Connoisseurship & Rembrandt’s denied studio practice, a ‘status questionis’

In memory of Walter Liedtke

What distinguishes a Rembrandt from a non-Rembrandt? The exhibitions in Amsterdam and The Hague at the occasion of the 350th anniversary of his death, are a good opportunity to look back on the history of the Rembrandt-connoisseurship and the problem of authenticity.

The lenses through which we look at art and its authenticity are differently tinted over the years. Rembrandt emerges now as a different figure as the one popularized in the 1950s. Between 1909 and 1951, the USA-customs administration registered the import in America of 9.428 Rembrandt paintings. The number of accepted ones in different catalogues went down from more than a thousand circa 1915, to 630 in Bredius’ catalogue in 1935 and to 420 by Gerson in 1968; Tümpel reduced it to 265 in 1988. Gary Schwartz in 1987, reworking Gerson’ catalogue, accepted 349 of which 35 with question marks. The Dutch Rembrandt Research Project’s (RRP), financed by the NOW (Nederlands Wetenschappelijk Onderzoek), reduced it to approximatively 250. Since 1968, initiated by Bob Haak, Joshua Bruyn and a group of Dutch art historians, the RRP’s main activity was to identify in a narrow sense the works by Rembrandt’s own hand as a restrictive ‘category A’ and the doubtful and rejected in categories B & C. Already in the second volume, the RRP announced its strategy: “an objective connoisseurship based on all the elements of scientific analysis in a consistent way for the whole oeuvre, against the ‘onberedeneerde varianten van het traditionele kennerschap.”

In 1993, four of the main authors of Corpus Volumes 1-3 (period 1625-42) left, after which, the remaining Ernst van de Wetering detailed the future of his research, alone on board. This resulted in Volume IV The Self-Portraits (2005) and Volume V The Small-Scale History Paintings (2010). The group learned a lot on the job, but ideological disputes arose, which lead to the end of the collaboration in 1993, van de Wetering assuming the sole responsibility of volumes IV and V. In the three first volumes, the RRP accepted 146 paintings as category A. PDF files of the five first volumes are made available by the RRP on the website. A substantial body of paintings was left uncatalogued, including the large history paintings, portraits (apart from self-portraits) and landscapes, all painted after 1642. In 2014, Ernst van de Wetering published ‘Rembrandt’s Paintings Revisited. A Complete Survey’, as a kind of personal scholarly approach and an assessment of Rembrandt’s pictorial intentions and processes (as a kind of Volume VI), reinstating some of the RRP’s rejected paintings (44). Ernst van de Wetering makes graciously no claim to finality of judgment in his remarkable achievement, but expects a detailed argumentation rather than just an opinion to argue a case. A number of art historians did that by stressing the value of some of the RRP pre-assumptions. In 1910, Valentiner already hinted at an early workshop practice in the two München portraits, speaking about the female portrait: ‘Möglicher war hier eine Gesellenhand mit beteiligt… Die Rolle welche in den frühen Amsterdammer Zeit die Schüler in Rembrandt’s atelier gehat haben, ist von der Forschung noch nicht klar ausgeschält’.Walter Liedtke stressed the absence of the assessment of ‘quality and variety in the workshop section of the exhibition’ in Berlin’s Altes Museum ‘Der Meister und seine Werkstatt’ (1991) and the non-definition of the notion ‘Rembrandt’s Werkstatt’ itself by the RRP.

Today, even after the end of the RRP and the insights it brought, the label ‘Rembrandt’s workshop in the 1630s’ is open to many contradicting definitions and inadequate a priori assumptions. The fact that the artist collaborated with his assistants in Leiden and in the first Amsterdam period is denied early on by the RRP. Since the 17th century, when ‘liefhebbers’ (amateurs) concentrated on the principal parts of the composition, connoisseurship evolved, differing its attitude to authenticity between then and now. The distinction between fully autograph and partly assisted works is strongly reflected in their present-day financial and art historical value. Ernst van de Wetering uses himself the label ‘Werkstatt’ in a letter of December 14, 2004 to Dr. Peter Wolf of the Dorotheum auction house about ‘The Angel leaves Tobias and his family’, Atelier van Rembrandt naar Rembrandt’s originaal in het Louvre: ‘Die Wiedergabe in ihrem (in Entwurf zugesandten) Katalogtext (Rembrandt Werkstatt) scheint korrekt. Wir haben das Bild vor vielen Jahren studieren können’. Also Anna Tummers reflects on the notion connoisseurship and Rembrandt’s studio practice. All over the history of connoisseurship, it was commonly accepted that 17th century artists would often collaborate with assistants.

Already in 1973, Gary Schwartz alerted us about the danger of the pre-conceptions in the RRP. It’s definition remains an exercise in connoisseurship until today and does not really fit the visual evidence of the best observers. Moreover, connoisseurship in this field still has different dimensions, depending on each of the scholars involved, without reaching a consensus. In 2004, Walter Liedtke proposed elements for a better understanding of Rembrandt’s relation with his pupils, dilettanti and professional collaborators in the 1630s’ in Amsterdam. Until today, consensus is lacking about Rembrandt’s studio practice. The exhibition catalogue in 1991-92: Rembrandt: De Meester & zijn werkplaats/ Schilderijen’. Ernst van de Weterings contribution: ‘Rembrandt’s schilderwijze: techniek in dienst van de illusie’ (pp 12-39), clarified greatly our understanding of Rembrandts painterly practice, ductus and intentions. Josua Bruyn’s : ‘Rembrandts werkplaats: functie en produktie’ (pp 68-89), represented the RRP’s concept of the studio practice. The latter initiated a vivid discussion.

The purpose of the present essay is not to criticize the RRP as such, but to reopen the discussion about Rembrandt’s eventual delegation of creativity in his workshop and to achieve a better assessment of the pictorial identities of individual artists in his circle. In 2002-03, working on my Ph. D. and tackling the problem of ‘Rembrandt attributions’, I had the pleasure to discuss problems in the attributions to Rembrandt with Walter Liedtke, Egbert Haverkamp-Begemann, Christopher Brown, Gerry Schwarz, Marc Evans, Jacques Foucart and Claus Grimm. Their connoisseurship and openness towards opinions of others, knowing how to distinguish quality in art, was of an extraordinary support to my endeavor.

The small number of collaborators and pupils from his first years in Amsterdam (1631-34) contrasts with the huge number of portrait paintings in Rembrandt’s style (often formal portraits), rejected by the RRP, but undoubtedly from his immediate circle and datable in the 30s’. Most of them remained without convincing paternity until now. Many of the RRP reattributions to individual members of his circle are doubted. Whether gifted collaborators had a limited participation in these portraits, leaving the most important part or some highlights to Rembrandt is clearly rejected by the RRP, with only two exceptions. The first is ‘the hands’ in the ‘Portrait of Johannes Wtenbogaert’. The second, considered by most scholars as a studio version with improvements by the master, is the second version of Hermitage ‘principael’, the Münich Sacrifice of Isaac, of which the under drawing is a transfer by sjabloon (carton) of the contours of the prime version. For sure Rembrandt did not need a transferred under drawing himself to create a second version. The Munich version is inscribed “Rembrandt verandert. En over geschildert 1636”. This suggests that at least at some occasions Rembrandt corrected work by pupils. Was this his standard practice in the 1630s? In the Chicago symposium (1969), Gerson stressed the importance of Sandrart’s remark (30 years later in 1675) that: ‘Next to Rembrandt’s pupils, a number of Amsterdam’s rich and prominent dilettanti fancied to participate for a short period in painting and drawing in Rembrandt’s studio’. Could it be possible to distinguish the hand of these dilettanti (such as Leendert van Beyeren and Constantijn van Renesse) from those of his professional pupils and collaborators?

After Isaac Jouderville in Leiden (1629/30-late 1631), Rembrandt’s only securely documented early collaborator in Amsterdam is Govaert Flinck. But his presence is not documented before1635-36 in Rembrandt’s studio and, according to Houbraken, only for one year. S.A.C. Dudok van Heel suggests that Flinck came to Amsterdam to work in the important workshop of the dealer and entrepreneur Hendrick Uylenburgh (Rembrandt’s future father in law’s brother) in the Lauriergracht. There, Rembrandt practiced and was able to find skilled collaborators, without necessarily having his own apprentices (1632-34) to assist him in his portraits’ secondary aspects. In 2004, Walter Liedtke listed in the appendix B of his article Rembrandt’s ‘Workshop revisited’, the documented and probable pupils in the order of their likely dates of presence in the master’s studio in Leiden (Gerard Dou when Rembrandt was in Leiden 1628-31 & supposedly also there: Isaac Jouderville) and in Amsterdam (19 artists from Govaert Flinck til Arent de Gelder, reaching from 1635 until the mid-1660s).

According to Dudok van Heel, “in 1631, Rembrandt travelled regularly to Amsterdam to work in Uylenburgh’s house and workshop on the corner of the Sint Anthonisbreestraat. In these two workshops, delegation and distribution of tasks was usual. Rembrandt finally settled permanently in 1633 in Amsterdam. There, he joined the painters’ guild only in 1634 (since all incoming members of the guild had to have resided at least a year before in Amsterdam.) He left Uylenburg’s workplace in 1635, while Flinck remained there as the masterpainter”. The latter had his training in Leeuwarden from 1629 on with Lambert Jacobs until 1631. Another collaborater, Ferdinand Bol, (arriving circa 1636) remained in Rembrandt’s studio until 1641/42 after his apprenticeship with Jacob Gerritsz Cuyp in Dordrecht. Both were 20 years old and already skilled ‘gezellen’ when entering Rembrandt’s studio. Would Rembrandt sell paintings by his pupils ‘as his own’ only on a few occasions? Was this possible without collaborating with them on individual works such as portraits, as was the usual studio practice (Rubens, Honthorst, Mierevelt) all over the Netherlands in those days?

Unlike the understanding we have about Rubens’ studio practice with fairly clear divisions of responsibility, the negation of Rembrandt’s studio practice, as seen by the RRP, is still based on a great number of unproven assumptions and pre-conceptions by the lack of adequate context and documents. The publication of scientific material by the RRP revealed a lot of interesting elements open to interpretation by all. It allowed the RRP to exclude a great number of so-called Rembrandts but also some later reaccepted ones. The RRP did not bring much clarity in the relation between the work of the master, the companions and pupils in his studio, in the light of their own production later independently. It left a great number of Rembrandt style portraits of the 1630s’ without paternity excluded often on the same scientific material arguments as some of the RRP accepted ones. Thus, based on the old-fashioned connoisseurship which the RRP loathed so much.

Without the discovery of new material or documentary evidence. Ernst van de Wetering and the RRP changed their opinion on a considerable number or earlier rejected paintings. A subjective approach, contradicting the supposed objectivity they claim. Later, Joerg Wadum brought some new insights through material analysis in Rembrandt’s pictorial strategy and the different hands in his circle. The core of Rembrandt’s autograph work is now much better defined. But, because of the lack of sufficient first hand reports by his pupils, skilled collaborators and witnesses of his time, the extent of Rembrandt’s participation in his studio’s activity is still unclear. The study of the master’s paintings, their contemporary copies and the relation with his drawings and etchings revealed some insights in Rembrandt’s impulsive creative process, but the eventual contribution of his assistants in some of his paintings remains uncertain. Also the dividing line between the corpus of each artist involved in a relationship with Rembrandt’s style, and the work of the master is not sufficiently defined, nevertheless the monumental achievement of Werner Sumowski and many others in the study of his circle.

Rembrandt’s case is not so clean-cut as Rubens’. The latter set out a precise range of grades of authorship of his pictures on offer and their corresponding price range in his correspondence with Sir Dudley Carrington. But his client’s expectations differed from the actual appreciation. Also, Rubens position in his studio relations was different then Rembrandt’s and Uylenburgh’s. Today, a painting invented by Rubens, but only reworked by him in a limited way, is usually described as ‘Rubens and workshop’, even if it had been sold by Rubens as autograph. A large supply of workshop paintings promoted an artist’s name brand all over Europe. Is this also applicable to Rembrandt’s early Amsterdam portrait practice, which is assumed by the RRP to be hyper-individual?

Rembrandt’s creativity and his painting strategy is still seen by the RRP as a creative exception in his time, compared to the studio practice of most successful artists of his time. This leaves space for an ongoing process of various interpretations and hypothesis, dealing with artistic intention and quality assessment. The spontaneity of the Rembrandt’s hand compared to that of his studio output is, for sure, harder to define than in the case of Van Dyck or Rubens compared to their studio. The delegation of execution and creativity, where a great number of assistants worked under close supervision at workshop replicas and even at originals, later retouched by the master himself in the most important parts, is well documented. The assistants collaborated and even executed a major part of the original production of series and decorative ensembles, such as in Rubens’ series for the Torre de la Parada, after the master’s sketches and drawings. This was also the case in the Gerard van Honthorst, Jan Mijtens and Michiel van Mierevelt studios in the Netherlands. At the contrary, Rembrandt’s monumental ‘The Night Watch’ is assumed to be entirely monograph, as is the ‘Claudius Civilis’, later cut up, composition. Also, Johannes Vermeer or Pieter De Hooch did not have a workshop. But, Rembrandt, being a successful artist as well as a teacher, an entrepreneur and a dealer, could have had a very active one in his early years in Amsterdam or otherwise, should have been very active himself, which could explain the variations in quality among his portraits of the 1630s’, as Christopher Brown remarked. Arthur Wheelock regretted that the RRP’s A-B-C distinction does not take in account the possibility of works produced collaboratively in the Rembrandt workshop and eventually corrected by the hand of the master. Wheelock’s hypothesis is further strengthened by the 1656 inventory of Rembrandt’s possessions, in which a number of paintings are described as “geretukeerd by Rembrandt” (retouched).

We will here not go into the discussion of individual reattributions. Many variables are involved in the assessment of Rembrandt’s changing studio practice over the timespan of his career, but of no other period than the 1630’s exist so many rejected portraits by the RRP. Egbert Haverkamp-Begemann assumed that a so self-conscious and narcissistic artist as

Rembrandt expected, such as Rubens did, his best pupils to go their own way. In Rembrandt’s time, pastiches after his work were only praised for the way they deceived would-be connoisseurs. In the Netherlands, deception about authorship was seen as a hilarious practice since Goltzius’ imitation of a Dürer etching. It opened the way for a Heyman Dullaert painting to pass in an Amsterdam sale as a Rembrandt.

The highlight of the ‘Rembrandt by Himself’ exhibition was the confrontation between two portraits of the young artist. The ‘Self-Portrait with Gorget’ (circa 1629) in the Germanisches Nationalmuseum Nuremberg was considered by the RRP (Vol I, p 225-30, A-21) as not by Rembrandt (not category A). After being examined by Josua Bruijn and Ernst van de Wetering in 1973, the other version is in the Mauritshuis’ was accepted as category A, described (wrongly) as an anchor stone of the master’s hand, but quite different of the other early self-portraits by its meticulous way of painting. The Nuremberg one was cataloged as an imitation done in Rembrandt’s circle. The RRP based its decision on the ’exceptional high quality evaluation, seen as superior to the Nuremberg one’ of the Mauritshuis version. This way of applying ‘old fashioned connoisseurship’ by the RRP contradicted its declared ‘objective ambition’.

Already in 1875, the Nuremberg version fostered a public debate. Both versions were seen by Alfred von Wurzbach (1846-1915) as studio paintings. Willem von Bode considered both authentic. Hofstede de Groot rediscovered on the Nuremberg painting the remains of the signature: RHLf, but it was ignored completely in the catalogues of the master’s oeuvre by Bredius, Bauch and Gerson. The third version in the museum of Copenhagen was seen as a studio copy.

In 1991, Claus Grimm’s pictorial quality assessment, critical stylistic comparison using photographed details and x-rays of all three versions, was an eye-opener. He, as a so-called ‘old fashioned’ connoisseur nevertheless using all technics of scientific analysis, considered the Nuremberg one was the ‘principael’; the other two versions he assumed to be copies. He concluded that the Mauritshuis version was second best and produced in Rembrandt’s studio by a close follower.

Only after Joergen Wadum and others found and published their discovery by IRR of the detailed under drawing of the outlines of the Nuremberg one, transferred with the aid of a cartoon, underneath the pictorial surface of the Mauritshuis version, the Nuremberg portrait was accepted as the prime version by the hand of the master (principael). They gave detailed additional information about the differing characteristics of painting technique and intention between both. A personal evaluation by Ernst van de Wetering (RRP) was also taken in account. IRR was used only in a few cases by the RRP. Since then, the Mauritshuis version is considered by most art historians to be by another hand (probably by Gerrit Dou). It is painted with considerable care, usually absent in Rembrandt’s spontaneous technique.

Eric Jan Sluijter defends the attribution of both versions to Rembrandt himself. He proposes to accept the hypothesis that Rembrandt, at this early stage in his career, was showing of his ability to paint in two different manners (handelingen): a smooth and a more spontaneous-painterly one. In Sluijter’s view, Rembrandt used in the Mauritshuis’ version some smooth elements, he recognizes in the ‘Portrait of Joris de Caulery dressed as an officer’ (San Francisco, 1632, at a moment when Govert Flinck was Rembrandt’s collaborator). He describes the technique of the San Francisco portrait: ‘Its more cursorily painted face’ and ‘the gorget with its smooth brushwork alternating with the thick impasto in the white catchlight on its standing rim’, are seen by Sluijter as similar to the Mauritshuis version of Rembrandt’s portrait.

In contrast, Christopher White writes that Rembrandt did not ‘draw in chalk on the panel’ as a preparation for his portraits, but that ‘he began directly by applying lightly brushed-in lines over which he painted freely in monochrome’ (under modelling). Wadum details the different manner of paint handling and brushstrokes, especially the absence of the extensive use of a soft brush to create smooth transitions between light and shadow (blending) in the Nuremberg one and writes: ‘All these points not only confirm that the Mauritshuis work is a copy, possessing many of the mistakes associated with copies’. In his monograph Rembrandt, the Painter at Work, van de Wetering writes: ‘Over the ‘coloured’ oil ground, Rembrandt would start by delineating the composition in lines of varying thickness. Then he would apply the dead-colouring or under-modeling in monochrome hues ranging from dark reddish brown to light ochre. Certain areas would be heavily toned and some even impasto; others would be semi-transparent, allowing the light ground to shine through’. Moreover, scientific analysis has revealed that ‘at this stage Rembrandt was probably working with paint bound in an aqueous medium’. This leads to the conclusion that the master used a different, more vivid and freely sculpting the face, technique, resulting in a different and distinct ductus, than artists in his immediate circle, following the contours of under-drawing of the face in a restrained manner.

Therefore, the Mauritshuis version and its brighter paint layers only could have been done in Rembrandt’s workplace in Leiden, shortly after the completion of the prime version. Did Rembrandt remodel in his early years the work of a pupil to the point that it deceived some of the great connoisseurs of their times? Or did he leave the production of copies to his studio under his supervision as most successful artists of his time did?

The comparison between the ‘Self-Portrait as a young man’, c1628 in the Amsterdam Rijksmuseum and the ‘Anonymous contemporary copy’ after the master (c1629) in Staatliche Museen Kassel, Gemäldegalerie Alte Meister, although the latter does not reveal under-drawing, leads to a similar conclusion. Wadum compares also the ‘Self-portrait with Gorget and Beret’ in the Indianapolis Museum of Arts, with the ‘Anonymous copy’ in the Atami, Mochiki Okada Association Museum. He is convinced that the painting technique of the Indianapolis one clearly points to Rembrandt. The relation between the Hermitage-prime and Munich-version of the ‘Sacrifice of Isaac’ is also an element to reconsider an indication of a possible workshop practice early on in the master’s studio. It is obvious that Rembrandt’s economic strategy changed later at the end of the 1630s. Rembrandt’s genius collaborates in a great variety of strategies with his apprentices, dilettanti and ‘gezellen’ (close associates).

From the above, it is clear that even early on in his career, Rembrandt had artists in his immediate circle closely imitating his style and execution. They only could have made that kind of copies after the original, present in the studio. What was the working relationship of Rembrandt with them? The accurate observations by some of the above mentioned connoisseurs, some called old-fashioned, permitted to correct quite some erroneous affirmations of the early RRP analysis, but it did not lead to a further explorations of a possible workshop practice and its evolution over the different stages in Rembrandt’s career. The comparative study of contemporary copies and versions, as well of the master as of his close associates, and their later achievements, is still an open field.

Science is not a sufficient arbiter of perceived quality in art, it needs the sharp eye of an open competition of connoisseurs. Scientific material analysis is objective, the therefrom derived diagnostics are not. Too often, they are biased by pre-assumptions about the master’s painterly profile. Science provides often data enabling to exclude impossible assumptions of paternity, but the distinction of the master’s hand and that of studio associates is a problem reserved for and a challenge to, those knowing how to look at quality in art. There is no more ‘old fashioned connoisseurship’ extant, since a real connoisseur is open to all the disciplines of art history and material analysis, interested in other people’s opinions. Open to all, this ongoing contest brings us nearer to the real genius of the artist.

Univ. Prof. Dr. Jan De Maere
University of Art&Design, Cluj-Napoca


1 RRP 1982-89, Vol. I, note 5

2 Translation: “unreasonable variations of traditional connoisseurship”.

3 E. van de Wetering, Burlington Magazine 135, 1993, pp 764-65 and E. van de Wetering & P. Broekhoff, New directions in the RRP, Part I, the 1642 self-portrait in the Royal collection, Burlington Magazine 138, 1 996, pp 174-75

4 /Rembrandt/cms/corpus.

5 München, Alte Pinakothek, inv. Nr & Nr

6 W. R. Valentiner, ;;;;;;;;;;;;;;; , Monatshefte für Kunstwissenschaft, III, 1910, pp 7.

7 W. Liedtke, Rembrandt and the Rembrandt Style, review of the Exh. In Apollo 3/1992, pp 140-50.

8 C. Brown, Jan Kelch & Pieter van Thiel, Rembrandt: De Meester & zijn werkplaats, schilderijen, Rijksmuseum Amsterdam-Waanders Zwolle, 1991, ill 89, pp 75-79.

9 Dorotheum sale catalogue 2015, … 167 : ’Der Erzengel verlässt Tobit’.

10 A. Tummers, The Eye of the Connoisseur. Authenticating Paintings by Rembrandt and His Contemporaries, Ed J. Paul Getty Museum, LA, 2011.

11G. Schwartz, Rembrandt bij de tandarts, in Hollands Maandblad, November 1973, 3-9.

12 W. Liedtke, Rembrandt’s ‘Workshop’revisited, Oud Holland, 2004, Vol 117, Nr ½, pp 48-73

13 Christopher Brown, Jan Kelch & Peter van Tiel, Rembrandt : De Meester en zijn werkplaats, schilderijen, Rijksmuseum Amsterdam-Waanders Zwolle, 1991.

14 J. De Maere, Neurosciences et Connoisseurship. La physiologie neuronale du Beau et l’attribution des tableaux anciens, Ph. D. Gent Univ. Press, 2011, chap. IV.3.1.1., pp 429-44.

15 RRP, 1982-89, Vol 2, pp 63,note 5.

16 RRP, 1982-89, Vol 3, pp 111-12, fig 10.

17 W. Lietdke, 1995-96, note 13

18 S.A.C. Dudok van Heel, De Schilder, zijn Leven, zijn Vrouw, de min en het Dienstmeisje, Kroniek van het Rembrandthuis, 2000, nr ½, pp 15.

19 W. Liedtke, Rembrandt’s ‘Workshop’revisited, Oud Holland, 2004, Vol 117, Nr ½, pp 48-73

20 S.A.C. Dudok van Heel, Kopstukken. Amsterdammers geportetteerd 1600-1800, in exh. cat.Amlsterdam’s Historisch Museum 2002-03, pp 50-53

21 S.A.C. Dudok van Heel, 1991-92, pp 54

22 J. Bruyn, 1991-92 pp 71 note 27.

23 H. Vlieghe, Rubens’s atelier and History painting in Flanders: a Review of the Evidence, in exhib. Cat. The Age of Rubens, Boston Museum of Fine Arts 1993/94, pp 158-70.

24 M. W. Ainsworth, J. Brealey, E. Haverkamp-Begemann and P; Meyers, Art and Autoradiography: Insights into the genesis of paintings by Rembrandt, Van Dyck and Vermeer, exhib. Cat. Art in the Making: Rembrandt, 1982, The Metropolitan Museum of Art, New York,

25 E. Van de Wetering, Rembrandt: The Painter at Work, Amsterdam 1997.

26 J. Wadum, Rembrandt under the Skin. The Mauritshuis Portrait of Rembrandt with Gorget in retrospect, Oud Holland, 114, 2000, : pp 164-87.

27 W. Sumowski

28 W. Liedtke, Review of ‘Study exhibition: The Flight of Loth and his Family from Sodom. Rubens and his Workshop, National Museum of Western Art Tokyo, 1993, Burlington Magazine, 135, pp 718-19.

29 C. Brown, Rembrandt at work: Methodological issues raised by recent research, (unpublished lecture) 1989.

30 A. Wheelock 1995, pp 207, note 21

31 Straus & Van der Meulen 1979, pp 351 and p 361.

32 E. Haverkamp-Begemann, 1969-70, pp 25, 29 & note 74.

33 W. Lietdke, 1995-96, pp 12, note 13

34 C. Grimm, Die Frage nach der Eigenhändigkeit und die Praxis der Zuschreibung, in T. W. Gaechtgens (ed.), Künstlerischen Austausch – Akten des XXVIII Int. Kongresses für Kunstgeschichte Berlin, 15-20 Juli 1992, Berlin, Akad. Verlag 1993, pp 631-48.

35 C. Grimm, Eine Neubewertung seiner Porträtkunst, Stuttgart-Zurich, 1991, pp 20-21, 24-28.

36 J. Wadum and C. van der Elst, Examination report in the Mauritshuis Conservation studio, June 1998.

37 J. Wadum and C. van der Elst, Attribution et désattribution. Les portraits de Nuremberg et de La Haye, Dossier de l’Art 61, 1999, pp 34-43.

38 C. White & Q. Buvelot (ed), cat exh. Rembrandt Zelf, London National Gallery-Den Haag Mauritshuis 1999-2000, ill cat nr 14a-14b, pp 112-17.

39 E. J. Sluijter, The Tronie with a Young Officer in the Mauritshuis. A second version by Rembrandt himself?, Oud Holland 114, 2000, pp 188-94.

40 E. van de Wetering, Rembrandt, the Painter at Work, Amsterdam, 1997.

41 J. Wadum, op cit, Oud Holland, 2000, pp 172