instead of “50.” donor appears to have been done by the The nails that initially attached the fingers, particularly the little finger. A right than the painted contour across her preparatory design: Thin, fine lines never painted. The lines in the sleeve and with a simple recessed frame and painted The size and shape of the numerals also same hand as the initial drawing of the hinges to the right panel were replaced red brown paint that registers in the ir bosom, and lower and straighter across describe the major contours of the com- face are generalized, and the paint does with a crucifix under a trefoil arch. The changed, and comparing the “6” in the other two donors and of the Virgin and with modern materials, while the nails on and irr and outlines forms should not be the forehead. Her hand and the cuff of position without hatching. This drawing not deviate significantly from their design latter does not appear to be original. date 1486 at the top of the panel with the Child. Thus both the underdrawing and the left panel are original. The paintings confused with the underdrawing. her sleeve were also drawn lower and to is only faintly visible and can be seen best except where the edge of the hood was The original surface can be seen in a new “6” in the age of the older woman on the paint application are comparable, and might have been joined at the same time the right. where the paint diverges slightly from shifted. The lines in the donor’s hands, small cleaning test that shows reddish paint layers: The paint is thin and was the frame shows they are not by the same only the initial preparation differed. The the modern staples were attached in the the indicated contour. The paint follows visible through the paint of his cassock, brown paint. applied in a straightforward manner. A hand. There are stars, also prepared with paint layers: The paint handling is simi- reverses were originally marbled. The the contours of the underdrawing fairly are painted, not underdrawn. donor panel, though joining possibly pre- white intermediate layer provided the mordant, before and behind each number lar to that in the Virgin and Child. The hair dendrochronology: present images on the reverses are most closely. ceded structural treatment. The dendro­ Earliest possible foundation for flesh tones, with strokes (as in the date at the top of the panel). under the young woman’s veil was appar- paint layers: The painting technique likely not contemporary with the creation chronological analysis of the left wing and felling date: 1474; earliest possible creation of thin lead white as a last campaign. ently worked wet-in-wet, then the veil was paint layers: Thin, superimposed trans- does not compare well with that of the support: Oak, 33.1  25 / 25.25. Single of the diptych. the date on the portrait indicate that the date: 1476; statistically more probable The curtain was laid in before the angels, scumbled over it. Final surface details in lucent layers are worked from light to dark. companion panel, nor with early Nether- plank with vertical grain. The reverse was former predates the latter. The portrait production date: after 1482. leaving areas in reserve so that the edges the flesh tones, such as the red coloration Shadows under the nose and chin and the landish painting in general. The buttery carved with a simple recessed frame. The may have been painted near Grenoble preparation of support: of the angels’ wings overlap a layer of dark on the older female donor’s cheek and the definition of features like the fingers are paint preserves marked brushwork, espe- White. The inte- original marbled finish (fig. 4) is now (where Bibaut was abbot at the Grande green; a lighter green then extends up to male donor’s stubble, were not achieved achieved with linear strokes of translucent cially in the donor’s white habit. Modeling gral frame has a chalk ground followed by covered by more recent paint depicting a Chartreuse) instead of the Low Countries. the tips of the feathers. The treatment with a glaze or a mix, but with scumbles brown paint. Similar translucent strokes of the garment was accomplished with two white intermediate layers, as shown chalice under a trefoil arch. The area between Grenoble and Avignon of the mouths, with a dark line separat- of opaque lines. of gray are used in modeling the Christ wet-in-wet blended strokes. A reddish in cross section. Cross section analysis was one of the few where walnut was used ing the lips, is identical in both panels. A The x-radiograph and irr both show child’s white drapery. The paint surface brown paint defines the deepest folds. The of the paint layers also reveals a chalk as a support for panel paintings. tan-colored paint was used to create the thicker, more opaque paint in the area generally shows little texture, with three flesh tones have an orange underlayer, lower ground followed by two intermedi- body of the throne, while light scumbles depicting the young woman than else- exceptions: the blue paint of the Virgin’s with the planes and features defined using ate layers of lead white, separated by a thin and dark paint later established zones where in the composition. The layer that drapery (a result of the coarse pigment opaque white scumbles and fluid high- layer of dark material, suggesting that of light and shadow. The Virgin’s eyes were blocked out the first donor in this position used to obtain the desired blue color); the lights. The donor’s garment was extended the panel and frame may share the same first painted open, then made to appear stops in the middle of the present donor’s gold embroidery on the Virgin’s garments; in several places over the red dots and gold preparation layers. The intermediate layers half-closed. The original irises became the hair. The x-radiograph has a slight indica- and the raised red dots that decorate the background, notably at the back of the are visible in the x-radiograph. pupils in the second version, so the new tion of a higher, smaller set of eyes (in gold background. A wash of red glaze cowl, the shoulders, and in the lower left preparatory design: There are two stages irises had to be enlarged. shades the left and top near the frame and corner. The hands were first painted with of underdrawing. In the first (not visible creates a shadow to the right of the figures. dark contours in a slightly higher position, to the naked eye as it probably lies under The red glaze of the shadowed area is dif- readily visible through the present surface. the intermediate layers), zones of parallel ferent from the color used for the dots. dashes model the Virgin’s drapery, zigzags indicate folds in the curtain near the

fig. 4 Macrograph from the reverse of the Three Donors, showing original red and brown mar- bling beneath surface paint

290 291 Cat. 25. Hans Memling ground plants. As was the case with Saint Cat. 26. Hans Memling of lighter and darker paint over a smooth Cat. 28. Jan Mostaert and with three “tails” beginning some- John the Baptist on the companion panel, medium-tone foundation. Final touches 3 what higher. Figures not present in the Saint Veronica was originally adorned with of warm brown glaze complete the model- underdrawing include the four individu- saint john the baptist virgin and child, 1487, and christ appearing to a nimbus, for which the rays were incised ing of the cheeks. Both highlights and als behind Saint Michael and Christ and and saint veronica maarten van nieuwenhove his mother in limbo and in the wet paint. The ring finger of Veroni- reflected color on the bottoms of the jew- the smaller figures at the far right. Musea Brugge, Hospitaalmuseum Sint- kneeling female donor with the ca’s right hand was lowered after an earlier els were described with separate strokes Janshospitaal, Bruges, inv. no. osj 178.1 redeemed of the old testament paint layers: The paint was freely applied. saint john the baptist paint stage. Her red robe overlaps the green of various colors over a base tone. The red Examined at smb: 29 September 2003 by Although no longer readily discernible, (db no. 12.23) with reverse: of the landscape, beyond the reserve left glaze of the Virgin’s robe was applied in 1 cm, rs, cvd, av 2 different hues were used for Mary’s gar- skull in a niche for it, at the lower left. dashes and dots in many of the shadowed christ appearing to his mother ments: the sleeve and bodice are pen- Alte Pinakothek, Munich, inv. no. 652 Documentation and analyses areas, giving a sense of texture to the fab- in limbo (db no. 40.79) etrated by irr, but the cloak is not. The Not examined, but x-radiograph made 2003 at smb: Phase One visible and ir ric. Light scumbles were used to indicate Rijksmuseum Twenthe, Enschede, chalice of saint john the angels and the background figures were available. Dendrochronology by pk, 2005 as well as Inframetrics irr by cvd, illuminated areas on the robe, and the rays inv. no. 13 evangelist (db no. 12.141); painted over the reddish background, in Munich av; binocular microscopy by cm, rs; of the halos were added in an x-ray opaque Not examined, but infrared documenta- reverse of saint veronica with the latter serving as the shadow tone. x-radiography by at paint, probably lead-tin yellow. tion and x-radiograph provided by rte The left side of Christ’s robe was higher in saint veronica (db no. 12.22) with frame: Not original. 2005 at nga: cross section analysis by mg The x-radiograph revealed that the Documentation and analyses an earlier paint stage and was shifted to reverse: chalice of saint john window to the left of the Virgin originally preparation of support: White. Barbes unknown date at rte: Art Innovation make room for Michael’s hand and a partly the evangelist matched the one to the right, with a cross- on all four edges. No intermediate layer virgin and child (db no. 13.24) Musis 2007 used for infrared documenta- obscured face. The figure to Michael’s National Gallery of Art, Washington, apparent. shaped division through which sky and tion (spectral responsitivity extended to frame: right was added where the archangel’s inv. no. 1952.5.46 Engaged, original. Upper joins a continuation of the landscape could be 5 1550 nm through coupling with photo- wing had been painted out, and the angel Examined at nga: 19 – 23 May 2003 by cm, preparatory design: A free, loose, unre- are half-miter, half-overlap without nails seen. The coat of arms added later in the cathode tube) entering from the far left was painted over ap, rs; 31 March – 2 April 2005 by cm, cvd strained sketch, apparently executed in or pins; lower joins have vertical tenons present left window was revised during 2003 at rte: dendrochronology by pk the staff of Michael’s banner. two distinct stages, with varied line width held in place with one pin each (see Ver- a second paint stage using a lighter gray Documentation and analyses that skipped over irregularities in the sup- ougstraete-Marcq 1989, 151, for a diagram). blue, then the arches and the grid of lead- lines, which continue through the figure paint stage, but the height of the horizon frame: Integral, outside profile cropped. unknown date and operator at nga: port preparation. Contours were restruck. The hinges are original. In the top frame ing were incised into the still-wet paint. of the donor and lead to a single vanishing shifted and the stained-glass window with kneeling female donor with the x-radiograph support: Oak, 26.7 cm  18.8 cm. Single Broad hatching and zigzag strokes indi- member there are corrosion residues at the Under a microscope, a more intense blue, point in the facing panel. The drawing for Saint George was a late addition, probably redeemed of the old testament 1986 at nga: dendrochronology by pk plank with vertical grain. Possibly thinned, cate generalized zones of shadow. The left from at least three removed nails and possibly from the earlier sky, can be seen the figure is free and full of overlapping coinciding with changes in the window to (db no. 40.80) 2005 at nga: Phase One visible and ir as with a modern veneer applied to the underdrawing was apparently not meant a hole in the center that appears to have under the coat of arms. The convex mirror curves. The sitter’s proper right eye was the left of the Virgin. Museo Thyssen-Bornemisza, Madrid, well as Inframetrics irr by cvd; binocular reverse. to be followed exactly, and the painting been drilled. Two pairs of holes in the left also appears to have been a later revision, drawn slightly lower than it was painted. Other elements in the composition inv. no. 1930.76 microscopy by cm reflects a more tightly controlled tech- frame member were drilled from back to for the right edge was painted after the Memling apparently first laid out the were also revised during the painting pro- dendrochronology: Earliest possible Examined at mtb: 13 – 21 September 2004 frame: Not original. nique than the drawing. The painted front. The reverse of the frame is flat. red lake surface of the Virgin’s robe, and contours of the portrait, including a wider cess, possibly in relation to the donor panel felling date: 1427; earliest possible creation by cm, rs, cvd, av shadow of the chalice is smaller than the stained-glass window at the left side in the Triptych of Benedetto Portinari. The date: 1429; statistically more plausible support: Single plank, 52.5  41.5, with the left edge overlaps the background Documentation and analyses support: Oak, 31.2  24.2. Single plank drawn one, and the snake was shifted. (without shutters), but then developed a column at the right was originally painted production date: after 1445. vertically oriented grain. Dark overpaint landscape. Moreover, the reflection in 2004 at mtb: Phase One visible and ir as with vertical grain. Top edge is unevenly The triangular shape above the chalice complex perspective system to correct the on a wider circular base instead of an currently covers the original red and black the mirror indicates that Maarten van preparation of support: White (est.). well as Inframetrics irr by cvd, av; binoc- trimmed. marks the convergence of perspective lines rendering of the windows. It is not clear octagonal one, as in the Portinari por- marbling on the reverse (fig. 5). Nieuwenhove is kneeling beside the table ular microscopy by cm, rs; x-radiography for the niche, which were corrected more if the drawing was made with a dry or trait panel. The donor’s brown cloak was preparatory design: From the available dendrochronology: Earliest possible at which the Virgin is seated, but the situ- by RayXArt, Madrid felling date: 1467; earliest possible creation than once. dendrochronology: Not possible ation was different in the underdrawing liquid material. changed extensively from an earlier paint images, it appears that the underdraw- because of engaged frame. In the underdrawing two stained- stage, with the entire section under the ing describes all primary elements in the frame: Integral, outside profile cropped. date: 1469; statistically more plausible paint layers: The painting technique is of the portrait (see below). glass windows originally appeared to the foreground arm added, covering portions composition, possibly in a liquid medium, Non-original strip nailed on all four sides. creation date: after 1483. This panel came controlled and exact. Opaque mixtures preparation of support: White, barbes left of the donor (like those to the right of the carpet and the sitter’s torso that had and its free handling is comparable to that Regilded. from the same tree as that for Saint John of paint were cleanly applied without on all four sides. An intermediate layer was maarten van nieuwenhove of the Virgin), with roundels seen from an been completed as well as landscape and of the companion panel, but with diago- the Baptist. hesitation. The straight lines in the lower not observed in the x-radiograph, but the (db no. 13.25) support: Single plank, 26.9 / 26.6  18.9 / angle. A crossbar beneath these roundels architectural elements at the right. The nal hatching and restruck, looping con- left corner of the niche were incised in wet underdrawing seems to skip over a rough 18.7, with vertical grain. Possibly thinned preparation of support: White. No frame: Comparable to the frame for the was at the same height as the crossbar prayer book was painted over a reserve that tours. Figures in the foreground may have paint. The highlights were added as short, surface, which may signal the presence with a veneer applied to the reverse. The intermediate layer evident. Virgin and Child, except that the reverse is bearing the coat of arms in the companion had been left for a parapet, following the been redrawn in a darker liquid medium, precise strokes over the darker underlayers. of an unpigmented intermediate layer. veneer has been partly removed. beveled on the interior edge of the bot- panel. The room was drawn significantly underdrawing. The book was less sharply but not the bearded figure behind Eve or preparatory design: The underdrawing No reserve was left for the snake. has the appearance of being an accurate preparatory design: The underdraw- tom frame member. The joinery is similar, lower, with ceiling beams and the junction foreshortened a later paint stage. background figures such as David and the dendrochronology: Not possible. ing appears to have been made with and the top frame member has a straight- of the ceiling with the wall visible above people surrounding him: the face of Eve copy of an earlier composition. It was preparation of support: White (est.). executed in a material that makes a uni- Summary both liquid and dry materials. It consists sided hole in the center and corrosion the donor’s head. In the left foreground is drawn over the legs of David, and her Summary Striations from an intermediate layer can form thin line, with occasional skipping The figure of Saint Veronica was prob- of contour lines with intermittent hatch- residues from two removed nails as well as a stone parapet was underdrawn, as in features as well as the left contour of her In several details this diptych for Van be seen. and grainy areas (as in the hands). Shad- ably copied from an existing composi- ing used to indicate shading. Incised lines remnants of a modern screw to the right the right wing of Memling’s Triptych face seem to have been completed in the Nieuwenhove was initially closely related ing was established with hatching and tion, while the chalice might have been in the architecture, which do not breach of center. Two pairs of holes in the right of Benedetto Portinari. The tip of the Virgin’s darker medium. preparatory design: The composition to the central panel and right wing of the crosshatching. The background landscape designed on the panel. The painting the figural elements, were probably made side roughly correspond to the holes in the red cloak in the lower left corner was not The wound in Christ’s hand does not is fully underdrawn, possibly in a liquid Triptych of Benedetto Portinari. Memling might is minimally indicated, with lines for the technique on both front and back is com- after the drawing was complete. left side of the Virgin and Child, also drilled planned in the underdrawing. The book appear in the underdrawing, while the material, and the manner is free and have used the same pattern drawings for hills and rough approximations for the parable. The x-radiograph of the Munich The irr and x-radiograph show that through the thickness of the frame. was apparently also not part of the origi- nipple and wound in his side were painted assured, though more controlled for the both works, but the large number and placement of the trees. Veronica’s blue panel shows a similar use of incisions in the Virgin’s fingers holding the apple were nal composition. A change in the donor’s in slightly different positions than they donor. The underdrawing is comparable support: Single plank, 52.3 / 52.5  41.6, nature of the late changes here imply the mantle extends farther to the left and the wet paint for the nimbuses and the repeatedly repositioned, both in under- thumb, which was drawn smaller than were drawn. The angel above Christ’s head in style, method, and conception to that in with vertical grain. Unpainted margins are donor’s close involvement in the work’s closer to the foreground in the underdraw- architecture of the reverse. drawing and in paint. Her ring finger was it was painted, also reflects the Portinari was more upright in the drawing than in the companion panel. Foreground figures recorded in the x-radiograph. The reverse final appearance. Van Nieuwenhove’s coat ing than in the painting. Her right hand drawn closer to the apple without touch- portrait. The inner edge of the fur-lined the painting. Saint Michael the Archangel were redrawn in a darker liquid medium, currently has dark overpaint obscuring of arms may have intended for the exterior was drawn in a cursory manner and not ing it, then painted touching the apple, outer garment at the donor’s proper right initially had a large wing to his proper left, as apparently in the facing panel, but there the original red and black marbling. of the diptych. It was suggested in Veroug- followed in the paint. The painted folds in and finally repainted farther back from was underdrawn at more of an angle than which was underdrawn and partly painted is little hatching. Broad, curved, widely straete-Marcq and Van Schoute 1989, 151, the mantle likewise bear little relation to the apple than it was underdrawn. Her preparation of support: As in the Virgin it was painted. but later blocked out. His banner was spaced, skipping lines mark the position little finger was drawn farther away from and Child, a slight gray striation is visible in that this might have been a stationary dip- drawn with an opaque horizontal band of cast shadows on the wall at the right. the drawing, and the painted landscape paint layers: The paint handling is com- the apple than painted. The Christ child’s places in the irr (i.e., in Maarten’s hands), tych, with the portrait on the moving wing. slightly higher than the current red band Similar lines appear above the wainscoting. adheres to the drawing only generally. parable to that in the Virgin and Child. The fingers were underdrawn closer to the and the underdrawing skips periodically, The donor panel’s elaborate perspective The floor tiles were underdrawn. imitation wood grain behind the donor’s paint layers: The paint was applied con- apple than painted. suggesting the presence of a striated inter- system, the bevel of the lower member on The donor and prie-dieu were head was created by incising into wet scientiously. Primarily opaque mixtures mediate later. the reverse of its frame, and the presence planned in underdrawing, but not until paint layers: The paint is thin and was brown paint, revealing a lighter brown were layered and blended to form smooth of marbling on both reverses do suggest after the overall composition was laid transitions in the figure. The foliage was applied with skill and care. The flesh tones preparatory design: The underdrawing layer beneath. The torso of the sitter was different handling for each wing, but the painted with small strokes of light green were built up with delicate liquid strokes is unusually elaborate and was prob- depicted without notable changes in the frame construction indicates a standing over a darker green base, suggesting ably created in separate campaigns. The diptych, not a hanging one. individual leaves and veining in the fore- perspective system of the architecture was established with a series of straight

fig. 5 Micrograph from the reverse of Virgin and Child, showing 1) the white ground layer with 2) original reddish marbling beneath 3) dark overpaint

292 293 out. This is illustrated by the elbow of the book was smaller and thinner, like a prayer Cat. 29. Netherlandish Artist dendrochronology: Earliest possible Cat. 30. Netherlandish Artists Cat. 31. Jan Provoost middleground figure in the donor’s nose book, but was enlarged in a second paint felling date: 1491; earliest possible creation and cheek and the underdrawn gown stage and her fingers extended over the date: 1493; statistically more plausible lentulus letter and virgin and child and two donors christ carrying the cross, of the redeemed figure in the center of the edge. The donor’s hands were first painted production date: after 1499. The planks portr ait of christ (db nos. 26.51 – 52) 1522 (db no. 41.81), with reverse: prie-dieu. with bent fingers, then painted longer for both panels are from the same tree. (db nos. 106.120 – 121) Koninklijk Museum voor Schone Kunsten, imitation porphyry; The donor’s portrait may have been and over the earlier book, then the book Museum Catharijneconvent, Utrecht, preparation of supports: The front Antwerp, inv. nos. 517 – 518 and portrait of a fifty-four- transferred from another source (see entry), was repainted to clarify the contour of inv. no. bmr 52 of both panels has a thin whitish ground Examined at kmska: 11 – 22 November year-old franciscan (db no. 41.82) as this drawing is less spontaneous than her fingers. Examined at sral: 15 – 27 September 2003 layer. Each work has barbes present on all 2003 by cm, rs, cvd, av with reverse: skull in a niche that seen elsewhere in the panel. The dark The figure in the upper group at the by cm, rs, cvd, av four sides. No intermediate layer is vis- Musea Brugge, Hospitaalmuseum Sint- contours defining the donor’s hat, visible far left may have had his head turned in Documentation and analyses ible in microscope examination or on the Janshospitaal, Bruges, inv. no. osj 191.1 in ir and irr, are a painted black outline, the opposite direction in an earlier paint Documentation and analyses 2003 at kmska: Phase One visible and ir x-radiograph. Examined at smb: 29 September 2003 by not drawing. Changes in the donor’s stage. The foot and leg of the Moses figure 2003 at sral: Phase One visible light as well as Inframetrics irr by cvd, av; cm, rs, cvd, av attire include her head veil, which was were shifted slightly to the left in a sec- and ir as well as Inframetrics irr by preparatory designs: For the Lentulus binocular microscopy by cm, rs underdrawn shorter, falling just below ond paint stage, after the staff had been cvd, av; binocular microscopyby cm, rs; Letter no drawing was observed with ir, 2004 at kmska: dendrochronology by pk Documentation and analyses her jawline, and continued at this length painted, and the cloth on the prie-dieu dendrochronology by pk irr, or in visible light. For the Portrait 2003 at smb: Phase One visible and ir frames: Integral. Partly original, but in an earlier paint stage (forms to the left overlapped this leg. The female profile face 2004 at nga: cross section analysis by mg of Christ the contours of the composition as well as Inframetrics irr by cvd, av; altered. The two panels, which now and right of her neck are x-ray opaque). in the back and the faces behind Moses, were underdrawn in a liquid material. binocular microscopy by cm, rs; frames: The applied frames are original share a single central frame member, no A wide necklace was drawn over her both of which enhance the visual con- Deviations between the underdrawing x-radiography by at and attached to the edges of the supports longer fold. This narrow central strip chest and shoulders, and the bottom nection between the two panels, were not and the paint surface include the side with hand-forged nails, hammered from was inserted where hinges would have frames: Original engaged frames. The border of her bodice was lower. The added until a second paint stage. of Christ’s mouth, shifted to the left, and the reverse. Nearly all of the original been, obliterating evidence of the origi- 6 corners are mortise-and-tenon joined x-radiograph revealed that the garment The irr assembly reveals that a sky the mustache, which was slightly raised. ground and gold leaf has been stripped nal hinges. The insert, measured on the with horizontal tenons. Each joint is was left in reserve from the rest of her or possibly a landscape was originally from the frames, but traces of both remain paint layers: The Lentulus Letter was x-radiograph, is 16 mm wide at the top, secured with a pin. The gold leaf has been attire. An underdrawn line indicates the painted above the wainscoting at the top in the corner joins. Wear patterns in the painted black, with the text in a yellow- 19 mm at the bottom. The reverse of the repainted, but the rebus on each is origi- end of her nose, which may have been right. Shapes suggestive of clouds and The underdrawing for Two Donors is top center of the upper frame member ish, x-ray-opaque mordant on which gold frame’s outer molding is cut into a rebated background paint lies directly on the paint nal, as is that on the reverse of the donor shifted higher in paint, and another marks hills, now painted over with deep blue, more schematic and less complex than that reflect long-term use of the hanging leaf was applied. A red glaze might have lip, which fits into a surrounding non- layers below, without intermediate grime. panel. The hinges and the closing device her mouth. appear to have been retained in an earlier for the Virgin and Child. It consists largely device. been applied to the background. original entablature. The inner lip of the irr revealed that the female donor’s are also original. Partly filled holes from The figure at the left holding a paint stage. The floor tiles at the lower left of liquid contour lines, with little elabora- head covering was changed from the The present hinges are modern Paint on the Portrait of Christ was frame for the Virgin and Child is slightly nar- earlier hanging devices are visible in the staff was drawn with horns, implying that were changed from the underdrawing and tion of the drapery. Multiple contours and attached with screws. The residues applied in a relatively direct and simple rower than that for Two Donors. The paint first paint stage, when it was larger and x-radiograph. The reverses of the engaged he was intended as Moses, and sketchy again in paint; ruled diagonal and hori- were used to position the female head of corrosion from old nails show the manner. The background was laid in first, on the lower inner bevel is cracked and deeper. Her attire was also changed, from frames are carved with a simple beveled lines near his knee may have represented zontal lines at every third join in the floor covering, and no hatching was observed. original location of the hinges, above then the flesh, followed by facial details appears original. The frames are regilded. a rounded neckline to the present white- recess, but original paint on the flat sur- the Tables of the Law. His staff was moved tiles at the lower right mark the location There were slight changes in the male and below the present upper and lower and hair. Finally, the contour of the profile edged black collar, and the lower edge faces imitates a recessed frame. from his hand to his elbow. Underdrawn of larger, lozenge-shaped floor tiles. supports: Oak, single plank. Both panels donor’s fingers and in the left contour hinges. Similar traces of a nail on the left was expanded from that left in reserve — at of the black collar was first rounded rather lines near the leftmost cupid at the top are 35  23.4. The reverses are coated with of his face. The female donor was under- supports: Single planks with vertical frame member of the Lentulus Letter and the the lip, nose, and forehead. There is a large than square. The upper contour of the may have established the position of his lead white, obstructing x-radiography. drawn with a round collar. grain. Christ Carrying the Cross is 49.9 / shaft of a nail in the right member of the area of damage in the lower left corner male donor’s thumb was enlarged in paint knees, while lines in the lower left corner Summary over his costume, which had been blocked 50.1  39.5 / 40.3; the donor panel is Portrait of Christ indicate that the diptych of the Portrait of Christ. dendrochronology: For the Virgin and paint layers: The Virgin and Child was suggest an earlier design for the floor tiles. These panels were obviously conceived as 50.05 / 49.9  39.8 / 40.3. Unpainted had a closing device, which is now lost. Child: earliest possible felling date: 1460; executed from light to dark, with reserves in without great care for delimiting the a pair. Traces of hinges were not observed, fingers. His shoulders were raised and margins are visible in the x-radiograph. paint layers: The painting technique in Hand-forged nails close the mitered joins earliest possible creation date: 1462; sta- left for the figures. Following the painting perhaps because the outer profiles of the widened, and his hair was extended to the figural group is very free. The floating frames were cropped. The similar treat- at the top. tistically more plausible production date of the composition, the fingers and faces dendrochronology: Not attempted figures were executed with fluid, confi- (assuming ten years for seasoning): after were outlined in reddish brown, and the the right. because of the engaged frames. ment of reverses and frames imply an supports: Oak, single planks: the Lentulus dent strokes over the dark background, 1477. For Two Donors: earliest possible felling final contours were delicately brushed on intervention before the panels were Letter is 36.4 / 37  26.7 / 26.5; the Portrait preparation of supports: White (est.). which serves as a shadow tone. The gar- date: 1491; earliest possible creation date: with a lighter paint. Cross section analysis separated. Numerous changes during of Christ is 36.9 / 37.1  26.7 / 26.5. The Summary No intermediate layer was observed. Each ment and red hat of the foreground figure 1493; statistically more plausible produc- shows that the blue background lies over the underdrawing stage and between reverses have rounded edges and traces These two panels, although they form a panel has a barbe along all four edges. were left in reserve, as were the donor’s tion date (assuming two years for season- black particles typical of surface dirt and underdrawing and the final paint stage of red paint applied directly on the wood, diptych, do not seem to have been pro- book and Moses’s cloak. Final touches ing): after 1498. crosses a crack that extends through all preparatory designs: An elaborate draw- make clear that these compositions were as confirmed by cross section analysis. duced simultaneously but paired at a later of gold leaf were applied over the donor’s lower layers. The blue background was ing for Christ Carrying the Cross defined all not copied from an existing prototype preparation of supports: White (est.). date. They have different dendrochro- necklace and embroidered neckline. A possibly painted at a later date, perhaps forms with both straight and zigzag lines. but were worked out on the panels. The No intermediate layer observed. nological dates, different underdrawing at the time the donor panel was attached, There are contours as well as hatching that fingerprint was observed at the tip of the drawings were done with confidence and techniques, and different painting tech- preparatory designs: The drawing for with the intention of visually uniting the suggests volume. Shading was described presumed Moses’s staff. accuracy in multiple stages, though the niques. The inner profiles of the integral the Virgin and Child, in both dry and liquid two images. with parallel hatching as well as some The donor was painted more precisely, foreground figures were developed more frames are slightly different. The Virgin’s media, is spontaneous, free, and angular, The gaze of the Christ child was at crosshatching in the deepest shadows. It but she seems to have been part of the completely than those in the background, background is more opaque in irr and with zones of parallel hatching. Many some point shifted upward (fig. 6). The is difficult to be certain about the under- original plan, as her figure does not over- and several airborne figures may not have more transparent in x-radiography than of the folds were struck with double lines, change in the placement of the pomegran- drawing medium; in some places it was lap other elements. Initially she was drawn been drawn at all. An area of uncertain that behind the donors, and cross sections some with a hooked end. Zigzags indicate ate was accomplished in successive paint clearly liquid but in other places the very wearing a white veil without a hat, but the size was left for the donor in the initial show the final layer of blue on that panel shading in the draperies. The faces of the stages. The Christ child’s head was left in fine line could have been dry. The under- latter was added in paint. She wore a wide, layout, which was apparently filled later lies over dirt and continues into old losses. Virgin and Child were each defined with reserve following the larger underdrawn drawing style is consistent throughout the rounded necklace, which was replaced at with a portrait taken from another source. The change in the direction of the Child’s a single broad contour, but the Child’s image, then painted smaller. His hand diptych and is fully comparable with that paint stage with a sheer v-necked bodice; Both underdrawing and painting display gaze, from outward to looking up at the elbow was drawn repeatedly. The Virgin’s was painted over the Virgin’s little finger, of other paintings by Provoost (see Spronk the heavier brocaded garment was drawn similar skill. Late changes and additions, Virgin, and the position of the Christ drapery was also altered, especially at the and his chest was painted farther to the in exh. cat. Bruges 1998, 1:184; 2:31 – 48). with deeper décolleté than painted. Her especially in small figures along the edges, child’s right hand, normally the blessing right. Her proper right eye was painted right, closer to his hand and to the Virgin’s seem calculated to strengthen the visual hand but here grasping the Virgin’s finger, higher than it was drawn, her breast breast, but was narrowed and shaded in relationship between the two panels. Some both might indicate that this painting at made larger, and the fingers on the hand the final painting. changes in the donor’s attire suggest an some point functioned as an autonomous holding Christ were shortened (the draw- Two Donors was painted more tightly interest in modesty and decorum. panel rather than as part of a diptych. The ing showed them almost curved around (paint conformed more closely to the ensemble probably represents a phased- the Child’s belly). The Child’s hand and outline), but the final touches that refine production pairing of uncertain date. The the pomegranate were drawn on his for- the flesh tones of the Virgin and Child are interventions to the frame suggest that ward knee but painted on his back knee, not seen here. Reserves were left for the a much later assembly — perhaps in the while his elbow and forward leg were heads and faces of both donors, although nineteenth century — cannot be excluded painted lower. that allowing for the hair of the male at this time. donor was not a specific shape. The blue

fig. 6 Micrograph of the proper left eye of the Christ child, showing the repainted iris

294 295 The thorns in Christ’s crown were had been an opening between the index Cat. 32. Bernard de Rijckere Adriaan’s proper left hand was carefully preparation of support: White (est.). The drawn larger than they were painted, addi- and middle fingers in the drawing that skull in a niche (db no. 41.125); outlined with a thin dry contour line, lack of an unpainted margin is evident in tional curls of hair were drawn at Christ’s revealed the back of the other hand, and reverse of portrait of a fifty-four- and there is a hint of a similar line on the the x-radiograph. Evidence of an interme- adriaan van santvoort and temple and ear, and the inner ear shell when the fingers were painted longer and year-old franciscan inside of his right thumb. Underpaint, as diate layer is present in ir and irr. his sons guillaume and adriaan, was not defined in the drawing. Several wider, they obscured the back of that hand. in the eldest boy’s right hand, can easily be preparation of support: White (est.). 1563, and anna van hertsbeeke preparatory design: The main con- folds in Christ’s robe were changed in the The collar of Christ’s lavender robe was mistaken for underdrawing. No intermediate layer apparent. and her daughter catharina tours of the composition were cursorily paint stage, and the rope was painted in a first painted following the underdrawn and son jan baptiste paint layers: The painting technique is sketched on the intermediate layer with a different position. Auxiliary figures were diagonal fold between the button and preparatory design: The underdraw- with reverse: coats of arms of assured and efficient, and the relatively soft dry material, such as a soft black chalk. also altered: the head of Saint John at the the sleeve at the right, and then changed ing is consistent with that on the fronts the van santvoort and van fluid paint was thinly applied in several The underdrawing is assured but not pre- upper left was drawn farther right than it to the present configuration in a second of the diptych panels. The right contour hertsbeeke families in a niche layers. The flesh tones were blocked in cise, and it might be a freehand transcrip- was painted and might have been depicted paint stage. The cross was painted with a of the skull was redrawn to the present Private Collection using a tan color, and areas of shading tion of a model at hand. The underdrawn more in profile, while the top of Mary’s wider plank, covering part of Mary’s face, configuration, and it appears the shading Examined at sral: 15 – 27 September 2003 were added in a dark underpaint that date at the bottom was more compact than head (just below) was drawn higher than and the new contour was incised into the in the forehead was redrawn to fit the new by cm, rs, cvd, av shows through the lighter-colored top lay- the painted date. it was painted, with a veil that crossed lower paint layers. Mary’s head covering, contour. This would imply that a later ers, as in the nose and corner of the mouth her forehead in a straight line but was which was underdrawn with a different drawing campaign was executed over an Documentation and analyses paint layers: The paint application was of the older boy. irr revealed that the gar- painted with a peaked fold. The figure contour, appears to have been painted in underpainting of the skull. The original 2003 at sral: Phase One visible light more summary than for the portraits, with ments were underpainted to leave reserves with the bulging eyes in the upper center different positions as well. A more signifi- underdrawing depicted a loose stack and ir as well as Inframetrics irr by less modeling and blending of the paint. for the hands and fingers. The reserves was not prepared in underdrawing; and cant change was the shortening of Christ’s of books supported on a pillow under the cvd, av; binocular microscopy by cm, rs; Pictorial elements are flatter (i.e., the inci- were smaller than the final painted forms the soldier in the top right had curls in right sleeve in a second paint stage, which skull. The pedestal was redrawn over an x-radiography by at; dendrochronology sion and restruck delineation of the v in throughout: as seen in the apple held by “ ” his hair, a larger mustache, and a different may have related to the changed position underpaint layer. A tooth halfway back by pk the banderole). The forms of the numerals the younger boy, the puff of fabric on helmet decoration. The face in the center of the rope: in the underdrawing the rope behind the jaw in the underdrawing was here differ from those used for the ages Adriaan’s proper left shoulder, and the to the right of Christ was painted slightly bound Christ’s sleeved rather than bare not painted. and dates in the portraits, especially the 6 adriaan van santvoort and his hands of the older boy. The left hand “ ” wider, with the nose made thinner and the arms. This probably has iconographical and “3.” Because the underdrawn numer- paint layers: The paint was handled in a sons guillaume and adriaan of the youngest was underdrawn without 7 mouth larger, while the large mustache meaning because of the rebus on the als resemble those of the painted date, it similar manner to the other images in the (db no. 51.97) the thumb. was indicated by only single lines in the frame (see entry). is possible that a different hand painted diptych. Lead white was used sparingly so The paint was masterfully manipu- underdrawing. The paint handling for the donor frame: Original rabbeted box frame. the inscriptions here, following the under- that it only registers in the x-radiograph lated, with traces of the brush having in a dry medium, for it appears to skip over The underdrawing for the donor panel is comparable, but a significantly Shallow indentations where now-missing drawing of the master. at the brightest highlights. A reserve was been dragged through wet paint to create anna van hertsbeeke and her the intermediate layer. The underdraw- panel is fully comparable in execution and larger number of changes were observed hinges were originally attached are filled left for the pillow under the stack of books. feathered ends in several details. The color daughter catharina and son ing at the lower right is surprisingly free, style. Some of the contours are double here between subsequent paint stages. with thin blocks of wood, and x-radiog- The lower right book was brought to a of the intermediate layer was efficiently jan baptiste (db no. 51.98) with zigzags and long diagonals roughly lines, such as the nostril. The thickly X-radiography revealed that the donor was raphy shows the remains of six nails for Summary beginning paint stage but not completed. used as a middle tone, as can be seen in the indicating zones of shadow. The drawing painted green background, which covers initially painted within a fully finished the top hinge and five for the bottom. A frame: Although this box frame is also The panels were planned and executed as left side of Adriaan’s face through the thin is intended to give a general indication much of the underdrawing for the left interior, with windows to his right, a wooden hanging device is possibly origi- original, it differs from that on the com- a diptych by a highly skilled painter. The surface paint. The gray brocade pattern on rather than precisely locate forms, and it eye and the outer contour of the face, is hearth to his left, and a kettle over a fire. nal but may date from the seventeenth panion panel in that a second face was construction of the panels and frames, Summary the garments was completed with a paint is largely followed in the paint. not penetrated by irr. The donor was These elements, some of which can also be century. The x-radiograph and wear on attached to the reverse with nails and pins. the sparse underdrawing, and the style This diptych, with its original hinges that has more body than the surrounding initially depicted within a fully furnished recognized in the irr assembly, were later the frame surface suggest that there was See above for a description of the hinges, paint layers: Comparable to the Adriaan and method of painting are typical of the and original closing device, is in remark- black paint and is visible in ir but is x-ray room, but the green background makes painted out with a thick layer of green once a closing device. X-radiography also hanging and closing devices, and original van Santvoort in the wet-in-wet painting later sixteenth century. The right wing ably good condition. A large number transparent. Despite the evident assurance it difficult to establish if details of the background paint. The donor’s head was shows foliate scrollwork that once adorned decoration. and the rendering of the gray brocade. The was meant to be opened and closed over of changes occurred late in production, with which the portraits were painted, interior (such as windows, kettle, and originally painted farther to the left, fol- the face and roundels at the centers and buildup of thin color layers is especially the stationary left wing, as confirmed by suggesting possible intervention by the some small changes were found with irr support: Oak, 84.6 / 85  66 / 65.5. Two fireplace) were underdrawn as well as lowing its underdrawn position, but the corners of each frame member, now visible in Anna’s temples near the hairline. the painted reverse of the right panel. The donor. Several changes seem to have had and x-radiography. The older son origi- boards. The reverse is painted with the painted. Among the deviations between face was moved to the right in a second painted black. The painting is now held Also like the companion panel, the reserves originally hinged pictures would have iconographical significance. Revisions nally looked to the left, and his lower right family coats of arms. Earlier reports that underdrawing and painting, the head paint stage, and, perhaps in a third stage, in place with a wooden batten attached were often smaller than the painted forms, been displayed either fully opened or fully in the Christ panel, including the addi- jaw was enlarged with the final light flesh the coats of arms were painted on a sepa- of the donor was drawn farther to the left, the head was enlarged at the crown over with screws to the reverse, but originally as seen in Anna’s fingers, while the chair closed. Probably in the seventeenth cen- tional figure at the top and enlargement tones. A rectangle behind Adriaan’s head rate panel could not be confirmed (see his proper left hand was drawn at a dif- the green background. The neckline was it was secured with nails along the panel’s finials were egg-shaped in the reserves but tury the frames were divested of hinges of the cross and Christ’s hands, increase is relatively opaque in irr and x-rays, but entry, note 2) ferent angle, and the fingers were drawn painted lower than drawn, and the cowl perimeter. painted round atop a square base. Anna’s and ornamentation, and the folding the sense of crowding and of Christ’s faint indications of folds are discernible, longer. What looks like a looped rope was initially painted larger and slightly dendrochronology (measured from x- eyes, like those of her oldest son, initially diptych was altered into two pendant suffering. Major changes in the donor support: Oak, 84.3 / 84.1  65.1 / 65.4. Two implying that he was initially depicted appears behind the donor’s thumbs in the lower. The hands were made smaller with radiographs): Earliest possible felling date: looked to the left (fig. 7). The daughter’s paintings. It is possible that the wooden panel — painting out the original interior boards: the left is 32.7 cm wide at the top against a backdrop of fabric. A similar underdrawing, perhaps representing a belt shorter fingers. In a final stage of painting 1544; earliest possible creation date: 1546; fingers were defined with underpaint that hanging devices were added then as well, setting, enlarging the man’s head, and and 32 cm at the bottom; the right is 32.6 rectangle can be seen behind Anna’s head for his habit, but echoing the rope around a thin line of gray was added over the flesh statistically more plausible production looks like underdrawing, but they were but this is uncertain. moving him toward the center of the cm to 33.2 cm. An old split on the right was with irr but not in x-radiography. These Christ’s wrists and the rebus at the top tone of the hands to efficiently indicate date: after 1552. Planks are from the same originally extended, not holding the ties panel — appear to emphasize the donor’s repaired with fabric and lead-white adhe- areas are no longer visible with the naked of the Christ panel; this rope might have that they are separated. tree as the companion panel. encircling the mother’s waist. The painted devotion. sive. The panel retains its original dimen- eye. The finials on the chair back were continued down to the lower left corner reverse complicates the reading of the sions. The verso has a 1 – 2 cm wide bevel painted round, though the reserves were preparation of support: White (est.). in the drawing, further connecting the x-radiograph, which does not show the imitation porphyry (db no. 41.124); along the lateral edges and wider curved egg-shaped, and they were originally X-radiography shows that the paint and wings of the diptych. rectangle behind Anna’s head that is seen reverse of christ carrying the bevels at the top and bottom that measure painted without a square upright. The left ground continue to the edge of the panel, in irr. It is possible that this form was paint layers: The paint for the Christ cross 2 cm in the corners but increase to 5 cm in arm of the chair was not painted in the with no unpainted margin. Over the not as fully worked as the area behind her panel was applied economically in opaque the center. first stage. The final paint layer was used ground a yellow-brown intermediate layer frame: Flat face with plain beveled recess. husband in the companion panel. layers. The artist employed wet-in-wet Painted with gray marbling. dendrochronology: Earliest possible to refine other shapes as well. The corner was broadly brushed that is visible in both blending in describing larger volumes, felling date: 1544; earliest possible creation of Adriaan’s sleeve trailer at the lower irr and x-radiography. adding detail with small, distinct opaque preparation of support: White (est.). left (under the older son’s proper right coats of arms of the van sant- date: 1546; statistically more plausible pro- preparatory design: Unlike the compan- brush strokes. Christ’s fingers and the No intermediate layer apparent. elbow) was defined with paint that is x-ray voort and van hertsbeeke duction date: after 1552. Planks are from ion panel, underdrawing here defines the back of his hand were painted larger than opaque but later covered with background families in a niche (db no. 51.98); preparatory design: None visible with the same tree as the companion panel. fur, chair, and shadows in the lower right they were drawn, emphasizing the picto- color, which is opaque in ir. The right side reverse of anna van hertsbeeke irr. corner, but there are no careful contours rial link with the donor panel. But there preparation of support: White (est.). of the sleeve ruffle on the older child’s left that describe the hands. Like the Adriaan frame: The frame has a flat profile with paint layers: Wet-in-wet and splattered There is no unpainted margin. A yel- arm was overpainted with black, and his van Santvoort, though, there is no under- a stepped inner rim. It does not seem to red and black paint. low brown intermediate layer is broadly purse was shortened. Adriaan’s beard was applied. The intermediate layer is visible drawing in the faces. The drawing may be have had the roundels and foliate decora- painted freely and larger at the left, with tion observed on the front. in both irr and x-radiography. the white collar added over it. preparatory design: Very little under- drawing was revealed in this panel.

fig. 7 Micrograph of the proper left eye of Anna van Hertsbeeke, showing the earlier position of the iris

296 297 Cat. 33. Jan van Scorel Scorel made several adjustments to Cat. 34. Michel Sittow of the headcovering and chemise. The paint layers: The paint was handled with Cat. 35. Jan Cornelisz Vermeyen examination of the paint surface with the portrait from the drawing to the paint lucretia (db no. 44.110); reverse of brush stroke and consistency of the paint skill and confidence. Over a smoothly binocular microscope confirm that the stages. The donor’s forehead was moved portrait of a man are manipulated to distinguish various blended base that expressed dimensional green drapery was executed with identical virgin and child and virgin and child and cardinal érard de la marck forward but his chin kept in the same posi- textures, from the bird’s feathers to the contours, small touches of feathery brush- paint mixtures and layer structure as on portrait of a man preparatory design: The underdrawing diego de guevara (?) and the holy family tion so that he appears to incline his head Child’s flesh and the wooly surface of the work define subtle detail. The fur collar the companion panel. is loose and free, with contours defined Examined at sral: 15 – 27 September 2003 more toward the other panel, and his neck carpet. The changes in the Child’s pose was painted wet-in-wet, with incisions Various changes appear to have been and sketched lines visible through painted by cm, rs, cvd, av virgin and child was widened in the first paint stage. The virgin and child and gaze were made in successive paint through the wet paint that uncovered a made during the painting process, but our skin tones. Dark squiggles at Lucretia’s Kartinaja Galeria, Tambov, inv. no. 13 proper left elbow was underdrawn larger (db no. 38.75) stages, without intermediate drawing. The tan underlayer, which then functioned as a Documentation and analyses observations were made before the over- knuckles echo the underdrawn treatment Not examined. than it was painted. The book at the lower Staatliche Museen zu Berlin, Virgin’s eyes appear to have been painted midtone. The artist changed several ele- 2003 at sral: Phase One visible and ir as paint from two restorations was removed of the donor’s knuckles and may have been left was not included in the underdrawing, Gemäldegalerie, inv. no. 1722 twice, judging by the x-radiograph. The ments here: the sitter’s eyes were painted well as Inframetrics irr by cvd, av; binoc- in 2006. The following findings should be meant to establish shading. Another zig- portrait of a man (db no. 44.88) and when it was added in paint, the fin- Examined at smbg: 22 July-1 August 2003 change in the placement of her ear was in at least two earlier positions; his proper ular microscopy by cm, rs; x-radiography confirmed after treatment. The reserve zag appears on her abdomen. with reverse: lucretia gers on the sitter’s left hand were changed by cm, cs, rs made in the first paint stage. right cheek was narrowed after the first by at; cross sections by cm; dendro­ for the angel at the upper left had a more Staatliche Museen zu Berlin, to spread across the book more convinc- paint layers: The painting technique is Documentation and analyses paint stage; while the top of his brocade chronology by pk horizontal contour at the top of the wings, Gemäldegalerie, inv. no. 644b ingly. The landscape, which was more fully efficient. Wet-in-wet scumbles drag paint 2003 at smbg: Phase One visible light diego de guevara(?) (db no. 38.76) doublet was originally lower, and the 2005 at nga: cross section analysis by mg whereas they were painted upright. Several Examined at smbg: 22 July-1 August 2003 drawn than the portrait, was also altered across color boundaries, with the curtain and ir as well as Hamamatsu irr by National Gallery of Art, Washington, white shirt was wider and had a lower folds in the sitter’s proper right sleeve by cm, cs, rs in several respects. A tree underdrawn at modeled by red lake scumbled over white cs; binocular microscopy by cm, rs; inv. no. 1937.1.46 neckline. Late additions include the fur cardinal érard de la marck were repainted in a different configura- the left was not painted, and the arch in and vice versa. With the exception of the collar, which was painted over the white tion. The fringed edge of the curtain was Documentation and analyses x-radiography by gs Examined at nga: 19 – 23 May 2003 by cm, (db no. 45.89) the rocks at right was drawn lower, then figure of Lucretia, the paint is not as x-ray shirt; and the cross of the Calatrava, added over highlights from the first paint 2003 at smbg: Phase One visible and ap, rs; 30 May-1 April 2005 cm, cvd Bonnefantenmuseum, Maastricht, on loan painted enclosing two tones of sky. opaque on this side as on the donor’s. frame: Not original. which was painted over the brocade stage that depicted a clean edge. Also, the ir as well as Hamamatsu irr by cs; from Rijksmuseum, Amsterdam, inv. no. Documentation and analyses and was originally gilded with red lake little finger and ring finger on the proper x-radiography by gs; binocular microscopy paint layers: The paint application was support: Oak, 33.1  25.5 / 25.7. A single a4069 1983 at nga: dendrochronology by pk embellishment. left hand were more curled in the reserve, by cm, rs free, efficient, thin, and exact. The paint piece of wood with a crack through the Summary 1988 at nga: x-radiography by Kristin frame: Not original. then extended in the final paint, while the 2004 at smbg: dendrochronology by pk was put down quickly in a single layer entire height. The original panel, with ver- This diptych shows the working methods Casaletto thumb and little finger were lengthened except in the sky, where a light-colored tical grain, was inserted into an auxiliary support: Oak, 62.8  54.35 (not including frame: Not original. of a single master using various models. 2005 at nga: Phase One visible and ir as Summary and the index finger on the right hand was initial layer provided a foundation (seen panel and cradled; the present unpainted added strips). Two planks, with vertical Although the language of all three under- well as Inframetrics irr by cvd; binocular The presence of original barbes at the top made larger. The most x-ray-opaque areas support: Oak, 66.1  45.4. Two boards. The in micrographs of the barbe). Many minor margins are part of the auxiliary panel. grain, with strips added on all edges. The drawings is the same, the Virgin and Child microscopy by cm and bottom of both panels proves that the of the fingers are in the top layer (the areas arch is slightly flattened at the top. adjustments were made in a second layer left plank is 25.3 cm wide at the top and 24 (which was not examined for this project) dendrochronology: Not possible heights of the two panels are within a half- of change). By contrast, it was the first of sky paint. The cloud behind the donor’s frame: Not original. cm at the bottom (without the added strip). dendrochronology: Earliest possible seems to be more fully worked out. The because of altered support. centimeter of each other. The paintings painting of the cardinal’s hat that used head was added only after completion Thinned, backed with a cradle. felling date: 1498; earliest possible creation quality of the painting is reportedly very support: Oak, 33.6  23.7. Single plank, are by the same artist, and the continu- denser pigment, and the hat was altered of the hair and a small figure at the right preparation of support: White. There date: 1500; statistically more plausible high, which does not suggest any major with vertical grain. Substantial non- ation of the carpet across the two panels dendrochronology: Earliest possible using a paint more transparent to x-rays. (apparently damaged by overcleaning are remnants of a barbe at all four edges, production date: after 1506. Both boards workshop involvement. The underdrawing original additions on both lateral edges. links them visually. The reworking of the felling date: 1511; earliest possible creation This change may represent later overpaint, and now largely disappeared). The cloud proof that the painted surface has not for this painting are from the same tree. for Lucretia is looser and more summary, The additions are not rectangular. The Christ child results not only in a greater date: 1513; statistically more plausible for the initial form of the hat (as seen in delimits the curls of hair, which were then been cut down. Technical examination while that for the portrait is less elabo- piece added at the left measures 4.4 cm at connection with the donor but with the production date: after 1519. the x-radiograph, which shows more fore- preparation of support: White (est.), brought back over the cloud. Reserves in revealed a thin, lightly pigmented inter- rated. Given the thin, efficient painting its widest and 0.6 cm at its narrowest. That viewer as well. Sittow often shifts the posi- head) resembles one Vermeyen depicted on with unpainted margins beyond the the sky were left for the small figure, rocks, mediate layer, creating vertical striations. preparation of support: White. Peach- technique throughout, it is possible that on the right is 2.5 cm at its widest and 1.5 tions of eyes in his paintings (see entry), De la Marck in an etching now in the Bib- barbes on all fours sides. There is a striated and tree branches at the right, though colored intermediate layer visible in cross the underdrawing was meant to show preparatory design: A dry material cm at its narrowest. which makes it difficult to interpret the liothèque Nationale, Paris. Yet the present white intermediate layer applied in ran- a reserve for the tree trunk was not left section. through the paint, defining shadow. Late was used for a sketch that was developed importance of changes in the donor’s and hat was copied in a version of the portrait dom manner. in the rock. Changes in the sitter’s collar dendrochronology: Dating not possible, paint changes involved the greater incline further in a liquid medium. The drawing Virgin’s gazes. preparatory design: Scant underdraw- that was in the Arenberg collection in 1905 meant that his chin was painted over an but oak was identified. preparatory design: The panel was in the donor’s head toward the Virgin and is free and searching, with both contours ing includes some contours in a liquid (see Lejeune 1948, 110). area blocked in as sky. underdrawn with a loose, freehand sketch the addition of a cloud behind the donor, and hatching. Numerous deviations preparation of support: White (est.). medium but no shading. Though the In addition to the usual retouch- It appears from the x-radiograph featuring expressive contours with zig- both of which may have been made to occurred between drawing and painting. Traces of a barbe at top and bottom. No drawing is most complex in the cardinal’s ing of edges and joins, some damage that the sky was laid in first, following zag shading in the landscape elements. strenthen the visual relationship between The Child’s body was drawn facing upward intermediate layer observed. wrists and hands, this should not be appears to have been inflicted deliberately. the underdrawing; the sitter’s head was the panels. The reverse of the Virgin and and painted facing out toward the viewer. confused with painted outlines that also Zigzags were not observed in the figure preparatory design: The lines in the Repeated scratches through the left eye painted slightly larger than the reserve Child apparently remained unpainted, His face also shifted, with his gaze moving register in irr. It is possible that the of the donor, in contrast to the Virgin hands appear to be have been made in suggest that the panel might have been left for it; then the cloud was brought to which implies that these relatively large from the Virgin toward the donor panel. same material was used for both the paint on the companion panel (see exh. cat. both liquid and dry media. Elsewhere the damaged by iconoclasts. the edge of the newly painted contour paintings functioned as a stationary dip- The Virgin’s eyes were likewise adjusted and the underdrawing, for many of the Utrecht 2000, 82 – 89). Scribbles behind underdrawing material appears to have of the head; and details such as the curls tych, with the portrait as the moving wing. so that they look more directly down at painted lines and washes appear similar to the knuckles of the proper right hand are been dry. The drawing was sparse, barely the holy family (db no. 45.90) of hair were added at the back of the head. the Child, while her ear was painted lower the drawing in irr. It is unclear if, and to comparable to marks on Lucretia’s hand on indicating the main contours. Shading Frans Hals Museum, Haarlem, inv. no. 683; The painting was harshly cleaned at some than it was underdrawn. what degree, earlier restoration materials the reverse. was not indicated. The left side of the on loan since 1948 from the Rijksdienst point, and much of the ultimate outline register in irr. The underdrawing lies paint layers: The paint handling was fur collar was also drawn narrower than Beeldende Kunst, inv. no. os 75 – 328 was damaged. A restoration was carried over the peach-colored intermediate layer. masterful and assured, despite the large painted, and the sitter’s eyes may have out with the goal of reinstating the artist’s Among changes that occurred between frame: Not original. number of changes. The artist employed been drawn in a different position than intended image, and retouching can be drawing and painting, the little finger and both thin washes of color and touches they are now painted. support: Oak, 64.3  54.6 / 54.5. Two seen to the left of the donor’s forehead, ring finger of the proper left hand were of thicker paint, with fine details planks with vertical grain. The reverse at his lips and chin, and along the top shortened, and the top right angel was not defined in the highlights on the hair, the is unpainted and may have always been of his head. followed exactly in the paint. These obser- filaments of the lips, and the lace edging so. Three horizontal battens were once vations should be confirmed after the attached to the reverse at the top, middle, present cleaning of the painting. and bottom, but squared inserts now span paint layers: The painting method the join, likely applied when the battens was direct, beginning with thin washes were removed. The join is 27.5 cm from the of dark underpaint to block in shadow left at the top, 27.2 cm at the bottom. The areas and continuing to work out the com- bevels on the reverse are distinct at the position in successively lighter paint layers, lateral edges, but barely noticeable at the with lead white used for modeling applied top and bottom. last. Vermeyen used fine, feathery brush- dendrochronology: Earliest possible work in the final surface throughout both felling date: 1520; earliest possible creation panels. This painterly handling of the fur date: 1522; statistically more plausible collar has been more visible in ir than in production date: after 1528. natural light owing to the darkened var- nish. Cross section analysis and additional

298 299 Cat. 36. Rogier van der Weyden paint layers: The paint was applied in drawing material may have been liquid. Cat. 37. Rogier van der Weyden The flesh paint has a pinkish tonality, Cat. 38. Rogier van der Weyden thin, opaque layers, using painstaking The Virgin’s face was drawn higher than softly modeled with a translucent reddish brush strokes to build up form, volume, it was painted. A second column with a brown. The underdrawing for the upper saint george and the dragon and virgin and child with reverse: virgin and child and philippe and detail. The astonishing degree carved figure on top was drawn on the left lid of the Virgin’s proper right eye is faintly virgin and child double pulley de croÿ with reverse: coat of arms of detail must have required the use of a side, nearer the Virgin, but not carried into visible through the whites and serves as a and jean de gros with reverse: coat of the de croÿ family magnification device, for magnification a paint stage, while the lower right corner shadow of the eyelid. A small admixture of arms of the de gros family saint george and the dragon is required to see some of the most distant of the architecture was drawn smaller and of blue pigment shades the lower por- (db no. 4.6) figures. A diagonal depression along the narrower than it was painted. tion of the eyeball. The small portion virgin and child (db no. 6.10) National Gallery of Art, Washington, right side at the top of Saint George’s lance virgin and child (db no. 5.8) of brocade sleeve that is original shows a Masonite, transferred from canvas, paint layers: The paint handling is com- inv. no. 1966.1.1 may be an incised line, for the craquelure Musee des Beaux-Arts, Tournai, inv. no. 481 sophisticated mixture of red lake glazes originally on panel, 50.8  33 parable to that in the Saint George. The paint Examined at nga: 19 – 23 May 2003 by cm, is primarily rectilinear, yet this line does Examined at sral: 15 – 27 September 2003 over broader masses of opaque red with (treatment record, Mark Leonard, is very thin and was applied with preci- ap, rs; 31 March-1 April 2005 by cm, cvd not continue the full length of the lance. by cm, rs, cvd, av highlights placed as precise lines and dots May 1994) sion. The larger background shapes were As is illustrated in the irr, the female fig- of lead-tin yellow. The Huntington Library, Art Collections, Documentation and analyses blocked in first using a monochrome layer, Documentation and analyses ure appears to have been painted over the and Botanical Gardens; The Arabella unknown date and operator at nga: then modeling and detail were added with 2000 at Tournai: dendrochronology by pk completed hill on which she kneels. The D.Huntington Memorial Art Collection, x-radiography thin washes and liquid lines. The Virgin’s 2003 at sral: Phase One visible and ir jean de gros with reverse: coat of x-radiograph also shows a continuous area inv. no. 26.105 2005 at nga: Phase One visible and ir as blue robe was executed with the thick- as well as Inframetrics irr by cvd, arms of the de gros family of increased density from the light green Not examined. well as Inframetrics irr by cvd; binocular est paint. As in the Saint George, detail is av; binocular microscopy by cm, rs; The Art Institute of Chicago, inv. no. hill behind her head. microscopy by cm extraordinarily fine, with two colors used x-radiography by at 1933.1052a-b Not examined. philippe de croÿ (db no. 6.11) frame: Not original. for each jewel in the Virgin’s crown. frame: Not original. virgin and child (db no. 4.7) Koninklijk Museum voor Schone Kunsten, support: Panel, 15.2  11.8. Original sup- Museo Thyssen-Bornemisza, Madrid, support: Oak, 38.7  28.5 / 28.6. Two Antwerp, inv. no. 254 Summary Examined at kmska: 11 – 22 November 8 port with vertical grain has been trimmed, inv. no. 1930.25 Summary planks, vertical grain. The left plank is 3.5 thinned, set into vertically grained panel, Examined at mtb: 13 – 21 September 2004 These two works were most likely the front cm wide at the top and 3.2 cm at the bot- This diptych has suffered tremendously 2003 by cm, rs, cvd, av from damages and restorations. The donor and marouflaged to horizontally grained by cm, rs, cvd, av and back of a single panel, which was tom; the right is 25 cm at the top and 25.4 Documentation and analyses preparation of support: White, extend- repainted narrower. The contour of the panel. A crack extends from the bottom up divided and the two sides marouflaged at the bottom. There are shallow bevels on panel had the back separated from the Documentation and analyses 2003 at kmska: Phase One visible and ir as ing to all edges. Cross section analysis little finger on the Virgin’s right hand through the dragon’s tail and the horse’s onto individual supports. When the paint- the reverse at all four edges. The paint on front, and only about a third of the Virgin 2004 at mtb: Phase One visible and ir well as Inframetrics irr by cvd, av; binoc- established the presence of a peach- departs slightly from the original, and her foot to the height of the horse’s flank. ings are viewed back-to-back, the crack in the reverse is not original. and Child is original. Yet sufficient evidence as well as Inframetrics irr by cvd, av; ular microscopyby cm, rs; x-radiography colored intermediate layer like the one eyes were lower and smaller in an earlier the Saint George aligns perfectly with the of the hand of Rogier van der Weyden dendrochronology: No longer possible micrographs by cm, rs; x-radiography dendrochronology: Earliest possible by at; dendrochronology by pk observed on Cardinal Érard de la Marck. paint stage. A change in the contour of the lower part of the left crack in the Virgin and remains to accept the work as his, and the because of the altered support. by RayXArt, Madrid felling date: 1453; earliest possible creation 2005 at nga: cross sections and cross blue drape to the left of the Virgin creates Child. The Virgin suffered further dam- fact that the main boards of each panel preparatory design: The underdrawing date: 1455; statistically more plausible section analysis by mp a broad diagonal band of highlight lead- preparation of support: White (est.). frame: Not original. age following the separation of the two are from the same tree adds weight to the was relatively summary, though more elab- production date: after 1469. One of the ing to her hand and reinforcing the trian- Hints of an upturn in the ground along paintings, with the first crack extending evidence provided by the nearly identical frame: Not original. orated than in the portrait. Areas where support: Panel, 15.8  11.4, with vertical planks is from the same tree as one used gular composition while emphasizing the all four edges suggest the support was the height of the panel and a new crack dimensions (known to be original thanks the style and method of drawing can be grain, which has been thinned, inset into for Jean de Gros. support: Oak, 51.5  33.6 / 33.45. Two green drapery in the background. As in the trimmed at the barbe. having formed. The intimate kinship to the presence of a full barbe on both best seen in the irr are in the blue drapery an auxiliary panel, and cradled. The visible boards. The original support is 27 cm wide portrait panel, these observations should of the underdrawing material and type, preparation of support: White (est.). paintings) for these panels to be consid- under Christ’s feet and in Joseph’s sleeve, preparatory design: The composition unpainted edges are part of the second- at the top, 26.7 at the bottom, with an be confirmed after the present treatment. the painting technique (including incised There is a barbe on all four sides. Faint, hori- ered two wings of a folding diptych. which was drawn farther down into the was drawn freehand with a liquid material, ary panel and thus not original. There are added modern strip on the right side. The lines), and the precision of the paint appli- zontal striations from an intermediate white garment on the Virgin’s shoulder. possibly executed with a pen. The dragon, two cracks extending from the bottom reverse is painted with the donor’s coat cation points to their execution by the layer are visible in the x-radiograph. Photo­ horse, and Saint George are fully described, through the right side of the Virgin’s of arms. paint layers: The paint handling was Summary same hand. micrographs suggest the intermediate as are the major hills, towers, and wall in robe: the one on the left extends the entire fully comparable to that in Cardinal Érard de These two paintings were recently identi- layer may be flesh-colored. dendrochronology: Earliest possible the background. The underdrawing is height of the panel, the one on the right la Marck. The green background draperies fied as subjects mentioned in the inven- felling date: 1449; earliest possible creation visible to the naked eye in some locations. only halfway. preparatory design: An assured contour are identical across the two panels, as seen tory of Margaret of Austria (see entry). date: 1451; statistically more plausible The female figure does not appear to have line in a liquid material describes the main in irr, and the structure and composition Strong iconographical, compositional, dendrochronology: Not possible production date: after 1465. been part of the original plan, as the line contours (where there is original paint). of the paint is the same. The finishing material, and technical indications do because of the altered support. of the hill on which she kneels continues No hatching was observed. The parts preparation of support: White. There brushwork was similarly fine and feathery, support their having once formed a pair. through her face, and all the lines in her preparation of support: White (est.). No of the painting that are original follow is a barbe on the top, left, and bottom. as in the Virgin’s white sleeve at the proper The sizes of the panels are virtually identi- attire and features that are visible in irr intermediate layer evident. The bottom the underdrawing closely. No intermediate layer is evident in the left shoulder, where a blended scumble cal, and differ from all of Vermeyen’s other can also be seen on the surface. The horse edge has a faint hint of a barbe. Losses and x-radiograph or irr, but cross section of gray was carefully worked with wet extant portraits. The preparation layers paint layers: The painting had been was drawn with a smaller head and rump exposed ground at right and left suggest examination reveals a fawn-colored layer streaks over the white. The mordant for are fully comparable, including the same severely damaged and was largely recon- and with a wide band across his withers, that the ground was upturned at the edges, under the paint overall, and a second, the gold leaf is mostly white, with some peach-colored intermediate layer beneath structed by restorer Joseph-Marie (Jef) van shoulders, and chest. implying that the painted surface remains lighter layer over it in the area of face. blue and black particles, and the center the underdrawing. The underdrawing de Veken in the late 1930s (see Verougstraete largely intact in these places. of the red carnation is gilded, with a red of the cardinal’s hands, being more devel- and Van Schoute 2001, 7 – 28; and exh. cat. preparatory design: An expressive glaze over the gilding (fig. 8). oped than the rest of the panel, may imply preparatory design: The arch, its col- Bruges 2004 – 2005, 62 – 77, no. 3). Still- outline describes the outer contours The irr assembly suggests that the the artist’s reuse of sketches of hands umns, their bases, and the trelliswork original portions include the forehead, of the face and neck, the upper line of the Christ child’s proper left hand was not that he had prepared independently. The inside are incised, and a spare outline eyes, and nose of Christ, the hair, head, sleeve, and the hand. The join of the lips planned in this position, for it overlaps underdrawing of the hands is similar to places the architecture, its carved embel- and neck of the Virgin with the exception is underdrawn, and there are faint indica- the paint used for Mary’s shoulder. There that seen in The Holy Family, which may also lishments, and the figures. The under- of three large losses through the nose and tions of the arc of the eyelids and curve appears to be underdrawing for the pres- use such sketches. The pigment mixture, drawing is schematic, with the Virgin’s mouth, the left cheek and forehead, and of the nose. The shading under the proper ent placement of the hand, but this may be layers, and painting technique in the green eyes indicated with circles and her mouth the left side of the neck. The Virgin’s upper left cheekbone is indicated with a cluster an example of the painter’s characteristic draperies are the same in both panels. The and nose with short upcurved dashes. The hand is largely original, as are the Child’s of fine lines. Thin, possibly ruled lines use of painted outlines. The Virgin’s robe paint handling throughout both panels lower abdomen, shins, and forward foot. mark folds in the green cloth background. was extended in this location, probably to relies on underpainting and direct model- The hand and the end of the sleeve were separate the hand visually from Joseph’s ing with many small adjustments, most drawn farther to the left, then redrawn drapery. Christ’s right foot was painted of which enhanced the relationship of the after the first paint stage had been applied. with more space between the toes, then panels to one another. The works appear to have been pendants, but because the reverse of the portrait was thinned and cradled, it cannot be excluded that they formed a stationary diptych.

fig. 8 Micrograph of the carnation in the Virgin and Child

300 301 Cat. 40. Follower of Rogier van preparatory design: The underdrawing preparatory design: The original donor was fluid and confident. It laid out all portrait and Saint Simon were drawn with epitaph and coats of arms der Weyden, Unidentified artist, important elements of the composition a coarse dry material that skipped over of joos van der burch and and Follower of Gerard David accurately and without hesitation (i.e., the texture of the ground. The drawing katheline van der mersch the hands, including fingernails, were includes both contour lines and paral- (db no. 9.179); reverse of joos van virgin and child and described with single smooth contours). lel hatching. It is descriptive without der burch and saint simon joos van der burch and saint No hatching was observed. The only area being exact. Zones of hatching define of jerusalem with multiple lines is in the hair. The the shadow. The present donor portrait simon of jerusalem with reverse: preparation of support: White. In a drapery folds were completely defined. (see below) is not underdrawn. The lines epitaph and coats of arms of cross section two layers of ground could No notable changes were made from the of drawing for the first donor’s face are vis- joos van der burch and be distinguished. No intermediate layer underdrawing to the paint stage. ible only intermittently, at the eyes, nose, katheline van der mersch is apparent. and midline of the mouth. No drawing Harvard University Art Museums, paint layers: The paint is thin and was was revealed in the landscape. preparatory design: None. Fogg Art Museum, inv. no. 1906a-b smoothly applied, with careful attention Examined at huam: 3 – 7 March 2003 by to linear description. Features such as lips paint layers: The paint was used in paint layers: The coat of arms was cm, ap, rs; and 9 – 18 May 2005 by nk, and eyes were rendered with flat planes opaque as well as transparent layers. Colors painted directly over a different coat of cm, rs of color, while details like fingernails, the mixed on the palette were laid adjacent to arms without blocking it out. The rest Documentation and analyses space between the lips, individual strands one another, sometimes with wet-in-wet of the reverse — including a large helmet 1930s at huam: x-radiography and of hair, and the stream of milk from the blending of edges. Glazes were used in the and red crest that embellished the initial stereo-x-radiography (right wing only) Virgin’s breast were painted with preci- green and red. coat of arms — was covered with a dark by Alan Burroughs sion. Eyelashes were created with short The current donor portrait was overpaint on which the letters for the 1996 at huam: Hamamatsu irr by rs lines of wet paint from the iris pulled over painted over the larger face and shoul- epitaph were applied in mordant gilding. 2003 at huam: Inframetrics irr by ap, rs; the flesh tones. The brocade pattern was ders of an original donor, who had been A cross section from an area of the red crest Phase One ir by rs painted with multiple layers of glaze, as depicted more in profile. X-radiography shows an earlier paint layer, most likely 2004 at huam: strati-x-radiography were folds in the red drapery. The land- and irr reveal the first donor’s eyes and of marbling. Both that layer and the red scape is thorough but lacks Rogier’s scru- nose. The final head was painted with were coated with varnish before further 9 by hl, rs 2005 at huam: Phase One visible and ir pulous touch. soft modeling, at which time the painter paint was applied. by cvd; binocular microscopy by cm, The coat of arms in the window was also extended the landscape over the paint layers: The painting method is rs; binocular microscopy by nk, rs; cross painted over a completed image of Moses. first donor’s hair. These elements have Summary deliberate, thoughtful, and assured. Folds coat of arms of the de croÿ sections and analysis by nk It is difficult to prove without cross sec- now darkened. The production history of the donor wing in the background cloth were rendered family (db no. 6.134); reverse of tions whether this paint is contemporary The painted architecture in the is highly complex. The original reverse was with successive applications of green glaze. philippe de croÿ with the image of the Virgin, but with background, which has suffered con- virgin and child (db no. 9.17) most likely marbled. The helmet and red Underneath the entire cloth lies a reflec- microscopy the paint of the coat of arms siderable damage, appears to have been preparation of support: White. A pink- crest on the coat of arms of the first donor tive base of silver leaf (fig. 9; confirmed frame: Not original. appears to cross over underlying cracks. largely overpainted. A layer of darkened ish brown isolation layer lies over the were painted atop this marbling. Later, by x-ray fluorescence spectroscopy [xrf] This, as well as evidence that the panels blue overpaint was found in two places in preparation. support: Transferred twice, first from the current epitaph was painted over the analysis). The flesh was blended wet-in- were painted by different hands, suggests the bishop’s bright blue garment at the panel to canvas on oak, then to a compos- helmet and red crest and the coats of wet, with defining lines of reddish umber preparatory design: There may be a that it was added slightly later, when the right edge of the panel (to the right of the ite board of redwood blocks, 56.6  36.8 / arms were changed. The present portrait in shadowed areas. The underlying fawn- layout of the contours in a broad curv- wings were joined. orphrey). A cross section from this area 37. There is no evidence of a join in the replaced that of the previous donor, pre- colored paint was used as a middle tone in ing line, but it is difficult to distinguish established that some of the overpaint (a surface. sumably at the same time — shortly after the fingers. The gold necklace was built underdrawing from the linear application joos van der burch and saint medium-rich mixture of azurite with red 1496 (see entry). Concurrently, Saint Simon up using three colors of paint, each lighter of black paint on the surface. preparation of support: White (est.). simon of jerusalem (db no. 9.16) and some black particles) leached into the of Jerusalem and much of the background in hue, superimposed dot upon dot. Each Extensive treatment has made it difficult, original paint (a less rich layer of azurite paint layers: The helmet and swirling frame: Not original. architecture appear to have been blocked bead of the rosary was described with at without cross sections, to determine without any visible admixtures), implying ribbons were underlaid with gold leaf, and out with a dark paint layer. During a sub- least four colors, and the small crucifix at if there is an intermediate layer. Despite support: Oak, 56.2 / 56.0  35.6 / 35.8. that the original layer was cleaned before the wings and shield were underlaid with sequent intervention the overpaint was its end was shaded with red. Highlights the many interventions, a barbe can be Single board, vertical grain. All four edges the overpaint was applied. A large area silver leaf. The red, white, and green verti- removed from the bishop saint, leaving on the velvet garment, which was painted seen at all edges, proof that the original have a shallow bevel. Epitaph and the behind the donor’s shoulder, starting at cal panels were painted around the shapes the second donor’s portrait largely intact with a mixture of red lake and azurite, dimensions are retained. donor’s coats of arms are painted on the the lower left corner of the clasp of the defined in metal leaf, which created sharp but confusing the contours of the first were achieved with strokes of white paint reverse. bishop’s robe, was also overpainted. silhouettes in the x-radiograph. The areas and second donor’s shoulders. This inter- brushed while wet to create feathered of gold and silver leaf (confirmed by xrf dendrochronology: Earliest possible vention probably also necessitated repair contours. The monogram in the upper left analysis) were articulated with linear shad- felling date: 1477; earliest possible creation of damaged paint surfaces, as in the rich was executed in paint, rather than in mor- ing and decorative jewels, and patterns date: 1479; statistically more plausible blue garment of the bishop at the far right dant gilding. The hand was left in reserve were applied with a sure hand. The inscrip- production date: after 1493. and in the background architecture. following the first drawing, and the first tion was painted with yellow paint. paint layer of the sleeve ended at the point preparation of support: White. In a indicated in that drawing. The hand and cross section, four layers of ground could sleeve were repainted after their position Summary be distinguished, three relatively coarse was redrawn. The studied but masterful handling layers and a finer top layer. There is a con- of paint is typical of Rogier, as is the tinuous barbe around all four sides. No adjustment in the positioning of the sit- intermediate layer is apparent. ter’s hands. Both silver and gold leaf were used in the depiction of De Croÿ’s coat of arms, and the presence of silver leaf in the background of the donor portrait and gold leaf in the background of the Virgin and Child may support the concept that these two panels originally formed a fold- ing diptych.

fig. 9 Micrograph from background of the Philippe de Croÿ, show- ing silver leaf under the green surface paint

302 303 Detail of fig. 1 showing a diptych hanging on the wall behind the kneeling figure of De Hondt

1 Master of 1499, Abbot Christiaan de Hondt

The term “Netherlandish” is applied to north- size. The courts of the dukes of Burgundy were ern Renaissance art produced in the region that among the richest in Europe, and young men encompasses present-day Belgium, The Nether- from other areas were routinely sent there to lands, and Luxembourg as well as Burgundy and be educated in noble and chivalrous behav- other northern French provinces. For most of ior. Moreover, the region was a financial and the fifteenth century this territory was ruled by mercantile hub of northern Europe. For most the dukes of Burgundy, a cadet line of the Valois of the fifteenth century Bruges was a center for kings of France. In the decade following the both banking and the production of luxury death of Duke Charles the Bold in 1477, the Bur- goods. In the early sixteenth century economic gundian Netherlands became part of the Holy dominance shifted to Antwerp, home of a lively Roman Empire, which also included Austria, import-export trade and the Bourse, or financial Germany, and Spain. exchange. Art also flourished, and the Nether- In the fifteenth and early sixteenth centu- lands, along with Italy, was one of the preemi- ries the power and influence of the Burgundian nent centers of painting in the Renaissance. Netherlands far outweighed its relatively small 2 Jan van Eyck, The Angel Gabriel with The Virgin Annunciate

what is a diptych? and they can also depict coats of arms or other images related to those inside. A diptych consists of two panels of the same size Unfortunately, the components of diptychs and shape that are framed and hinged so they are easily separated and their frames and hinges can be opened and closed like a book, as seen in lost or altered. Indeed, only two diptychs in the the Bibaut diptych, which depicts the Carthu- exhibition, those by Memling and by Provoost sian abbot Willem van Bibaut praying to the Vir- (see figs. 7, 8 – 9), still definitely have both gin and Child (see cover). The format has a long their original frames and hinges. Nonethe- history, and antecedents can be found in classi- less, technical examination of diptych panels cal Greek and Roman writing tablets as well as often reveals traces of now-lost hardware or in the numerous hinged, carved ivory diptychs framing elements, allowing for the reconstruc- that survive from late Roman times through the tion of diptychs that have been divided. This Gothic era. The primary images are protected exhibition brings together halves of numerous on the interior of the diptych and are revealed diptychs that are now dispersed and residing only when the ensemble is opened. The subject in distant collections, and these are presented matter on the interior is usually religious but alongside diptychs that have remained intact also includes secular portraits. The exteriors are over the centuries. often painted to look like marble or porphyry, how were diptychs used? prayers and private devotion in northern europe Most surviving Netherlandish diptychs are small or modest in size. Those with religious The popularity of diptychs in the Low Coun- subjects likely accompanied private devotion, tries coincided with a change in religious prac- primarily in a domestic setting, although use tice there. A movement known as the Modern in the family chapel of a church is also possible. Devotion (Devotio Moderna) was founded in the Images in manuscripts and paintings provide Netherlands by two clerics, Geert Groote and some information about ways they were dis- Florens Radevrijns, in the late fourteenth cen- played. The portrayal of Abbot Christiaan de Hondt tury. Emphasizing private, personal devotion, (fig. 1) by the Master of 1499 shows a diptych they urged their followers to focus on Christ’s hanging on the back wall of the prelate’s room. humanity more than his divinity, to emulate Yet because diptychs could be folded, many were his humility and empathize with his suffer- probably kept in drawers or special containers ings. Solitary meditation on Christ’s Passion until needed and then set upon a table or prie- and redemption, on one’s own death, the Last dieu (a small desk for prayer), possibly on a cush- Judgment, heaven, and hell was essential. In ion, or opened at an angle so that the diptych could stand upright.

3 Jan Gossaert, Virgin in the Church with Antonio Siciliano and Saint Anthony 4 Robert Campin, The Trinity with Virgin and Child the course of the fifteenth century the Mod- frame imitates a yellowish stone. The figures ern Devotion found adherents throughout the of Gabriel and Mary and the plinths on which Netherlands and Germany. Its precepts were they stand seem to be carved from a brown- further disseminated in texts such as The Imita- ish yellow stone with the warmth and sheen of tion of Christ by Thomas à Kempis, which reached alabaster. Thanks to Van Eyck’s mastery of light an increasingly literate public. In this context and shade, the statues appear to occupy a shal- small works of art such as diptychs that pro- low space, and the edges of the plinths seem to vided a focus for private worship enjoyed wide project slightly beyond the edges of the frames. popularity. Clients included bankers, mer- It is not known who commissioned this dip- chants, and other members of the middle class tych, but the Annunciation, the moment when as well as nobles and clerics. Gabriel announces the coming birth of Christ, was evidently an important part of the owner’s jan van eyck and robert campin personal devotion. Van Eyck’s influence on the next genera- Among the earliest Netherlandish diptychs tion of Netherlandish painters is evident in are those by Jan van Eyck and Robert Campin, Jan Gossaert’s Virgin in the Church, which copies rightly acclaimed as the founding fathers of a major painting by Van Eyck now in Berlin. Netherlandish painting. In The Angel Gabriel Gossaert’s panel is paired with Antonio Siciliano and The Virgin Annunciate (fig. 2) Van Eyck cre- and Saint Anthony, which portrays the donor with ated an astonishingly convincing painted his patron saint (fig. 3). Antonio Siciliano was illusion of sculpture. The outermost frame is secretary to the Duke of Milan, who sent him fictive red marble or porphyry, while the inner on a diplomatic mission to the Netherlands in Wearing a voluminous fur-lined robe, Mary 1513. Because neither of the frames is original holds the Christ child in her lap and extends her and all trace of hinges consequently lost, it is right hand toward the fire, perhaps to warm it difficult to confirm that these two paintings or to make sure the fire is not too hot for her son. formed a diptych. Possibly they were pendants, The ewer, basin, and white towel imply that the that is, a pair of paintings intended to be seen Child has had or will be given a bath, and at the side by side. same time they symbolize the Virgin’s purity. The personal nature of diptychs is under- This diptych thus juxtaposes Christ’s divinity, scored in Campin’s Trinity with Virgin and Child death, and resurrection on the left panel with (fig. 4). The combination of subjects is possibly his human existence and his mother’s tender unique in Netherlandish painting and probably care on the right. reflects the owner’s wishes. On the left wing God the Father, wearing a papal tiara, holds the rogier van der weyden and the lifeless body of Christ the Son, displaying the devotional portrait diptych wounds in his side, hands, and feet. Together One of the most widespread types of Nether­ with the dove of the Holy Spirit on Christ’s landish diptych has an image of the Virgin shoulder, these figures were meant to inspire and Child on one panel, often the left, with a the viewer to contemplate the mystery of the portrait of a donor on the facing panel in a pose Trinity. The right wing depicts the Virgin and of supplication. It is generally agreed that this Child in a contemporary domestic interior. format was created by Rogier van der Weyden,

5 Rogier van der Weyden, Virgin and Child with Philippe de Croÿ (right panel before treatment) 6 Michel Sittow, Virgin and Child with Diego de Guevara (?) who ranks with Jan van Eyck as one of the great- is Michel Sittow’s Virgin and Child from Berlin est artists of the fifteenth century. An outstand- and Diego de Guevara (?) in the National Gallery ing example of the devotional portrait diptych of Art’s collection (fig. 6). The Child gently is Rogier’s Virgin and Child with Philippe de Croÿ reaches up to touch Mary’s chin while holding (fig. 5). Identified by inscriptions and the coat in the other hand a goldfinch, an allusion to his of arms on the reverse of his portrait, the sitter future suffering and crown of thorns (because was then about twenty-seven years old and a ris- of the bird’s supposed taste for thorny plants). ing star at the Burgundian court. Described by The infant rests on an oriental carpet that covers a contemporary chronicler as proud, eloquent, a foreground parapet extending onto the right- ambitious, and talented, Philippe de Croÿ dis- hand panel, a feature that confirms the two tinguished himself in later life as both a dip- pictures originally formed a pair. The donor has lomat and a warrior. Here he is shown in rapt been identified on the basis of the Cross of the adoration, holding a rosary. On the left wing Order of Caltrava that appears partly obscured the Virgin supports an energetic Christ child, on his doublet just above his hand. In 1517 who stands unsteadily on a richly textured red Diego de Guevara was appointed to a high posi- and gold cushion and leans forward to play with tion in the order by Charles v, the Holy Roman the clasps of an elegant leather-bound book, Emperor. A Spaniard by birth, Guevara spent perhaps a Book of Hours (a book of prayers used more than forty years in service at the Burgun- in private worship at home). Philippe was a book dian court and was entrusted with diplomatic lover and collector and would have delighted in missions to Spain and England. Perhaps because an image of such a deluxe volume. Painted with he is contemplating Christ’s sacrifice and the superb precision and technical finesse, these Virgin’s sorrow, his expression is imbued with works are filled with a sense of religious intensity. an ineffable and restrained sadness that conveys Another exquisite devotional portrait dip- the sitter’s inner emotions to a degree rarely tych reunited in this exhibition (see also fig. 5) found in Netherlandish painting. 7 Hans Memling, Virgin and Child with Maarten van Nieuwenhove

A significant change in the devotional por- diptychs and religious orders trait diptych was achieved by Hans Memling Well suited to solitary contemplation, dip- in his Virgin and Child with Maarten van Nieu- tychs were used by members of religious orders, wenhove (fig. 7). Instead of the dark, featureless including Carthusians, Franciscans, and Cister- background favored by Rogier van der Weyden cians. One of the most unusual works in this and his followers, Memling depicted the holy category (fig. 8) is by Jan Provoost, who was figures and the donor in a spatially coherent active in Bruges. The interior left wing depicts room in Van Nieuwenhove’s house. On the left Christ Carrying the Cross: his hands are bound, panel the Virgin offers an apple to the Christ and the crown of thorns pressed into his head child, who looks toward the donor (identified causes rivulets of blood to run down his face. by his coat of arms in the window behind her). Behind the cross at the right the grotesque faces A circular mirror on the back wall reflects the of his tormentors may reflect an awareness Virgin from the back and Van Nieuwenhove in of Hieronymus Bosch, while at the left John profile, establishing that the two figures are the Evangelist and the Virgin can be glimpsed together in the same room. On the right wing shedding tears. With a look of gentle but pained the elegantly attired donor is portrayed in an resignation, Christ turns toward the donor, a attitude of prayer. The oriental carpet on the Franciscan whose age, fifty-four, is written on parapet in front of him also appears on the left the frame. The monk looks fervently toward the wing, a device that further unifies the two pan- savior, his fingers steepled in prayer. The imag- els. Maarten van Nieuwenhove, only twenty- ery makes abundantly clear one of the major three years old when this diptych was created, precepts of the Modern Devotion, for the monk was a member of a prominent family in Bruges, seems to witness directly Christ’s pain and where he later served as burgomaster (mayor). humiliation as he goes to be crucified. As one His career ended prematurely in 1506 when he of the two diptychs in the exhibition with its died at the age of thirty-six. 8 Jan Provoost, Christ Carrying the Cross with Portrait of a Fifty-four-year-old Franciscan

9 Imitation Porphyry and Skull in a Niche, outer wings of fig. 8 original frames and hinges (see also fig. 7), this from diptych to pendant work is also exceptional in having two painted Although religious images dominate this exhi- reverses in excellent condition (fig. 9): the bition, purely secular subjects are also repre- reverse of the left wing is painted to resemble sented on Netherlandish diptychs. Technical porphyry, while the reverse of the right wing examination of Bernard de Rijckere’s portraits bears the image of a skull in a niche, a reminder of the prosperous Antwerp merchant Adriaan of the inevitability of death and the need for van Santvoort and his family (fig. 10) revealed repentance. 10 Bernard de Rijckere, Adriaan van Santvoort and His Sons Guillaume and Adriaan with Anna van Hertsbeeke and Her Daughter Catharina and Son Jan Baptiste that the panels initially formed what can be of images on the inner wings produced an inti- called a folding stationary diptych. The original mate dialogue that was particularly conducive frames were once hinged together, and because to prayer and contemplation. This relationship the panels are relatively large and the back of the could be intensified by opening the diptych to a left one is unpainted, it is thought to have been particular angle to enhance the communication attached to a wall, while the right panel — with between the figures on the two panels, whether the family’s coats of arms painted on the reverse — it was Gabriel addressing the Virgin or a donor folded over it. At some point, perhaps in the before the object of his or her devotion. To illus- seventeenth century, the hinges were removed trate, the exhibition adopts this mode of display and the diptych was transformed into two pen- for a number of diptychs. dants meant to be hung side by side. It is hard to pinpoint the exact reasons for the decline in the popularity of the diptych for- mat in the late sixteenth and early seventeenth century. The aftermath of the Protestant Refor- For more than a century and a half the diptych mation and the political and religious turmoil was an essential component of Netherlandish that engulfed the Netherlands beginning in art. The inner wings of diptychs often attained the 1560s were contributing factors. Whatever a remarkable unity through either a composi- the cause, the religious climate was altered and tional device such as a setting that continued the intimacy inherent in diptychs was displaced through both panels or the exchange of glances by the more doctrinaire public statements of between holy figures and donors on adjoining Counter-Reformation altarpieces. panels. Artists realized that the juxtaposition programs concert in honor of the exhibition lecture series

Sunday, November 12, 2006, 6:30 pm Early Netherlandish Painting West Building, West Garden Court East Building Auditorium, 10:15 am J. Russell Sale The Suspicious Cheese Lords (male vocal ensem- ble) will sing music by northern Renaissance This series presents an overview of painting composers. Concerts at the National Gallery of in the Low Countries during the fifteenth and Art are open to the public, free of charge. First- sixteenth centuries. come, first-seated admission begins at 6:00 pm. October 5: The Age of Jan van Eyck and Rogier van der Weyden lectures October 12: Crosscurrents of Piety and Humanism East Building Auditorium in the Later 15th Century Sunday, November 26, 2:00 pm October 19: The Rise of Genre Painting and Landscape Unfolding the Netherlandish Diptych: An Introduction to the Exhibition October 26: Nature, Reform, and the Lure of Italy John Hand, curator of northern Renaissance in the Sixteenth Century paintings, National Gallery of Art Catherine A. Metzger, senior conservator of family workshop paintings, National Gallery of Art It Takes Two!

Saturdays, January 13, 20, and 27 Saturday, December 2, 2:00 – 5:00 pm An Inside Look at Netherlandish Diptychs Morning workshop: 10:30 am to 12:30 pm Illustrated lectures by noted scholars of Nether- Afternoon workshop: 1:30 to 3:30 pm landish art, Al Acres, Carol J. Purtle, and Victor ages 9 to 12 M. Schmidt. A panel discussion will follow. Registration begins January 3 Meet at West Building Lecture Hall gallery talks Parents and children will learn about portrai- All Gallery talks will begin at the West Building ture and symbolism; then create their own dip- Rotunda. tychs to take home. Led by museum educator Nathalie Ryan. Open Discussion: Early Netherlandish Painting in the National Gallery of Art Workshops are free, but preregistration is October 26, 27, and 28 at 12:00 noon required. To register, fill out the online registra- October 29 and 31 at 1:00 pm tion form at www.nga.gov/kids or call 202.789.3030. November 2 at 10:30 am J. Russell Sale, senior lecturer, teacher workshop National Gallery of Art Unfolding the Netherlandish Diptych Prayers and Portraits: Unfolding the Saturday, December 9, 2006 Netherlandish Diptych (Saturday, December 16, 2006, if repeated) December 7, 8, and 9, 12:00 noon Emphasizing the fabrication and conservation December 12 and 14, 1:00 pm of panel paintings, this program is particularly J. Russell Sale appropriate for chemistry and studio art teachers. Fee: $20. To register, call 202.842.6796. audio guide The exhibition is organized by the National Gallery of Art, Washington, and the Koninklijk An audio tour is available at the entrance to Museum voor Schone Kunsten, Antwerp, in the exhibition for $5. Narrated by National association with the Harvard University Art Gallery of Art director Earl A. Powell iii, this Museums, Cambridge. tour includes commentary by John Oliver Hand, curator of northern Renaissance paintings, Additional support is provided by the National Gallery of Art; Ron Spronk, associate Flemish government. curator for research, Straus Center for Conser- vation, Harvard University Art Museums; and The exhibition is supported by an indemnity Laura D. Gelfand, associate professor of art his- from the Federal Council on the Arts and tory, University of Akron. To reserve audio tours the Humanities. for groups, call 202.842.6592.

This brochure was written by John Oliver on the web Hand, curator of northern Renaissance paint- http://www.nga.gov/exhibitions/diptychinfo.htm ings, and produced by the department of exhibition programs and the publishing office. catalogue Copyright © 2006 Board of Trustees, National The exhibition is accompanied by a fully illus- Gallery of Art, Washington. trated, 352-page catalogue by John Oliver Hand and Catherine A. Metzger of the National Gal- list of illustrations lery of Art and Ron Spronk of the Straus Center cover Master of the Magdalen Legend, Virgin and Child, for Conservation, Harvard University Art 1490/1500, and unknown French artist, Willem van Bibaut, 1523, Private Collection (no. 23) Museums. Produced by the publishing office 1 Master of 1499, Abbot Christiaan de Hondt, 1499, Koninklijk of the National Gallery of Art and published Museum voor Schone Kunsten, Antwerp (no. 21), with detail in association with Yale University Press. 2 Jan van Eyck, The Angel Gabriel with The Virgin Annunciate, Hardcover $75.00. c. 1435 / 1437, Museo Thyssen-Bornemisza, Madrid (no. 8) 3 Jan Gossaert, Virgin in the Church with Antonio Siciliano and Saint Anthony, c. 1513, Galleria Doria Pamphilj, Rome (no. 13) general information 4 Robert Campin, The Trinity with Virgin and Child, c. 1433 – 1435, The State Hermitage Museum, Saint Petersburg (no. 6) Hours: Monday-Saturday 10:00 am – 5:00 pm, 5 Rogier van der Weyden, Virgin and Child, c. 1460, The Sunday 11:00 am – 6:00 pm. Gallery Web site: Huntington Library, Art Collections, and Botanical Gardens; The Arabella D. Huntington Memorial Art Collection; with Philippe www.nga.gov. For information about accessibility de Croÿ, c. 1460, Koninklijk Museum voor Schone Kunsten, to galleries and public areas, assistive listen- Antwerp (no. 38) 6 Michel Sittow, Virgin and Child, c. 1515 / 1518, Staatliche Museen ing devices, sign-language interpretation, and zu Berlin, Gemäldegalerie, Property of the Kaiser Friedrich- other services and programs, inquire at the Art Museums-Vereins; with Diego de Guevara (?), c. 1515 / 1518, National Gallery of Art, Washington, Andrew W. Mellon Information Desks, consult the Web site, or call Collection (no. 34) 202.842.6690 (tdd line 202.842.6176). 7 Hans Memling, Virgin and Child with Maarten van Nieuwenhove, 1487, Municipal Museums, Bruges, Hospitaalmuseum Sint- Admission to the National Gallery of Art and all Janshospitaal (no. 26) 8 Jan Provoost, Christ Carrying the Cross with Portrait of a Fifty- of its programs is free, except as noted. four-year-old Franciscan, 1522, Municipal Museums, Bruges, Hospitaalmuseum Sint-Janshospitaal (no. 31) 9 Imitation Porphyry and Skull in a Niche, reverses of fig. 8 10 Bernard de Rijckere, Adriaan van Santvoort and His Sons Guillaume and Adriaan with Anna van Hertsbeeke and Her Daughter Catharina and Son Jan Baptiste, 1563, Private Collection (no. 32) material and technical aspects of the netherlandish diptych

14 15 This essay is excerpted from Prayers and Portraits: Unfolding the Netherlandish Diptych, Copyright © 2006 Board of Trustees, National Gallery of Art, Washington, available November 2006.

technical examination has been a vital component of our and that have research facilities. We started at two of the organizing insti- the format provides inherent protection. The reverses of both wings in a Larger panels consisted of multiple boards glued together, with the joins plan since an exhibition on the Netherlandish diptych was first conceived tutions: the Harvard University Art Museums in Cambridge, Massachu- folding diptych are typically also painted, though often with secondary often reinforced by dowels. The direction of the grain usually follows the in 1997. We hoped — and expected — that material and technical findings setts, and the National Gallery of Art in Washington, dc. We also worked imagery such as coats of arms or marbling.2 The two general exceptions, largest dimension of the panel, which in our examinations was always would answer basic but essential questions about the production, original at the Koninklijk Museum voor Schone Kunsten in Antwerp; the Staat­ both of which appear to have been less common, concern relatively large vertical (we saw no horizontally oriented diptych wings). Until the end use, and subsequent history of these objects. Our first goal was to gather lichen Museen zu Berlin, Gemäldegalerie; the Museo Nacional del Prado or relatively small objects. Large “stationary” diptychs had only one mov- of the first quarter of the sixteenth century most Netherlandish paintings as much information on individual works as possible, and we felt encour- and the Museo Thyssen-Bornemisza in Madrid; the Stedelijke Musea in ing panel, painted on both sides, while the other wing was affixed to a wall were framed before the painting process started, and this does not seem aged when our efforts revealed significant new information on several Bruges; and the Philadelphia Museum of Art. A number of works from and had a reverse that usually remained roughly finished and unpainted. to have been any different for diptych wings. After a support panel arrived paintings. In addition to this general approach, we had specific questions. other museums and from private collections were transported to the Stich- Such works would serve as altarpieces or epitaphs in public settings. With at the painter’s workshop, it was covered with a paste of animal glue and First, because so many diptychs have been dispersed over time, a funda- ting Restauratie Atelier Limburg in Maastricht and to the Straus Center respect to one of the smallest works we studied, the Berlin Annunciation chalk (calcium carbonate) to provide a light colored, smooth surface for the mental issue was establishing whether two now-separated panels were for Conservation and Technical Studies at Harvard to be studied and diptych by the Master of 1499 (cat. 20), the reverses have smoothed sur- paint layers. An underdrawing was often executed on this ground layer to once attached to one another. Second, we needed to determine if currently documented there. faces coated with a monochrome black paint but no ground layer. A Virgin lay out the main forms of the composition. This underdrawing can often paired panels retained an original relationship or if they had been joined Most of the tools used in our investigation have gained long- and Child attributed to the Master of the Magdalen Legend (the left wing be detected with infrared photography or, achieving better results, with later. Third, if two paintings did form an original pair, could we demon- standing recognition in the field of technical art history. We requested or of cat. 23) was also treated this way,3 while the exterior of the book-shaped infrared reflectography. The highly absorbent ground layer was often strate that they constituted a folding diptych or were autonomous pendant commissioned x-radiography in advance of our travels so that we would diptych with a Portrait of a Man and Portrait of a Woman attributed to the coated with an intermediate layer that functioned as a sealant, which paintings? Were they hung on a wall or displayed in another way? Are there have this reference material available during examinations. Peter Klein Master of the Benson Portraits (cat. 18) has similar surfaces. The latter could lie either under or over the underdrawing. The actual paint layers clues that the paintings were meant to be seen at a particular, ideal angle performed new dendrochronological analyses and shared earlier findings panels we believe may have had a leather or cloth binding.4 Although small consist of an underpainting, over which the final paint layers and medium- when opened? with us. Whenever possible, the paintings were unframed for our study. diptychs without decoratively painted reverses might have been designed rich glazes were applied. X-radiography was used to study the buildup Initial examinations in a number of cases revealed surprising We scrutinized the paintings in good light, along with their frames and for permanent display, it is equally possible that some were intended as of the paint layers, although infrared studies can also be informative in differences in the method of production between two paired paintings, reverses. Two members of the team, Catharina Van Daalen and Adriaan hand-held, portable devotional aids and may have had decorative mate- this regard. The precise nature of the layers of paint in a specific area of a which led us to formulate the working hypothesis that such works were Verburg, documented the works using infrared reflectography and rela- rial such as leather, velvet, or brocade on the reverses. In any case, the great picture can be determined through the study of paint cross sections.6 probably produced at different times, possibly by different hands. By exten- tively high resolution digital photography, in both the visible and the majority of folding diptychs had paintings on all four surfaces, which is an sion, if one image from a popular pairing appeared in greater numbers infrared regions of the electromagnetic spectrum — the latter being a important factor in distinguishing diptych panels from pendant paintings. d i p t yc h fr a me s than the other panel — for example, the Virgin and Child from a devotional recent innovation in the field.1 The other two team members, Catherine The construction methods that we observed in the wings of dip- The construction of the frames provides vital information about the works portrait diptych — we surmised that the more frequently recurring image Metzger and Ron Spronk, concentrated on visual inspections and on study tychs correspond to the prevailing practices for other early Netherland- in question, and twenty-nine of the sixty-five panels we studied still had could have been produced “on spec” for the open market, with a client later of the works with the binocular microscope (for a complete record of this ish panel paintings, the only difference being that the two panels were some or all elements of their original framing (see cats. 3, 8, 18, 19, 20, 21, 23, commissioning a portrait panel to complete a diptych if desired. Hence a research on paintings in the exhibition, see the technical appendix later in attached to one another by means of two metal hinges. Shallow recessed 24, 26, 28 – 32, and p. 10, figs. 6 a,c). Following standard production meth- fourth research goal was to look for indications that some of these paired this catalogue). areas carved into the sides of adjoining frame members typically accom- ods for Netherlandish panel paintings, different types of frames were used pictures might have been painted in a staggered or “phased” process Based on painstaking examinations, extensive documentation, and modated hinge plates that lay flush with the surface of the frame and for diptychs, generally depending on the size of the panels. The frames and of production. meticulous review and double-checking of the data, we can offer several allowed the diptych to be fully opened and closed while the hardware support panels of small diptychs (less than about 30 cm high) were usually A collaborative research grant from the Getty Foundation allowed general observations about material aspects of Netherlandish diptychs and remained inconspicuous when the ensemble was open. The lacunae thus carved out of a single plank, with this “integral” frame having a raised a research team of four to examine and document twenty-five pairs their production. These findings have implications regarding the practical reveal the location of hinges even if the hinges have been removed. Some profile only on the front of the panel and a flat surface on the reverse. of paintings, fourteen single panels that were thought to have once been use of the objects as well as the processes of their design and execution. In wood also had to be carved out of the edges of the frames immediately Occasionally larger diptych panels with an arched top, such as the Aachen part of diptychs (many of them now dispersed), as well as one autonomous some cases, we found clear indications for workshop participation in the above the hinges to allow for the insertion of hinge rods. Ecce Homo and Mater Dolorosa (cat. 3), had integral frames, and the use of panel that influenced the subsequent tradition of Netherlandish paired production and for close involvement of the donor in the final appearance The structure of Netherlandish panel paintings is relatively con- integral frames continued well into the sixteenth century. Indeed, sixteen paintings. Ideally, we would have examined such a large group of pictures of the work. sistent throughout most of the fifteenth and sixteenth centuries. The of the twenty-nine panels with any remnant of original framing had inte- with the same equipment under identical conditions, but that was not As described in greater detail in the preceding essay, a diptych con- manufacture of the panels and the frames required specialized skills, and gral frames (see cats. 3, 18, 20, 21, 23, 24, 28, and 30). feasible because we had to travel to see many of the paintings on site. Thus sists of two hinged panels of the same size that fold open and closed like a it is assumed that a woodworker rather than a painter would have pro- the selection of the works studied was made on pragmatic grounds. For book. Because the main imagery is visible only when the diptych is opened, duced these.5 Support panels were made of oak imported from the Baltic reasons of efficiency and economy, we focused our resources on those col- region, which can be dated with dendrochronology (or tree-ring analysis). lections that house relatively large numbers of diptychs or diptych wings

16 17 Diptychs with two-sided integral frames do exist, with raised profiles on both front and back of the panels. For the Virgin in the Church and Abbot Christiaan de Hondt now in the Koninklijk Museum voor Schone Kunsten in Antwerp (cat. 21), each wing was made from a single oak plank, arched at the top, and integral frames project at the front and back of both panels. All four surfaces are fully finished, and the frames are elaborately painted as imitation stone. Two-sided integral frames are also found on the Master of the Saint Ursula Legend diptych in Antwerp (cat. 24) and on the donor wing of the Master of the Magdalen Legend diptych in a private collection (cat. 23). Rectangular panels of medium size (30 – 45 cm high) or larger are 10 usually provided with “engaged” frames, made out of four grooved pieces of wood that are placed around the edges of the panel and joined at the corners, often with tenons and pins for reinforcement (those who fash- Its method of construction is fully comparable with that of other frames ioned the frames apparently varied the type of the joinery according to from the period. Although the support panel has shrunk considerably over whether a work would be hanging or displayed in a standing position).7 time, and relatively broad strips of unpainted wood have become visible at The four grooved members, attached to the panel before the ground the top and lateral edges within the frame, there can be no doubt that the 11 and paint layers were applied, created a frame on both the front and the panel was painted within this frame: traces of the red dots on the gilded reverse of the panel. Even when such frames were later removed, one can background of the panel can be observed on the frame using a microscope

recognize this type of framing by the unpainted margins on the panel’s 9 (fig. 10), and there is a corresponding topography in the break between the edges (initially covered by the frame) and by the presence of “barbes,” or paint film on the panel and that on the frame. Knowing that the frame is ridges of ground and paint between the unpainted margin and the paint original, it becomes significant that there are two slots carved for hinges surface where the frame edge and panel originally came together (fig. 9). vertical members of the interior frames were carved from the same plank in the right frame member. The hinges are now missing but were attached Moreover, the presence of unpainted margins and a barbe indicate that as the panels, but the moldings at the top and bottom, with the grain with hand-forged nails, some of which are still visible in the x-radiograph a panel that has lost its engaged frame still has its original dimensions. running across that of the panels, were produced out of separately carved (fig. 11). The x-radiograph also reveals corrosion from a nail hole just below Eight diptychs from which we studied one or both wings, now all dispersed, pieces of wood. Finally, an exceptionally large portrait diptych by Bernard the vertical center of the left frame member, which must have served to either still had their engaged frames or showed barbes at their perimeters de Rijckere (cat. 32) has “box” frames constructed of mitered boards, attach a closing device. This nail was hammered “horizontally” into the indicating that they were originally painted inside such a frame. Of these which were placed around the painted panels after the painting process left outer edge of the frame, parallel to the support panel — an important eight ensembles, six were of medium size (cats. 15, 25, 37, 38, 40, and p. 10, was completed. indication that the panels originally constituted a diptych, for only that figs. 6 a – d), while the other two were much larger and presumably func- The close examination of the original frame can hold crucial format would be closed by a hook and eye placed on the outer edges of the tioned as stationary diptychs (see pp. 6 – 7, figs. 4 a – c; and Jean Fouquet’s information for the reconstruction of a diptych. For example, our study wings’ frames. (The wings of a triptych, which close side by side like win- “Melun Diptych”). of the frame on the Virgin and Child by the Master of the Saint Ursula Legend dow shutters, would have closing devices on the reverse faces of the frame, In addition to the integral and engaged frames, we encountered in Harvard’s Fogg Art Museum has definitively confirmed the proposal attached with nails hammered perpendicular to the support panel.)9 The three other types of frames in this project, but they seem to represent rare that the panel originally functioned as the left wing of a folding diptych, decisive evidence linking the Fogg’s Virgin and Child with Philadelphia’s por- occurrences. The wings of one relatively small work, the Lentulus diptych adjoining the portrait of Lodovico Portinari in the Philadelphia Museum trait of Lodovico Portinari was found during our examination of the reverse from Utrecht (cat. 29), have original “applied” frames, for which the four of Art (see figs. 6 a,b). The frame on the portrait of Lodovico Portinari has been of the latter, where ample remnants of a green paint appear on the outer separately carved frame moldings were nailed onto the outer edges of the lost, but the engaged frame on the Fogg’s Virgin and Child is clearly original.8 perimeter of the crimson background (fig. 12) that are identical to the color front surface of the panels. Jan van Eyck’s Annunciation diptych in Madrid of the inner bevels of the frame on the reverse of the Virgin and Child. 12 and the Saint Peter and Saint Paul panels from Antwerp (cats. 8, 19) have “semi-integral” frames, combining integral and applied framing elements:

fig. 9 fig. 10 fig. 11 fig. 12 Macrograph of left edge of Joos Macrograph of top edge of X-radiograph of the Master Micrograph of left edge from van der Burch and Saint Simon the Master of the Saint Ursula of the Saint Ursula Legend, the Master of the Saint Ursula of Jerusalem (cat. 40), showing Legend, Virgin and Child (p. 10, Virgin and Child (paint on the Legend, Coat of Arms of the unpainted margin and barbe fig. 6a), showing background reverse also registers) Portinari Family (p. 10, fig. 6d), dots overlapping frame showing remnants of green paint

18 19 We also had the opportunity to examine briefly the engaged frames covered with black paint (p. 219, fig. 1), presumably in the seventeenth but not associated until much later, when they were placed in the present of two diptychs that are still fully intact: Hans Memling’s Virgin and Child century when this style became popular.14 Each panel has a hanging device composite frame. with Maarten van Nieuwenhove (cat. 26) and Jan Provoost’s Christ Carry­ing the that consists of a wooden shank attached by hand-forged nails (fig. 13). We The Man of Sorrows and Mater Dolorosa in the Fogg Art Museum (cat. 4), Cross with the Portrait of a Fifty-four-year-old Franciscan (cat. 31). Both are medium- are not certain if these devices date from 1563, for they are fully comparable which are also attributed to Albrecht Bouts, showed significant differences size half-length devotional portrait diptychs (measuring roughly 50  40 with similar objects from the seventeenth century,15 when this diptych was in production method as well, most notably in the handling and style cm, including the frames). The molding profiles of the frames differ from modernized and made into two pendant portraits. of their underdrawings and in the ways their gilded backgrounds were one diptych to the other, but all surfaces of both pairs were covered with Only three diptychs that we studied still appear to have their origi- adorned with reddish dots. Yet these works have probably always been a either paint or gold leaf. There are differences in the ways the joins were nal hinges — the works by Provoost and Memling discussed above and that pair. The painting technique is fully comparable for both pictures, and constructed, however, which may hold information on the practical handling by the Saint Ursula Master in Antwerp (cat. 24) — but because a folding dendrochronological analyses established that the support panels came of these works. The joins of the frame for the Provoost diptych have hori- diptych by definition has hinges, we expected to find evidence of hinges from the same tree.16 It is no longer possible to determine in what exact zontal tenons secured with pins, similar to the construction of frames for whenever we encountered original frames. Indeed, x-radiography nearly format these paired images appeared, because the pictures have lost their the wings of triptychs, and may imply that the panels were meant to be always revealed the nails that were used to attach the hinges or corrosion original frames, and their reverses were planed when cradles were applied. hung. The frame for the Memling diptych is more complex: joins at the traces that indicate their earlier presence. In some cases when we did not Yet we surmise that this and other pairings of the Man of Sorrows or Ecce top corners are half-mitered and half-overlapped, without tenons, whereas find signs of hinges, the absence could be explained by a later alteration in Homo and the Mater Dolorosa functioned as pendant paintings rather joins at the bottom are mortise-and-tenon, with vertical tenons secured a frame. In the diptych by Jan Mostaert, for example, now divided between than as folding diptychs, because we know of no fifteenth- or sixteenth- with pins.10 Vertical tenons are typically found in the frames of the central, Enschede and Madrid (cat. 28), only the inner moldings from the inte- century examples with painted reverses. standing section of a triptych and may signify a standing display for the gral frames survive, so it is possible that existing evidence of hinges was Although we assume that most diptychs, unlike pendant paintings, Memling.11 The lack of tenons or pins in the upper joins on the Memling removed when the outer edges of the frames were trimmed off. In other were not on permanent display, it is known from inventories and picto- point to a non-supportive role there as well. cases, however, the absence of hinge marks on original frames could indi- rial sources that some did hang on a wall or at the head of a bed, possibly The different moldings used on the frames for the diptychs by cate that the panels were never part of a folding diptych, especially if the between devotional uses. In addition to the De Rijckere diptych, only one Memling and Provoost may reinforce the suggestion of distinct uses for reverses remained roughly finished and unpainted. other work in our study still has what appear to be its original hanging the two works. The interior frames of the wings of both diptychs have devices: the Lentulus diptych (cat. 29) has metal hangers in the form of a identical profiles: a flat field between concave profiles, with a sloping tr ac e s o f o ri g i n a l d i s pl ay keyhole, with a loop on a triangular base (p. 204, fig. 3). The style and place- bottom molding for the Provoost; and a concave curve between two convex The Aachen Ecce Homo and Mater Dolorosa (cat. 3), attributed to Albrecht ment of the hardware are typical for the period.17 The frames are worn elements for the Memling.12 The exterior frames were treated differently, Bouts, present an unusual situation. Now combined to form the center at the locations of the metal hanging devices, perhaps a testament to its however: there are identical moldings for both halves of the Provoost (flat section of a triptych, this configuration is not original and was most likely repeated use. The Lentulus diptych is relatively uncommon as a type in with a short beveled edge); but there are dissimilar moldings on the two created in the nineteenth century. Based on our technical examination, early Netherlandish painting, however, so we should not try to derive any exterior frames for the Memling (the reverse of the Virgin and Child framed we believe that these two images might have functioned independently general conclusions for other diptychs. in a flat profile with no beveled edges, and the reverse of the donor wing 13 or as pendant images but were probably not attached to one another. We Nearly all of the original frames that we studied showed evi- framed in a flat profile with a beveled interior edge at the bottom). Ver- observed significant disparities in production methods between the paint- dence of some fastening at the center of the top frame member.18 These ougstraete and Van Schoute have suggested that this subtle distinction in paintings changed dramatically. Originally installed as a stationary ings — with respect to their ground layers as well as the technique and style traces — such as drilled holes with straight edges, actual nails still present, the Memling told the user which wing to open, an idea they based on the diptych — with the left panel attached to the wall and the right panel of underdrawing and paint handling — and dendrochronology established or corrosion traces of now-empty nail holes — might indicate hanging strikingly different perspective systems for the interior panels.13 It must be hinged so that it could fold closed over the fixed wing — the two paintings a twenty-five-year difference in the earliest possible felling dates for trees devices that were later removed, but it is impossible to determine if they emphasized that much evidence has been lost over time, and it is not clear now hang separately as pendant portraits. The hinges were removed, and that were used to make the panels. Additionally, cross sections revealed dis- were original. (We did find several examples of clearly modern interven- how much we should infer from so few remaining original frames. Yet we wooden shims were set in their places. Closing devices were also removed, tinct restoration histories for the paintings. It seems prudent to conclude tions, such as screws and industrially produced nails, evincing more recent hope these hypotheses will invite further study. although traces of their use remain visible on the edges of the frames. that the two paintings now framed together in Aachen were not produced means of display.) When the joinery of a frame was designed to accom- The technical examination of original frames can sometimes docu- X-radiographs revealed that the faces of the frames had once been elabo- in tandem. Though it is not impossible that the Mater Dolorosa was painted modate hanging (see cat. 31), with horizontal tenons to increase strength in ment changing tastes and fashions. The wings of the large De Rijckere por- rately decorated with mordant gilded scrollwork, but the frames were later in the Bouts workshop to be paired with an existing Ecce Homo, it that position, we might view traces in the top frame member as a sign of an trait diptych (cat. 32) from 1563 are still housed in their original box frames, seems more likely that the paintings were executed in the same workshop but probably sometime in the seventeenth century the presentation of the

fig. 13 Detail from the reverse of Bernard de Rijckere, Anna van Hertsbeeke and Her Daughter Catharina and Son Jan Baptiste (cat. 32), showing hanging device on top of frame

20 21 e vi d en c e o f wo r k s h o p involvement Around the same time that the diptych format enjoyed a sharp increase in popularity, the open art market was also coming to the fore. In response to the growing demand, painters streamlined their production and mar- keting methods, a development that appears to have benefited diptychs especially. Workshop assistants could produce multiple versions of favorite images based on pattern drawings and stock compositions, without needing much creative input from the master painter. The predominant appearance of the Virgin and Child on the left wing of a devotional half- length portrait diptych thus allowed the shop to work “on spec” for the open market and to finish this wing relatively independently. When a client commissioned a portrait to complete the diptych, the master would become more involved. And after the panels were joined, refinements and revisions could be made to enhance the visual relationship between the two paintings — for example, in the sight lines and gestures of the main figures — usually reflecting the hand of the master. Our technical exami- nations revealed ample evidence for studio participation throughout the stages of production. A Virgin and Child that formed a folding diptych with Joos van der Burch and Saint Simon of Jerusalem (cat. 40), both at the Fogg Art Museum, bears all the hallmarks of workshop involvement. It is one of myriad copies of a detail from Rogier van der Weyden’s famous composition of Saint Luke Draw- ing the Virgin (fig. 14). Many variations of Rogier’s Virgin and Child motif appeared after its introduction in the mid-1430s. Indeed, the Virgin and Child in Cambridge and four other close versions of the composition may have all 14 been produced by studio assistants from a single pattern drawing.19 The left wing of a diptych by the Master of 1499, with the Abbot Christiaan de Hondt as the donor wing (cat. 21), has Jan van Eyck’s famed Virgin original hanging device. By contrast, when the joinery appears to suggest in the Church in Berlin as its ultimate source. Yet several features indicate that a standing display (see cat. 26), with vertical tenons in the bottom joins, we the Master of 1499 based his painting on a linear source, probably a draw- might interpret similar traces to derive from a later alteration. Most origi- ing, not on the original painting by Van Eyck. Intriguingly, when Simon 15 nal hardware on Netherlandish paintings has been lost over time, but the Bening used a detail from the Virgin for an illumination in the Dublin small number of original (or possibly original) hanging devices that we did Rosarium (fig. 15), he adopted the color scheme of the Master of 1499 rather find supports the assumption that most diptychs were probably not hung than Van Eyck,20 and he depicted the Virgin with a relatively narrow crown, dust that passes through the holes in the paper and leaves a series of dots whose workshop is generally believed to have produced these images in but laid on a pillow or stood upright on a table or altar. In these positions, a change that the Master of 1499 introduced very late in the painting pro- on the new support. These dots can then be connected and the excess pow- remarkably large numbers.21 The underdrawings for the Harvard paint- the wings would not be fully opened but would rest at an angle in relation cess (p. 288, fig. 2). der brushed off to create a replica of the pattern drawing on the prepared ings as well as for an Ecce Homo and Mater Dolorosa in Aachen (pp. 44 – 45, to one another, a hypothesis that accords well with other findings from our One way to transfer a model or pattern drawing to a new surface is surface. We found traces of pouncing in the Man of Sorrows and the Mater figs. 3 a,b) do share a general visual language but differ widely in style and technical examinations discussed below. by means of pouncing. A pattern drawing’s main contours are pricked, and Dolorosa at Harvard (pp. 54 – 55, figs. 2 a,b), both attributed to Albrecht Bouts, execution, an indication of workshop participation.22 Interestingly, the the sheet is placed on the intended support. An open-weave bag with fine hands of Christ in these images are sometimes underdrawn differently black powder is then repeatedly pounced on the sheet, releasing powder

fig. 14 fig. 15 Rogier van der Weyden, Saint Simon Bening, Virgin and Child, Luke Drawing the Virgin, c. 1435 – facing a prayer that addresses 1440, panel. Museum of Fine Mary as Queen of Heaven, Arts, Boston, Gift of Mr. and from the Rosarium manuscript Mrs. Henry Lee Higginson (w. 9, fol. 44v – 45r). Chester Beatty Library, Dublin

22 23 than the rest of the panel, suggesting that this detail might have been of a Mater Dolorosa, but she was later transformed into a more regal fig- above. A stone parapet covered with a carpet is depicted over both panels, Female Donor with the Redeemed of the Old Testament (cat. 28), includes what added later to characterize either a Man of Sorrows or an Ecce Homo.23 ure. At the same time, what appears to have been an earlier Holy Face was and the presence of this strong horizontal element, placed parallel to the is probably a posthumous portrait of Mary of Burgundy. The source of Workshop collaborations were observed not only in the under­ transformed into Christ as Savior by the addition of his hands and scepter, bottom edge, virtually excludes the credible display of the wings at an Mostaert’s portrait may have been a secular depiction of Mary in a costume drawings but in the paint layers as well. In the case of the Harvard Man while indications of cropping at the left side of the panel suggest that angle. Moreover, Sittow is known to have frequently repainted the eyes with a rather deep décolleté, but the diptych is thought to have been com- of Sorrows and Mater Dolorosa, the same master, a superbly skilled craftsman, it was trimmed to match the size of the Virgin at Prayer so that they could of his figures.28 missioned by Margaret of Austria, Mary’s daughter, and at late stage in must have painted the flesh tones in both panels, considering their con- function as a diptych. painting Mary was given a higher neckline more befitting her piety, per- sistent painting technique and high-quality execution (pp. 54 – 55, figs. 1 A separate category of changes, all initiated very late in the pro- s i g n s o f d o n o r i n flu en c e haps at Margaret’s request. and 3). One or both of the gilded backgrounds (which would have been duction process, seem to have served to correct or improve the patterns The individuals who commissioned diptychs apparently also instigated Donor portraits were sometimes completely replaced. In the completed before the figures were started) might have been done by an of communication between the figures on the two wings. These adjust- changes, some of which were quite dramatic. On the right wing of diptych of the Virgin and Child with Three Donors by the Master of the Saint assistant, for the organization of the fields of dots differs dramatically ments appear to be directly related to the angle at which the panels were Provoost’s diptych pairing Christ Carrying the Cross with the Portrait of a Fifty- Ursula Legend, the young woman at the far right was underdrawn and between the two works; clearly they were the work of two different hands. meant to be viewed,27 for we encountered them primarily in standing four-year-old Franciscan in Bruges (cat. 31), the monk was initially depicted in painted on top of a different donor whose image was first painted out Our examination of The Trinity by the Master of the Lille Adoration (cat. 22), diptychs, which would necessarily have the wings set an angle to achieve a fully furnished room. It seems likely that it was the donor himself who (p. 168, figs. 4 a,b). The reasons for this radical change remain unknown, which we believe related to the Saint Jerome as a pendant rather than part a stable position, and in diptychs that were presumably placed on a pillow requested that the interior setting be painted out, perhaps in keeping but in other cases it has been possible to gather information that sheds of a folding diptych, also revealed workshop collaboration in the under- as a prayer book or held in the hands, such as Jan van Eyck’s Annunciation with his vows of poverty, perhaps to increase the visual focus on Christ’s light on similar alterations. For instance, in the devotional half-length drawings and the handling of both the background and the figures of the diptych in Madrid (cat. 8). Throughout the process of painting these pan- Passion. At an even later stage the monk’s head was enlarged so that it portrait diptych of Joos van der Burch at the Fogg (cat. 40) the donor’s like- putti. Because multiple pairings of the Trinity and Saint Jerome apparently els, Van Eyck appears to have made minute refinements that influence the partly overlapped the new background. Infrared reflectography also ness appears to have replaced that of his son Simon at some point after Joos’ existed, it seems that the imagery was in demand, and it may have been angle at which the wings were to be viewed as well as the proximity of the revealed an early change in the underdrawing: the wrists of the donor death in 1496. The function of the picture also seems to have been dra- worthwhile to produce the works in quantity. viewer to the panel surfaces. were originally bound together with a rope that appears to have led to the matically altered, with Simon van der Burch’s name saint painted out and Striking stylistic distinctions between the wings of the diptych In some devotional portrait diptychs it appears that the panel with bound wrists of Christ on the other panel. The rope binding Christ’s hands his coats of arms on the reverse replaced by an epitaph for Joos and his wife. of the Virgin and Child and Three Donors by the Master of the Saint Ursula the Virgin and Child was to be seen frontally, while the donor panel was was painted as planned, but this element was abandoned on the portrait The diptych was mounted over the couple’s grave in the Church of Saint Legend (cat. 24) have suggested that these panels were painted by different at an angle. This is suggested by the position of the figures relative to wing, most likely reflecting the donor’s wishes. An important part of the Walburga in Furness, in which Simon was later interred as well.29 hands,24 and the technical examination confirmed distinctive working the viewer. Additionally, the relationship of the donors to the Virgin was artist’s conception — and seemingly related to the rebus on the frames In a variation on this theme a second donor portrait was added methods for the two panels. Multiple pattern drawings might have been carefully described. In the diptych of the Virgin and Child with Three Donors above the images, which reads “Franciscan cords carry (or draw) the most to the diptych of the Virgin in the Church with Abbot Christiaan de Hondt by used to create the Virgin and Child,25 and the resulting stock image, with its by the Master of the Saint Ursula Legend, for instance, we found a signifi- hearts” — this detail may have created an impression that did not concur the Master of 1499 (cat. 21). Originally commissioned by De Hondt, who fully frontal composition and the absence of sight lines between the fig- cant change in the sight lines between the main figures. The Virgin was with the monk’s vows of modesty. was the abbot of the Cistercian abbey of Ter Duinen between 1495 and ures, might have functioned as an autonomous image that was later con- initially painted looking to the right, toward the donors, but her gaze was We encountered another major revision probably prompted by the 1509, the reverse of the abbot’s portrait was painted with a red and black verted into the left wing of a diptych. later directed down toward the Christ child (p. 168, fig. 3). The final image donor in Hans Memling’s diptych of the Virgin and Child with Maarten van porphyry imitation. But this surface was later adorned with the portrait A small number of panels originally represented a different subject creates a convincing circular pattern of communication, with the donors Nieuwenhove (cat. 26). The arched stained-glass window at the top left of the of Robrecht de Clercq, the abbot of Ter Duinen between 1519 and 1557. altogether, and it appears that the images were changed by the painter to addressing their prayers not directly to the Christ child but to the Virgin, Virgin and Child had initially been rectangular with clear glass and a vertical De Hondt was directing his prayers to the Virgin in the left-hand panel, facilitate their being paired with other images. The initial underdrawing who looks down at the Child, who in turn looks at the donors. This change and horizontal division, like the window to the right, but was changed and De Clercq now directed his prayers to the Salvator Mundi on the panel for the Mater Dolorosa at Harvard, for instance, described a centrally placed occurred late in production, perhaps only after the panels were joined. to display the coats of arms of the Van Nieuwenhove family. The back- to the right (the reverse of the wing portraying the Virgin). At the same head, in a frontal position — possibly the face of Christ or a Holy Face in One obviously needs to be careful in interpreting such changes. ground landscape visible at the right had initially continued through both time the portrait of De Clercq was added, the trompe-l’oeil stone frame the Byzantine tradition (p. 55, fig. 2 b). This face was not taken into the We also observed, for example, several adjustments to the sight lines in windows, as in a Virgin and Child by Memling that once functioned as the was copied from the Salvator Mundi, effectively creating a double diptych. paint stage; instead, the artist simply superimposed the underdrawing and the diptych of the Virgin and Child and the portrait of Diego de Guevara(?) by center panel of his Triptych of Benedetto Portinari (see p. 184, figs. 3 a,b, 4). It is A small but telling change was also revealed at the lower left corner of the painting of the Mater Dolorosa and paired it with the Man of Sorrows. Another Michel Sittow, now divided between Berlin and Washington (cat. 34). Yet conceivable that the coats of arms were first planned for the reverse of the Salvator Mundi, where De Clercq had the coat of arms of the abbey replaced example of the conversion of existing images into the wings of a diptych these revisions are most likely not related to the issues of display described portrait of Maarten van Nieuwenhove but that the donor later favored the by the arms of his own family (see p. 141 and p. 148, fig. 7). is Quentin Massys’ Virgin at Prayer and Christ as Savior in Antwerp (cat. 15).26 present, more prominent placement. In this case an earlier stage showed the Virgin wearing the somber garb Some changes are less obvious but still suggestive of a donor’s influence over the final composition. The right wing of a diptych by Jan Mostaert, representing Christ Appearing to His Mother in Limbo and a Kneeling

24 25 16a 16b 17 18

Occasionally our examinations established that panels long assumed ery on the reverses of diptych wings.31 These panels were probably part of a interior wings of a triptych rather than a diptych. This finding accords The specific findings on the individual works in the exhibition are to be the wings of a diptych actually functioned in a different format. now-disassembled triptych. It has also been suggested that the Saint George with their relatively tall and narrow dimensions. Our technical examina- presented in greater detail in the following catalogue entries, with mate- Bernard van Orley’s Christ Among the Doctors and the Marriage of the Virgin at and the Dragon at the National Gallery of Art in Washington and the Virgin tion of a diptych of the Virgin and Child with Willem van Bibaut from a private rial and technical aspects integrated into the art historical discussions. the National Gallery of Art in Washington (figs. 16 a,b) are presented as a and Child at the Museo Thyssen-Bornemisza in Madrid by Rogier van der collection (cat. 23) confirmed that these two panels were originally wings Although technical examination provided a wealth of new information diptych in hinged, modern frames.30 Yet our examination of the barbes Weyden (cat. 36) formed a diptych.32 It has now been determined, however, of a diptych — but not the same one. The attribution of the Virgin and Child on these engaging paintings, we must be cautious in formulating general and unpainted edges of these panels made clear that the panels originally that these two images are most likely the separated front and back of the to the Master of the Magdalen Legend might well be correct, but the por- conclusions from our data. Much of the material evidence for Netherland- had a frame that was shaped as an inverted t (fig. 17), ruling out an original same panel, which has a vertical split beginning at the center of the lower trait of the Carthusian abbot was probably painted by a French master, per- ish diptychs has been lost over time, and numerous paintings have not yet pairing as a folding diptych. Furthermore, the reverse of Christ Among the edge that corresponds exactly in the two paintings.33 haps from the Grenoble /Avignon area. At an unknown date the two panels been thoroughly examined using technical means. It is our hope that the Doctors depicts a Putto with the Arms of Jacques Coëne (fig. 18) as if seen from the The two panels of Saint Peter and Saint Paul attributed to the Master were made into a new diptych. present study will inspire more research into the subject. left, in contrast to the frontal perspective system commonly used for imag- of the Female Half-Lengths (cat. 19) are hinged as a diptych, but our tech- nical examination led us to conclude that these works were originally the

figs. 16a,b fig. 17 fig. 18 Bernard van Orley, Christ Detail from lower right corner Bernard van Orley, Putto with Among the Doctors and The of Van Orley’s Marriage of the the Arms of Jacques Coëne Marriage of the Virgin, c. 1513, Virgin (fig. 16b), showing an (reverse of fig. 16a) panel. National Gallery of Art, irregular barbe Washington, Samuel H. Kress Collection

26 27