Rembrandt's Portrait(S)

National GalleryTechnical Bulletin volume 31

National GalleryCompany London

Distributed by Yale University Press This volume of the Technical Bulletin has been funded by the AmericanFriends of the National Gallery, London with agenerousdonation from MrsCharles Wrightsman

Series editor Ashok Roy Photographic credits

All photographs reproduced in this Bulletin are ©The National Gallery, London unless credited otherwise below. ©National GalleryCompanyLimited 2010 AMSTERDAM ©Rijksmuseum, Amsterdam: fig.12, p. 103

All rights reserved. No partofthis publication maybe BERLIN Kupferstichkabinett ©Photo Scala, Florence /BPK, transmitted in anyformorbyany means, electronic or Bildargentur für Kunst, Kultur und Geschichte,Berlin: fig.34, p. 21 mechanical, including photocopy, recording,orany CAMBRIDGE ©Fitzwilliam Museum, Cambridge: fig.51, p. 31 information storageand retrieval system, without the prior permissioninwriting of the publisher. CANTERBURYRoyal Museum and ArtGallery©The National Gallery, London /courtesy of RoyalMuseum and ArtGallery, Articles published online on the National Gallerywebsite Canterbury: fig.3,p.79; fig.17, p. 85; fig.20, p. 86; fig.21, p. 86; maybedownloaded for private studyonly. fig.28, p. 88; fig.31, p. 91; fig.33, p. 92 CAVA DE’ TIRRENI Badia della SS.Trinità photo ©Antonio Biasio: First published in GreatBritain in 2010 by fig.29, p. 90; fig.30, p. 90; fig.32, p. 91; fig.34, p. 92 National GalleryCompanyLimited EDINBURGH The National GalleryofScotland, Edinburgh St Vincent House,30OrangeStreet ©National Galleries of Scotland: fig.52, p. 31 London WC2H 7HH FLORENCE Galleria degli Uffizi, Florence,Gabinetto dei Disegni edelle www.nationalgallery. co.uk Stampe ©Soprintendenza Speciale per il Polo Museale Fiorentino, Gabinetto Fotografico,Ministero per iBeni eAttività Culturali: fig.49, British LibraryCataloguing in Publication Data p. 29; fig.54, p. 32; fig.55, p. 34; Museo del Bargello ©Photo Scala, Acatalogue record for this journal is available from Florence –courtesy of the Ministero per iBeni eAttività Culturali: the British Library fig.38, p. 24; fig.39, p. 25; San Lorenzo©DeAgostini Picture Library/Scala, Florence: fig.6,p.10. ISBN 978 185709 495 4 KASSEL ©Staatliche Museen Kassel: fig.17, p. 107 ISSN 0140 7430 1018117 LONDON ©The Trustees of The British Museum: fig.53, p. 31; Victoria and AlbertMuseum, London Project managerJan Green ©V&A Images/Victoria and AlbertMuseum, London:fig. 41, p. 25 Editor Rebecca McKie OXFORD Christ Church, Oxford ©Bypermission of the governing Designer Libanus Press body, Christ Church, Oxford: fig.33, p. 21 Picture research Karolina Majewska and Suzanne Bosman PA SADENA, CALIFORNIA ©The Norton Simon Foundation, Production Jane Hyne and PennyLeTissier Pasadena, California: fig.4,p.100 Repro by Alta Image, London Printed in Hong Kong by Printing Express PISTOIA Duomo ©Photo Scala, Florence: fig.40, p. 25 PITTBURGH, PENNSYLVA NIA ©Carnegie Museum of Art, Pittsburgh: fig.3,p.66 FRONT COVER ROUEN ©Musée des Beaux-Arts de Rouen: fig.1,p.78; fig.35, p. 95 Andrea del Verrocchio, TheVirgin and Child with TwoAngels, NG 296, detail of fig.18, page 16

TITLE PA GE Andrea del Verrocchio, TheVirgin and Child with TwoAngels, NG 296, photomicrographs (see page 17 for details) Rembrandt’sPortrait(s?) of Frederik Rihel

marjorie e. wieseman

In 2008, afull X-radiographwas recorded of concluded. He became director of the familyfirm after Rembrandt’s Portrait of Frederik Rihel on Horseback Guillelmo’sdeath in 1658, and wasappointed guardian

(NG 6300) for the first time (FIGS 1and 2). The X- of the young Bartolotti children after the death of radiographassemblyrevealed another,apparently Guillelmo’swidow, Jacoba, in 1664.4 Rihel also unrelated composition beneath the grand equestrian managedtwo profitable businesses of his own, one portrait nowvisible: afull-length figure of aman, in partnership with Guillelmo’sson. Documentary orientated with the canvasturned 90 degrees anticlock- evidence suggests thathewas an ardent horseman.5 wise.Heholds astaff(?) in his right hand, and appears He never married, and apparentlynever owned ahouse to be standing in alandscape with some trees to the left. in Amsterdam. He did, however, become acitizen of the The discovery wasquite unexpected, as the visible image city on 5April 1662, very likelyinorder to fulfil the gave no evidence of masking an earlier design, and prerequisite to becoming amember of the city’scivic the painting has been well studied since its acquisition guard. In 1677, aged about 56, Rihel wasappointed by the National Galleryin1960.1 The radical difference to the prestigious post of ensign, or standard bearer between the twocompositions suggested the pragmatic (vaandrig), in the militiafor aperiod of one year.Hedied reuse of the largecanvassupport. This practice wasnot in Amsterdam on 6January1681. uncommon in the Rembrandt studio: several of his The commission to portray Rihel might have come works are known to have been painted on canvases or about through Rembrandt’slong-standing association panels thathad been used previously, including roughly with the Trips,another of the leadingmerchant- aquarter of his self portraits.2 However, the National industrialist families in the Netherlands. Rembrandt had Gallery’spainting is by farthe largest example painted twomembersofthe Trip familyin1639 (Aletta discovered to date,and appearstobethe onlyinstance Adriaensdr.and Maria Trip,widowand daughter in which Rembrandt (or anymember of his studio) respectively of Elias Trip),6 and in 1661 he painted the adopted this practice for acommissioned portrait. statelyportraits of the aged Jacob Trip (Elias’s brother) The subject of the picture,Rembrandt’sonly and his wifeMargaretha de Geer which are nowalso equestrian portrait, wasfirstidentified as the Amsterdam in the National Gallery(NG 1674 and NG 1675). The merchant Frederik Rihel by the noted Rembrandt expert portraits of Trip and de Geer were probablyordered and collector Abraham Bredius (1855–1946), based by their son Hendrick Trip to adornhis magnificent on an item in the inventoryofRihel’sproperty drawn new house on the Kloveniersburgwal in Amsterdam, up after his death in 1681: ‘Het conterfijtselvan de completed in 1662. Trip’sclose neighboursonthe overledene te paertdoor Rembrandt’ (‘The portrait of Kloveniersburgwal were Rihel’spatrons and business the deceased on horseback by Rembrandt’).3 Rihel was associates the Bartolottis, whoalso had extensive borninStrasbourg in 1621. His father,apaper manu- business ties with the Trips. Although no specific and facturer,sent him to Amsterdam in 1642 for afive-year direct links between Rihel and the Trips have been termofemployment with Guillelmo Bartolotti, one of established, he must certainlyhave known them. In the most powerful merchants in Amsterdam. Rihel this context, it is worthnoting thatseven yearsafter retained close business and personal ties with the the forced dissolution of the Trip-de Geer monopoly Bartolotti familylong after his apprenticeship was on the importofarmsfrom Sweden in 1662, this

96 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

FIG.1Rembrandt, Portrait of FrederickRihel on Horseback (NG 6300), probably1663. Oil on canvas, 294.5 × 241 cm.

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 97 Marjorie E. Wieseman

FIG.2NG 6300, X-radiographmosaic.The composite image has been digitallyprocessed to reduce the effects of the stretcher bars. lucrativetrade wasquietlytaken over by the Rihel- Netherlands in the seventeenthcentury. Bartolotti firm.7 The Portrait of Frederik Rihel on Horseback is painted It is plausiblyassumed thatRembrandt’sportrait, on three pieces of canvassewn together,each with dated 1663, wasmade to commemorateRihel’spartici- apparentlythe same or similar weave characteristics.9 pation in the honourguard thatprovided aceremonial The upper seam runs horizontallythrough the horse’s escortfor MaryStuartand the young Prince William III bit and the pommel of the sword just belowRihel’s of Orangeontheir entryinto Amsterdam on 15 June elbow; the lowerseam runs horizontallyfrom between 1660. The architecture dimlyvisible on the left has been the bottom of the horse’sright front hoof and the top of identified as the Heiligewegspoort(since destroyed); its left front hoof.The canvasisprepared with adark the ridersinthe coach at left the prestigious visitors.8 brown ‘quartz’-type ground composed of silica with a It is ahighlyunusual portrait for Rembrandt to have little brown ochre,similar to thatused in Rembrandt’s painted: not onlyisitone of the very fewofhis paintings Portrait of Hendrickje Stoffels of about 1654–6 which makereference to acontemporaryevent, but it (NG 6432) and Self Portrait at the Age of 63 of 1669 is also his onlyequestrian portrait –indeed one of just (NG 221). Rembrandt began using coarselytextured twolife-sized, or near life-sized, equestrian portraits of ‘quartz’ grounds after about 1640, and this method of ordinarycitizens (thatis, not stadholdersormembers preparation seems to have been unique to him and the of the nobility) known to have been painted in the artists in his immediate circle.10 It is not clear whether

98 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

FIG.3NG 6300, infrared photograph.

the choice wasmade for economic reasons (the ‘quartz’ dark paint can be seen, for example,inthe torso of grounds used cheaper materials than the more common Rembrandt’sunfinished Portrait of aBoy of about double grounds) or aesthetic ones (the rougher surface 1655–60 in the Norton Simon Museum of Art, of the ‘quartz’ grounds mayhave provided better tooth Pasadena (FIG.4), or the right armofthe Man with a for building up thick layers of impasto). Magnifying Glass in the Metropolitan Museum of Art, a Infrared photographs taken during cleaning and work of the early1660s.11 The infrared photograph restoration of the painting in 1960 showthatthe figure shows several changes made as the equestrian portrait and horse,aswell as details in the foreground and evolved: most notably, Rihel’shat wasoriginallytaller background of the composition, were boldlyunder- in the crown and wider at the brim; this alteration is drawnwith broad strokes of blackish paint applied faintlyvisible on the surface of the painting in available with abrush (FIG.3). Similar preliminarysketches in light. The infrared image also discloses changes to the

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 99 Marjorie E. Wieseman outline of the horse’shead and the position of its legs, bar (showing light in the earlier X-ray image)largely and reveals features in the background of the painting – obscured the points of the collar thatmight have provid- such as the carriage with figures to the left of the ed an suggestive identifying ‘context’ for the marks. mounted figure,the architecture beyond, and the Several paint samples were taken from the painting additional rider(s) to the right –with greater clarity than in 1975 and 1988. Rembrandt used arather more is nowpossible in available light, because the paint varied palette than wascustomary, undoubtedly layers have become more transparent over time and the because the scale and subject matter of agrand eques- dark brown ground more prominent. trian portrait required agreater rangeofcoloursthan Before 2008, X-radiographs had onlybeen made of his usual compositions with an indoor setting.One of afew discrete areas of the painting,such as the head of the more unusual pigments identified is brazilwood lake, the rider and areas aroundthe horse’sforelegsand the found in the underlayers belowthe rider’scoat(see FIG. rider’sstirrup. The selection of these areas for imaging 7). Smalt is used extensively throughout the painting, wasprobablyintended to clarify or amplify the changes for both colouristic and textural effect. The grey-blue of evident in the infrared photograph. Yetsome oddfea- the sky, for example,isconstructed of twounderlayers tures, clearlyvisible in these X-ray images, seem to have of smalt mixed with lead white (the lowerincorporating gone unremarked. In the detail X-radiographofthe stir- alittle red lake), glazed with smalt; for the dull green ruptaken in 1960 (FIG.5), light horizontal streaks with foliage, smalt is mixed with yellowpigments. More aknobbled terminus, unrelated to the visible image, coarselyground particles of smalt are used to ‘bulk up’ extend from the right edgeofthe image. In the context the paint and build texture in areas of heavy impasto.It of the full X-ray image, we nowunderstand these marks is also likelythatRembrandt incorporated smalt into as corresponding to the tasselled ties of the underlying paint mixtures and glazes for its siccative properties. figure’scollar,rendered with swift strokes of X-ray- When viewed in cross-section, some of the samples absorbent paint. Yetthere is no evidence thatthese appeared particularlycomplex. In asample taken from curiouslywell-defined marks were investigated further the lowerright corner of the yellowskirtofthe rider’s in 1960 or subsequently, perhaps because the stretcher coat, atranslucent blue-greylayer containing ahigh degree of smalt wasobservedbeneath the layerof

opaque yellowpaint (FIG.6); it wasconjecturedthatthe projecting edgeofthe skirthad been painted over the blue-greyofthe background sky. 12 Another sample, taken from the fold in the skirtjust belowthe rider’sleft wrist, wasfound to have dark underlayers containing

smalt and red lake(FIG.7), whilst beneath the light grey of the horse’schest is athick layeroffaintlygreenish dark-brown paint with asecond dull brown-green layer

below(FIG.8). The complexlayer structureofthese and other samples taken from the central portion of the canvas, with lowerlayersquite different from the layers directlyrelated to the visible design, would seem to sug- gest asuperimposition of forms deriving from composi- tional changes or other modifications made by the artist. Without asense of an overall design,however,there was no way to connect these isolated pieces of information. More definitive evidence of amodification to the design wasdiscovered in the area of the skyatupper right. Asample taken from near the right edgeofthe FIG.4Rembrandt, Portrait of aBoy,1655–60. Oil on canvas, canvasatthe level of the brim of the rider’shat consist- 64.8 × 55.9 cm. Pasadena, California, The Norton Simon Foundation, inv. F. 1965.2.P. ed of athick layerofpaint containing smalt and lead

100 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

FIG.5 NG 6300, detail X-radiograph taken in 1960, showing Rihel’s stirrupand traces of an underlying image at right.

FIG.6 NG 6300, paint cross-section from the lowerright corner FIG.7 NG 6300, paint cross-section from the fold in the skirtjust of the yellowskirtofthe rider’scoat. Atranslucent blue-grey belowthe rider’sleft wrist. The dark underlayers contain smalt layercontaining ahigh degree of smalt can be seen belowalayer and red lake(incorporating cochineal and brazilwood dyestuffs) of opaque yellowpaint consisting mainlyofyellowochre and with other pigments.The brown ‘quartz’ ground is visible at the lead white. base of the sample.

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 101 Marjorie E. Wieseman

FIG.8 NG 6300, paint cross-section from the horse’schest. Beneath the light grey layerisathick layerofdark greenish- brown paint containing primarilysmalt, earth pigments and black, with asecond dull brown-green layerbelow. white,glazed over with smalt, which concealed two earlier layers of green and yellow-green paint (FIG.9). This suggested thatRembrandt had first laid in thickly painted foliage in this area, then obliterated it with a layerofsky,also thicklypainted. In the context of the visible design, the green underlayerswere accounted for by supposing amore denselyfoliated background to the equestrian portrait, which Rembrandt subsequently ‘lightened’ with the introduction of apatch of smoky grey-blue sky. The discovery via full X-radiographassemblyofan earlier composition beneath the Portrait of Frederik Rihel on Horseback allows some of the puzzling details encoun- tered in isolated X-radiographs and paint cross-sections to become partofamore coherent picture.Yet many questions remain,particularly as regards Rembrandt’s apparentlyanomalous use of apreviouslyused support

FIG.10NG6300, X-radiographdetail showing the whole figure.

for what wassurelyanimportant portrait commission. The X-radiographassemblyshows aman standing at full length, positioned just to the right of the centre of

the canvasinits horizontal orientation (FIG.2).13 The figure is posed frontally, with his proper left foot slightly forward of his right and his head turned three-quarters FIG.9 NG 6300, paint cross-section from greyish-blue skyat to the left; he holds astick or staffinhis right hand and upper right, taken from near the right edgeofthe canvasatthe level of the rider’shat-brim. Several layers of smalt and and lead his left hand is propped on his hip (FIGS 10 and 11). The white,with afinal glazeofsmalt, conceal twoearlier layers of features of his garments visible in the X-ray image yellow-green foliage paint containing yellowochre,smalt and suggest thatheiswearing ridingcostume,consisting of azurite.The lowermost layerofthe skyincorporates alittle red lakepigment. asleeveless thigh-length kazak,orriding coat, worn over

102 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

FIG.11NG6300, X-radiographdetail of the head.

FIG.12School of Thomas de Keyser, Equestrian Portrait of Johan Wolfert van Brederode, c.1650. Oil on canvas, 104.5 × 90 cm. Amsterdam, Rijksmuseum, inv. SK-A-3972.

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 103 Marjorie E. Wieseman

FIG.13NG6300, X-radiographdetail of lowerright of the underlying image.

adoublet with sleeves decorated with light bands Other elements of the underlying composition are less (probablyindicating gold braid) and full knee-length legible,however,and maynot have been as ‘finished’ breeches. An extension at the outer contour of his when the painting wasset aside.Broad undulating proper left calf suggests adecorativeribbongarter or contoursappearing light and dark in the X-radiograph flared boot-top.The relatively small flatcollar and the extend from the left side of the composition, suggesting overall silhouette of the figure suggest adateofabout alandscape setting.Atthe left are twotrees, with what 1650–5; averysimilar costume is depicted in an appearstobeathird tree –orthe stump of one –just to equestrian portrait of Johan Wolfertvan Brederode, the right. Although ashadowy extension of the ‘stump’ painted in about 1650 (FIG.12).14 runs verticallyapproximatelytothe level of Rihel’s The clarity of the figure in the X-radiographsug- shoulder,abroad continuous strokeofthe light paint gests thatthe abandoned first composition wasneither thatdefines the landscape contour passes over this painted over nor scraped out before Rembrandt painted extension.The horizontal landscape contoursappear to the equestrian portrait of Rihel over it. The figure seems continue to the right of the figure; the dark vertical mass fairlywell developed, constructed with broad, vigorous to the right of his elbowisananomalyresulting from brushstrokes not unlikethose used to provisionally the matching of individual X-radiographs in the assem- describe forms in other late paintings by Rembrandt. bly. An area of light radioabsorbent paint surrounds

104 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

FIG.14NG6300, X-radiographdetail of the centre of the lowerhalf of the painting,tothe left of the figure.

the lowerhalf of the figure,possiblyapplied to help figure,suggesting an area left in reserve.Possiblyrelated distinguish it from the surrounding landscape setting. to this elusive formare several long,sweeping strokes At lowerright is alight-coloured shape,possiblycylin- of radioabsorbent paint near the loweredgeofthe drical, with narrowdark horizontal bands. In the dark composition, including ahorizontal strokeextending area just to the left of this form(belowthe left foreleg from near the mouth of Rihel’s horse,and twomore of Rihel’shorse,clearlyvisible in the X-radiograph) is a vertical marks thatcross, then angle back. distinct but unidentifiable motif in light radioabsorbent Guided by the image revealed in the X-radiograph paint (FIG.13). Reading the shadowy formtothe left of assembly, in 2009 additional paint samples were taken the figure is complicated by the superimposition of the with twospecific goals: to determine whether the com- head, chest and forelegsofRihel’shorse in this area position of the ground layerwas consistent throughout

(FIG.14). The positioning of the standing figure slightly all three pieces of canvas; and to obtain as much to the right of centre would suggest the inclusion of a additional information as possible about the underlying fairlysubstantial compositional element here.The light composition. Where the layerstructure of the sample paint of the background landscape stops shortofthe wascomplete,asingle brown ‘quartz’-type ground was horizontal stretcherbar (just belowthe nose of Rihel’s present in each of the three canvaspieces (FIG.15). This horse), and continues again to the left of the standing would indicate thatthe entire canvasassemblage was

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 105 Marjorie E. Wieseman

containing smalt and lead white (FIG.9)undoubtedly relate to foliage from these trees. As noted, the painting over which Rembrandt painted the Portrait of Frederik Rihel on Horseback seems to have been fairlywell advanced but in all likelihood uncompleted. The bold brushwork used to describe the figure,most apparent in areas of radioabsorbent white paint, mayrepresent an initial laying-in of forms, although such details as the dangling collar ties would suggest agreater degree of finish. None of the samples FIG.15NG6300, paint cross-section from the foreground, lower left corner.The lowest layeristhe single brown ‘quartz’ ground; studied in cross-section gave evidence of containing an over thatare three layers of paint: athin warm brown consisting intermediaryvarnish layer, which might suggest that of earth pigments and some red lake; agreenish brown layer comprising earths, red lake, azurite and smalt; and athick upper the underlying painting wasfullycompleted when the layerofalighter greenish brown, similarincomposition to that portrait of Rihel wasbegun. Nor wasevidence found in immediatelybelow, containing much coarse smalt. anyofthe cross-sections analysed to indicate that Rembrandt applied an intermediarypaint layeroverall almost certainlyprimed at the same time and, if it is or partofthe first painting before embarking on the acceptedthatthe ‘quartz’ ground is unique to equestrian portrait. It would therefore appear that Rembrandt’sstudio,thatthe entire prepared canvas the underdrawn sketch in black paint for the equestrian must have originated there.15 It seems reasonable to portrait, which can be seen in the infrared photo- assume,therefore,thatthe underlying painting of a graph(FIG.3), wasdone directlyonthe earlier painting. man standing in alandscape also originated in Farfrom obliterating the original image, in fact, Rembrandt’sstudio,and thatits dimensions were Rembrandt appearstohave salvaged afragment of the roughlyequivalent to those of the present canvas. underlying figure –the gold bands of the sleeveand Unfortunately, attempts to learnmore about the doublet at his right shoulder –and incorporated it compositionbeneath thevisible image –specifically thefigure andthe shadowyforms flanking him –by cross-section analysis were less successful, as the condi- tion of the paint surface yielded no appropriate areas for sampling in the immediate area of the underlying figure.But with the X-radiographassemblyasaguide,it waspossible to reassess samples taken close to this area in earlier campaigns, paying particular attention to the lowerlayersofthe cross-sections possiblycorresponding to the earlier design. The translucent blue-greylayer observedbeneath the thick layerofyellowinthe sample taken from the skirtofRihel’scoat(FIG.6)probably relates to apatch of skytothe left of the standing figure.

The sample from the horse’schest (FIG.8)was taken close to the vertical of the staffheld by the standing figure,and the dull greenish-brown lowerlayersmay relate to this or to the rather amorphous larger form, which might (verytentatively)beread as amass of foliage. The sample from the skyatthe upper right of the equestrian portrait wastaken from the area between the twotree trunks; the twothick layers of green and yellow-green paint visible beneath the lighter layer FIG.16NG6300, detail showing decorativemotif on Rihel’sboot.

106 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

as afanciful decorativemotif on Rihel’sboot (FIG.16). It is riskytoventure an attribution for the under- lying painting based solelyonthe style and quality of the X-ray image. Allowing for the inherent differences between X-radiographs of finished and unfinished paintings, the free and assured laying-in of the under- lying figure nonetheless seems reasonablycompatible with X-radiographs of paintings by Rembrandt him- self.16 Dry, scumbled strokes, such as those delineating the collar ties, find parallel in several presumably unfinished paintings by Rembrandt from the 1650s and

60s, such as the Portrait of aBoy in Pasadena (FIG.4), or the slightlymore finished Portrait of an Elderly Man of 1667 in the Mauritshuis. If the underlying painting is by Rembrandt, the representation of afull-length, near life-sized figure in an expansive landscape setting would appear to be unique in his oeuvre.Onlytwo examples of individual full-length standing portraits of men by Rembrandt are known, both from the 1630s (Portrait of Maerten Soolmans,dated 1634, and Portrait of Andries de

Graeff,dated 1639 [FIG.17]),17 and after the completion of TheNight Watch in 1642, he did not paint another portrait of afull-length standing figure.Ifthe under- lying painting is not by Rembrandt, apossible author might be sought among the pupils active in Rembrandt’s studio in the yearsinwhich (based on the costume) the FIG.17Rembrandt, Portrait of Andries de Graeff,1639. Oil on underlying portrait waspresumablypainted. Abraham canvas, 200 × 124.2 cm. Kassel, Staatliche Museen, inv. GK 239. vanDijck, Willem Drost, Heyman Dullaert, Jacobus Leveck, Nicolaes Maes and Constantijn vanRenesse simplydoes not provide enough information for any were all working with Rembrandt during the earlyto useful formoffacial recognition.19 mid 1650s, but none is known to have painted this type Whether the underlying painting wasby of large-scale portrait composition.18 It is unlikely, Rembrandt or by amember of his studio,itwould seem moreover, thatsuch alargecanvas(and presumably thatany work left abandoned or unfinished in the prestigious commission) would have been entrusted to a Rembrandt atelier wasaviable target for reclamation. pupil or assistant, no matter howadvanced. In addition, Approximatelytwo dozenworks from the Rembrandt Rembrandt’sdecision to incorporateasnippet of the studio have been identified as having been painted over underlying figure into the new composition suggests areused support; for convenience these have been adegree of proprietarypragmatism thatlends weight termed ‘palimpsests’.20 Just over half (thirteen) of the to the argument in favour of his also having painted ‘palimpsests’are self portraits, representing more than a the earlier work. quarter of the total number of portraits of Rembrandt If proposing an attribution for the underlying painted either by himself or by pupils or workshop painting based solelyonthe appearance of an X-ray assistants.21 ‘Palimpsests’ are farrarer in the case of image is risky, attempting an identification of the figure Rembrandt’s‘tronies’and subject pieces; most of these itself would be downright foolhardy. The man appears concernsmall works from either the artist’sLeiden to have amoustache similar to thataffected by Rihel, period, or his earlyyearsinAmsterdam.22 Apartfrom and possiblyasmall goatee (which Rihel apparently afew self portraits, onlythree ‘palimpsests’from the did not, in 1663), but beyond thatthe X-radiograph 1650s (and none from the 1660s)have hitherto been

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 107 Marjorie E. Wieseman recorded: the version of Joseph accused by Potiphar’s Wetering concluded thatmost of the ‘palimpsests’ Wife in the Staatliche Museum, Berlin, dated 1655, is issuing from the Rembrandt studio concerned paintings painted over the unfinished figure of an old man seated; thatwere not made on commission,25 although this the Man in Armour in Glasgow,dated 1655, is painted did not necessarilymean thatRembrandt did not intend over afigure of an elderlybearded man, perhaps partof to sell these works. Indeed, one of his ‘palimpsest’ an unfinished composition; and a Portrait of aYoung paintings wasowned by Frederik Hendrik, stadholder Man in the Louvre, (falsely) signed and dated 1658, of the Netherlands, and twoothers, givenbythe stad- is painted over ascene of awoman bending over a holder to the Earl of Ancram, eventuallyfound their cradle.23 Another example of a‘palimpsest’ by afollower way to the collectionofCharles Ibefore 1639.26 Van of Rembrandt is the National Gallery’s Seated Man with de Wetering also observedthatsofar as wasknown, aStick of about 1650–60 (NG 51), which is painted there wasnot asingle case of aportrait made over a over adepictionofthe crucified Christ(FIG.18).24 rejected painting,27 although the subsequent discovery Evaluating the known examples, Ernst vande of another painting beneath the Portrait of Frederik

FIG.18FollowerofRembrandt, Seated Man with aStick (NG 51), c.1650–60, X-radiographmosaic showing underlying painting of crucified Christ.

108 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

Rihel on Horseback would appear to contradict this. part.33 If,however,the underlying figure does represent There seems to have been no consistent patternto Rihel, his risingcivic status in the earlyyearsofthe Rembrandt’stechnique in approaching a‘palimpsest’. 1660s maywell have prompted artist or patron to The majority would seem to have been made from propose achangefrom an impressive yetinformal unfinished or abandoned canvases and panels lying likeness depicting the subject in alandscape setting about the studio,although at least one or two (possiblysurrounded by attributes/accessories of the ‘palimpsests’ were made from paintings probablynot hunt?) to agrand and manifestlyimposing equestrian originating in the Rembrandt workshop.28 Some of the portrait. underlying imagesshowevidence of having been at The Portrait of Frederik Rihel on Horseback wasan least partiallyscraped or wiped away.29 In some audacious undertaking for both artist and patron. Apart instances Rembrandt appearstohave painted directly from adozen or so imagesofvarious stadholders, on top of the underlying image without applying an life-sized equestrian portraits are exceedingly rare in intermediarylayer;30 in others, an intermediary layer seventeenth-centuryDutch painting.34 Onlyone other wasindeed added to coverall or partofthe underlying of acivilian is known, Paulus Potter’s Equestrian Portrait image.31 of Dirck Tulp of 1653 (Six Collection, Amsterdam). As the present painting is in all likelihood the only While Tulp wasamember of Amsterdam’selite regent known ‘palimpsest’ portrait by Rembrandt or any class with legitimate pretensions to ‘nobility’, Rihel was member of his studio,itisworth considering howor arelatively recent transplant, possessing wealth and whythis might have come about. Documentary business acumen but lacking established familial evidence indicates thatportrait paintersacquired the lineage, making the adoption of this princelyformat supportsfor commissioned works in various ways. A even more remarkable.Significantly, while the costume patron could deliveranagreed supporttothe artist or worn by the standing figure connotes the aristocratic could opt to have the artist supplythe support;the price pursuits of riding and the hunt, Rihel’sbuffleather coat of the supportwas often calculated separatelyfrom (kolder)–aswell as his sword, pistol, bandolier and the price of actuallypainting the likeness.32 No specific fringed sash –had more specificallymilitaryconnota- information concerning acontract to paint the Portrait tions.35 If the current formatofthe Portrait of Frederik of Frederik Rihel on Horseback has survived, but it would Rihel on Horseback indeed represents such aradical appear thateither Rembrandt or Rihel must have re-presentation of the subject’slikeness, the transforma- ventured the unusual proposal to use apreviously tion can be interpreted as abid to proclaim not only painted canvasfor this monumental portrait. Rihel’swealth and equestrian pretensions, but his Without positive identification of the underlying recentlyenhanced civic status, achieved in part figure,itisunclear whether the Portrait of Frederik Rihel throughhis participation in the prestigious procession on Horseback represents pentiment or ‘palimpsest’: a welcoming the young prince of Orangeand his dramatic reworking of asingle portrait of Rihel thatwas mother to Amsterdam, and as aproud new citizen of the first envisagedasastanding figure in alandscape,or capital city. the appropriation of acanvaspreviouslyused for an unrelated (and unfinished) depiction of an unknown sitter (or sitters). The costume would suggest thatthe first portrait wasconceivedabout adecade earlier than the equestrian portrait completed in 1663. There is no documentaryevidence for or against Rihel having Acknowledgements commissioned aportrait from Rembrandt at this earlier date,although there would have been less immediate Fortheir generous assistance and advice in the research reason for him to have done so.Furthermore,toleave and preparation of this article,Iwould liketothank acommissioned portrait unfinished in the studio for Rachel Billinge, Michiel Franken, NorbertMiddelkoop, several yearswould not onlybeimpractical, but would Petria Noble,Joe Padfield, Ashok Roy, Jaap vander Veen, very likelyhave prompted some legal action on Rihel’s Ernst vandeWetering,Colin White and Martin Wyld.

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 109 Marjorie E. Wieseman

Notes 11 Rembrandt, Portrait of aBoy,(oil on canvas, 64.8 × 55.9 cm, The Norton Simon Foundation, Pasadena, inv. F. 1965.2.P); and Man 1D.Bomford, A. Roy, and C. Brown, with contributions from J. Kirby with aMagnifying Glass,(oil on canvas, 91.4 × 74.3 cm, The and R. White, Art in the Making: Rembrandt,National Gallery, Metropolitan Museum of Art, New York, inv. 14.40.621). London, 1988, pp.134–9; N. Maclaren, National Gallery Catalogues: 12 Technical reportbyJ.Plesters, dated 31 October 1985, in the TheDutchSchool 1600–1900,rev.edn. by C. Brown, 2vols., London Scientific Department dossier. 1991, vol. 1, pp.358–62; and D. Bomford, J. Kirby, A. Roy, A. Rüger, 13 In descriptions of the underlying image, all directional indications and R. White, Art in the Making: Rembrandt,rev.edn., National are givenaswith the canvasinits horizontal position, with the Gallery, London 2006, pp.184–9. The current studyisindebted Portrait of Frederik Rihel on Horseback turned 90 degrees anticlock- to the research presented in these publications. wise.Inthe X-ray mosaic,the individual plates have been digitally 2E.van de Wetering,‘Rembrandt’sSelf Portraits: Problems of joined and processed to reduce the effects of the stretcher bars. Authenticity and Function’, in E. vandeWetering,with contribu- The white rectangles at the top and bottom of the image are the tions by K. Groen, P. Klein, J. vander Veen and M. de Winkel, A metal plates bracing the hinged stretcher construction. Corpus of Rembrandt Paintings (hereafter Corpus), vol. 4, Dordrecht 14 School of Thomas de Keyser, Portrait of Johan Wolfert vanBrederode, 2005, p. 96. c.1650 (oil on canvas, 104.5 × 90 cm, Amsterdam, Rijksmuseum, 3A.Bredius, ‘Rembrandtiana, I: Rembrandt’szoogenaamd portret inv. SK-A-3972); and also the anonymous (Dutch?) drawing of a vanTurenne bij Lord Cowper te Panshanger’, Oud Holland 28, Hunting Accident, c.1655, illustrated in M. de Winkel, Fashion and 1910, pp.193–5; for biographical details see also I. H. vanEeghen, Fancy: Dress and Meaning in Rembrandt’s Paintings,Amsterdam ‘Frederick Rihel een 17de eeuwse zakenman en paardenliefhebber’, 2006,fig. 48.Inadditiontothe examples of similar dress worn Maandlblad Amstelodamum 45, April 1958, pp.73–81; N.H. by men in ahunting context cited by de Winkel, compare Schneeloch, Aktionäreder We stindischen Compagnie von1674. Die GovaertFlinck’s Portrait of aGentleman,1642, (oil on canvas, Verschmelzung der alten Kapitalgebergruppen zu einer neuen 124 × 98.5 cm, Amiens, Musée de Picardie), and, for an earlier Aktiengesellschaft,Stuttgart1982, pp.297–301 et passim;and example of this type of costume,Thomas de Keyser’sportrait of Maclaren and Brown 1991 (as in note 1), pp.358–62. The only Constantijn Huygens of 1627 in the National Gallery(NG 212). writer to argue against the identification of the rider as Rihel was 15 Summaryofanalysis by A. Roy, 18 December 2009; see also note 9. R. vanLuttervelt in ‘De Grote Ruiter vanRembrandt’, Nederlands 16 Compare the X-radiographofAristotle’ssleeveinRembrandt’s KunsthistorischJaarboek 8, 1957, pp.185–210,who suggested that Aristotle contemplating the Bust of Homer, dated 1653 (New York, the subject wasJacob de Graeff(1642–90). Metropolitan Museum of Art), illustrated by H. vonSonnenburg in 4G.Leonhardt, Het Huis Bartolotti en zijn bewoners,Amsterdam Rembrandt/Not Rembrandt in TheMetropolitan Museum of Art: Aspects 1978, p. 91 et passim. of Conoisseurship,vol. 1, Paintings: Problems and Issues,exh. cat., 5Van Eeghen 1958 (as in note 3), pp.78–81; see also Maclaren and New York, The Metropolitan Museum of Art, 1995, p. 34, fig. 26. Brown 1991 (as in note 1), pp.359–60, for amore detailed list of 17 Rembrandt, Portrait of Maerten Soolmans,1634 (oil on canvas, the equestrian trappings described in Rihel’sinventory. 207 × 192.8 cm, Paris, private collection); and Portrait of Andries de 6Rembrandt, Portrait of Aletta Adriaensdr.,oil on oak, 65.5 × 55.5 cm, Graeff,1639 (oil on canvas, 200 × 124.2 cm, Kassel, Staatliche Rotterdam, Museum Boijmans vanBeuningen (Stichting Willem Kunstsammlungen, Gemäldegalerie Alte Meister,inv.GK239). vander Vorm); and Portrait of Maria Trip,1639, oil on poplar, 18 Also among those assumed to have studied with Rembrandt in the 107 × 82 cm, Amsterdam, Rijksmuseum (on loan from the Familie late 1650s is his son Titus vanRijn; Bruyn’s suggestion that vanWeede Stichting). Titus might have been responsible for painting Rihel’sadmittedly 7P.W.Klein, De Trippen in de 17de eeuw: Een studie over het onderne- rather stiffhorse has not found acceptance (J.Bruyn, ‘A nunknown mersgedrag op de hollandse stapelmarkt,Assen 1965, pp.455–6; also assistant in Rembrandt’sworkshop in the early1660s’, Burlington G. Schwartz, Rembrandt: His Life, His Paintings,New York 1985, Magazine 132, October 1990, pp.715–8). pp.337 and 340. 19 Twoother portraits of Rihel are listed in the 1681 inventory 8R.van Luttervelt (as in note 3) wasthe first to suggest that (neither of which are known today), including ‘een conterfijtsel Rembrandt’sportrait commemorated this historicevent. The daer hij te voet gaet’ (another portrait [of the deceased] in which he identification of the Heiligewegspoortisnot entirelyconvincing; a is walking), which could well describe afull-length portrait in a quite different structure is seen in JanAbrahamsz. Beerstraaten’s landscape.Both this and the equestrian portrait were listed as drawing of TheHeiligewegspoort on the Heiligewegsburgwal,1662 hanging in the ‘zaal’, presumablythe largest and most important (Amsterdam, Stadsarchief; Collectie tekeningen VanEeghen) and room in Rihel’shome.J.van der Veen (in correspondence with other contemporaryimages. the author) has questioned whether the composition visibleinthe 9Unless otherwise noted, technical information in the following X-radiographofthe National Gallerypainting might not have passage is derivedfrom London 2006 (as in note 1), pp.184–9. begun as acopyafter this now-lost portrait. 10 H. Kühn, in ‘Untersuchungen zu den Malgründen Rembrandts’, 20 VandeWetering 2005 (as in note 2), p. 96. Jahrbuchder Staatlichen Kunstsammlungen in Baden-Würtemberg 2, 21 ‘Palimpsest’ portraits of Rembrandt described in the Corpus include 1965, pp.189–210, wasthe first to identify Rembrandt’suse of a the following works; titles and dates are those used in the Corpus: ‘quartz’ ground and to suggest thatthe practice might be restricted ‘Tronie’with Rembrandt’s Features,1629 (oil on panel, 89.5 × 73.5 to Rembrandt’sstudio.Further research at the National Galleryand cm, Boston, Isabella StewartGardner Museum, Corpus vol. 1, no. elsewhere corroborated his findings(see London 1988, as in note 1, A20), painted over alarger,more broadly-based figure; ‘Tronie’ esp.pp. 27–31). No use of abrown ‘quartz’ ground has yetbeen with Rembrandt’s Features, c.1629 (oil on panel, 38 × 30.9 cm; identified outside Rembrandt’simmediate circle.The most complete Nuremberg, Germanisches Nationalmuseum, Corpus vol. 1, no.A discussion of Rembrandt’suse of grounds is: C.M.[K.] Groen, 21 copy1and vol. 4, p. 597); Rembrandt Workshop (Isack de ‘Grounds in Rembrandt’sWorkshop and in Paintings by his Jouderville?), ‘Tronie’with Rembrandt’s Features,1630/1 (oil on Contemporaries’, in Corpus vol. 4(Dordrecht 2005), pp.318–34. panel, 69.7 × 57 cm; Liverpool, Walker ArtGallery, Corpus vol. 1, Recently, Groen demonstrated in ‘Earth Matters: the origin of the no.A33), painted over afull-length standingfigure; Self Portrait, material used for the preparation of the Night Watch and many 1632 (oil on oak, 64.4 × 47.6 cm; Glasgow,The Burrell Collection, other canvases in Rembrandt’sworkshop after 1640’, Art Matters: Corpus vol. 2, no.A58), aportrait of aman with the head placed Netherlands Technical Studies in Art 3, 2005, pp.138–54, that rather higher up is visible in X-radiograph, with some elements Rembrandt used specificallyafluvial clay, similar to thatused in visible on the surface of the painting; studio of Rembrandt, Bust the manufacture of bricks and tiles and as rawmaterial in the of Rembrandt with an Architectural Background,1637 (oil on oak, manufacture of pottery. 80 × 62 cm, Paris, Musée du Louvre, Corpus vol. 1, no.B10),

110 |NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 Rembrandt’sPortrait(s?) of Frederik Rihel

painted over both an earlier version of the visible bust portrait cm, Berlin, Gemäldegalerie,Staatliche Museen zu Berlin, inv. 828H) and, turned through 180 degrees, asmall-figured scene possibly Man in Armour,1655 (oil on canvas, 137.5 × 104.4 cm, Glasgow, related to Rembrandt’setching of Christ and the Woman of Samaria Kelvingrove ArtGalleryand Museum, inv. 601) and Portrait of a of 1634; ADead Bitternheld High by aHunter, c. 1639 (oil on oak, Yo ung Man,1658 (oil on canvas, 75 × 60.5 cm, Paris, Musée 120.7 × 88.3 cm; Dresden, Gemäldegalerie Alte Meister, Corpus du Louvre,inv.1749 (as ‘Studio of Rembrandt’)). On the Berlin vol. 3, no.A133), apparentlypainted over an unfinished painting,see most recently: J. Kelch, in Rembrandt: ein Genie auf der figure seen almost to knee length, wearing acap or turban; studio Suche,exh. cat. Berlin, Gemäldegalerie,Staatliche Museen zu Berlin, of Rembrandt?, Rembrandt in aBlackCap, c.1637 (oil on oak, 2006, pp.366–9; on the Glasgow picture,C.Brown and A. Roy, 64 × 49 cm; London,The Wallace Collection, Corpus vol. 3, no.C96 ‘Rembrandt’s“Alexander the Great”’, TheBurlington Magazine 134, and vol. 4, p. 604), painted over an uncompleted portrait of a 1992, esp.p.290 and figs 3and 13. woman; Self Portrait in aFlat Cap,1642 (oil on oak, 69.9 × 58.4 cm, 24 London 2006 (as in note 1), pp.214–7. Though not strictlya Windsor Castle,HMQueen Elizabeth II, Corpus vol. 4, no.1), under- ‘palimpsest’, Saskia vanUylenburgh in Arcadian Costume (NG 4930) lying portrait is possiblyalso aself portrait, from about 1633, is ‘drasticallyaltered’ from an underlying depiction of Judith with painted in the Rembrandt workshop but not necessarilyby the Head of Holofernes,partiallycovered with an intermediate grey Rembrandt, with evidence of apossible third painting beneath; Self layer; see London 2006, esp.pp. 93–4. Portrait, c.1645/8 (oil on oak, 73.5 × 59.6 cm, Karlsruhe, 25 VandeWetering 2005 (as in note 2), p. 97. Staatliche Kunsthalle, Corpus vol. 4, no.5), painted over aportrait 26 Respectively: Minervainher Study,1631, Berlin, Gemäldegalerie, of aman, probablynot from the Rembrandt workshop; Self Portrait, Staatliche Museen zu Berlin (Corpus vol. 1, no.A38); Bust of an Old 1654 (oil on canvas, 72 × 58.5 cm, Kassel, Staatliche Museen Woman,currentlygiven to JanLievens, at Windsor Castle (Corpus Kassel, Gemäldegalerie Alte Meister, Corpus vol. 4, no.9), painted vol. 1, no.A32); and ‘Tronie’with Rembrandt’s Features,1630/1, over an unfinished portrait of awoman in millstone ruff,possiblya Liverpool, Walker ArtGallery(Corpus vol. 1, no.A33). work of the 1640s, by apupil or assistant of Rembrandt; Self 27 VandeWetering 2005 (as in note 2), p. 97. Portrait with Sketchbook, c.1655 (oil on canvas, oval, 74.5 × 61 cm, 28 The Self Portrait of about 1645/8 in Karlsruhe (Corpus vol. 4, no.5) San Francisco,Fine Arts Museums of San Francisco, Corpus vol. 4, and the Self Portrait of about 1655 in Florence (Corpus vol. 4, no. no.10version 2), copyafter alost original by Rembrandt, painted 12). See note 21 above. over an unfinished paintingofanunidentified subject; Self Portrait, 29 David with the Head of Goliath beforeSaul,1627 (Basel, Corpus vol. 1, 1655 (oil on oak, 64.3 × 50.8 cm, Vienna, Kunsthistorisches no.A9), panel probablypartiallyscraped smooth before starting Museum, Corpus vol. 4, no.11), painted over asketch of aseated the new image;Rembrandt workshop (Isack de Jouderville?), female nude,possiblyderivedfrom Rembrandt’s Bathsheba of 1654, ‘Tronie’with Rembrandt’s Features,1630/1 (Liverpool, Corpus vol. 1, although the brushwork is not reminiscent of his manner; and Self no.A33), underlying image scraped away before the new image Portrait, c.1655 (oil on canvas, 69 × 59 cm, Florence,Galleria degli wasbegun; Minervainher Study,1631 (Berlin, Corpus vol. 1, no.A Uffizi, Corpus vol. 4, no.12), with an open book discernible at lower 38), portions of the underlying composition mayhave been wiped right, possiblyfrom astill lifeand probablynot originating in away while the paint wasstill wet; and Self Portrait in aFlat Cap, Rembrandt’sstudio.None of the ‘palimpsest’ self portraits date 1642 (Windsor Castle, Corpus,vol. 4, no.1), underlying portrait after about 1655. VandeWetering has suggested thatthe absence partiallyscraped away.See notes 21 and 22 above. of later ‘palimpsest’ self portraits mayreflect the fact thatwhile the 30 Forexample, Bust of aMan in Gorget and Cap, c.1626/7 (Corpus vol. earlier self portraits were made on the artist’sown initiative,the 1, no.A8), Joseph accused by Potiphar’s Wife,1655 (Berlin), and Self later ones are more likelytohave been done on commission; see van Portrait (Karlsruhe, Corpus vol. 4, no.5). See Corpus vol. 4, p. 423. de Wetering 2005 (as in note 2), p. 97. 31 Forexample,the Bust of an Old Woman at Windsor Castle (Corpus 22 Ibid.; examples noted in the Corpus are: Spectacles Pedlar (Sight), vol. 1, no.A32); in the case of the David with the Head of Goliath c.1624–5 (oil on panel, 27 × 17.8 cm, private collection, Corpus beforeSaul,1627 (Basel, Corpus vol. 1, no.A9), the partiallyscraped vol. 1, no.B3), painted over aseated nude figure; Foot Operation (oil panel waspossiblycovered with ayellowbrown ground before on oak, 31.8 × 24.4 cm, Switzerland, private collection, Corpus starting the new image;alight intermediarylayer mayhave been vol. 1, no.C11), X-radiographshows unclear indications of brush- applied to the panel supportof Bust of Rembrandt with an work not associated with known paint layers; Bust of aMan in Architectural Background before the final painting (1637, Paris, Gorget and Cap, c.1626/7 (oil on oak, 40 × 29.4 cm, private collec- Corpus vol. 1, no.B10); and the portrait underlying the Self Portrait, tion, Corpus vol. 1, no.A8), painted over the head of an old man at 1654 (Kassel, Corpus vol. 4, no.9)was partiallycovered with a asignificantlylarger scale; David with the Head of Goliath before flesh-colouredlayer,applied locally. See Corpus vol. 4, p. 423. Saul,1627 (oil on oak, 27.2 × 39.6 cm, Basel, Oeffentliche 32 J. Bruyn, ‘Patrons and EarlyOwners’, in J. Bruyn, B. Haak, S. H. Levie, Kunstsammlung, Corpus vol. 1, no.A9), painted over ahead, P. J. J. vanThiel and E. vandeWetering, Corpus vol. 2(1986), p. 91. turned 90 degrees clockwise; circle of Rembrandt, Bust of an Old 33 Filippo Baldinucci (and others) remarked upon Rembrandt’sslow Woman, c.1628? (oil on panel, 35.4 × 28.9 cm; Essen, collection H. working methods, which sometimes required twoorthree months vonBohlen und Halbach, Corpus vol. 1, no.C42), possiblypost-sev- to complete aportrait (F.Baldinucci, Comminciamentoeprogresso enteenth century, under-lying painting of ahalf-length figure; Bust dell’arte dell’intagliareinrame colle vite di molti de’più eccellenti maestri of an Old Man with aTurban,c.1627–8 (oil on oak, 26.5 × 20 cm, della stessa professione,Florence 1686, p. 79). Rembrandt also collection Foundation Aetas Aurea, Corpus vol. 4, addendum no.3); occasionallyleft commissioned paintings unfinished, completing Interior with Figures (‘La main chaude’), c.1627–9 (oil on panel, 21 × them onlywhen pressedbythe patron or by economic need. One 27 cm, Dublin, National GalleryofIreland), painted over abust- notable example is the series of paintings of the Passion for Frederik length portrait of aman, turned 90 degrees clockwise; An Artist in Hendrik,stadholder of the Netherlands, begun in 1632, with the his Studio, c.1629 (oil on panel, 25.1 × 31.9 cm; Boston, Museum last paintingdelivered onlyin1646. On Rembrandt’slegal disputes of Fine Arts, Corpus vol. 1, no.A18), possiblypainted over another with patrons, see P. Crenshaw, Rembrandt’s Bankruptcy: TheArtist, painting; Minervainher Study,1631 (oil on oak, 60.5 × 49 cm, His Patrons, and the Art MarketinSeventeenth-Century Netherlands, Berlin, Gemäldegalerie,Staatliche Museen zu Berlin, Corpus vol. 1, Cambridge2006, esp.pp. 110–35. no.A38), astanding figure with aheaddress and other,less legible, 34 See In het zadel: Het Nederlands ruiterportret van1550 tot 1900,exh. forms are visible in X-radiograph; and Man with aGorget and Plumed cat., Leeuwarden, ’s-Hertogenbosch and Assen, 1979–80, pp.99ff. Cap, c.1630–1 (oil on panel, 65 × 51 cm; Los Angeles, J. Paul Getty 35 On the historyofthe kolder and its connotations, see E. Sint Museum, Corpus vol. 1, no.B4), painted over an earlier picture: a Nicolaas and H. Stevens, ‘Kolders. Vanmodieus militair kledingstuk half-length figure of ayoung man, turned through 180 degrees. tot slagveldreliek’, Bulletin vanhet Rijksmuseum 54, 2006, 23 Joseph accused by Potiphar’s Wife 1655, (oil on canvas, 113.5 × 90 pp.266–89; and de Winkel (cited in note 14), pp.108–9.

NATIONAL GALLERYTECHNICAL BULLETIN VOLUME 31 | 111