Lot 10
  • 10

Rembrandt Harmensz. van Rijn Leiden 1606 - 1669 Amsterdam

bidding is closed


  • Rembrandt Harmenszoon van Rijn
  • Study of an elderly Woman in a White Cap
  • oil on panel
  • 21 by 14 3/4 in. 53.3 by 37.5 cm.


Fürst Wenzel Anton Kaunitz (1711-1794), Kaunitz Stadtpalais, Dorotheergasse, Vienna (his cypher stencilled to the bottom right corner of the picture when enlarged; see Fig. 8);
Probably by descent and inheritance to his nephew Fürst Aloys Kaunitz;
His sale of the contents of the Kaunitzer Stadtpalais held at No. 1025, Obere Bräuner Straße, first floor, with Artaria & Co., Vienna, 13 March 1820, lot 138: `Rembrandt, portr. veille femme, h 20 by l 13' [inches];
Probably with John Smith, London;
From whom probably purchased by the Hon. George John Vernon, London;
By whom (anonymously) sold or offered, London, Christie's, April 15-16, 1831, for 60 Guineas;
The Right Hon. Lord (Robert?) Vernon, London, by 1836;
A. Watson, Sandford Manor, Woodley;
Eldridge R. Johnson, Moorestown, New Jersey, by 1931, by descent in the family until consigned to Newhouse Galleries;
With Newhouse Galleries, New York, from whom purchased in 1971 by  F. Howard Walsh and Mary D. Fleming Walsh, Fort Worth, Texas, and thence by descent in the family.


Detroit, Detroit Institute of Arts, Paintings by Rembrandt, May 2-31, 1930, no. 40;
Amsterdam, Rembrandthuis, September 22nd - November 29th 2005.


Artaria & Co., Catalogue des Tableaux provenants d’une ancienne Galerie célebre (Prince Kaunitz), Vienna 13 March 1820, lot nr 138;
J. Smith, Catalogue Raisonné of the Works of the Most Eminent Dutch and Flemish Painters, London 1836, vol. VII, p. 172, no. 540 (as Rembrandt);
C. Hofstede de Groot, A Catalogue Raisonné of the Works of the Most Eminent Dutch and Flemish Painters of the Seventeenth Century, London 1916, vol. VI, p. 328, no. 690 (as Rembrandt, a portrait of his mother);
W.R. Valentiner, Rembrandt, Klassiker der Kunst, Stuttgart 1921, p. XVI, no. 13, illus. (as Rembrandt, about 1630, a portrait of his mother);
C. Hofstede de Groot, Die holländische Kritik der jetzigen Rembrandt-Forschung und neuest wiedergefundene Bilder, 1922, p. 6 (as Rembrandt, circa 1645/50, a portait of his mother);
W.R. Valentiner, Wiedergefundene Gemälde, Berlin 1923, p. XXIV, no. 59;
W.R. Valentiner, Rembrandt Paintings in America, New York 1931, no. 92, illlus. (as Rembrandt, circa 1645);
W. Sumowski, Gemälde der Rembrandt-Schüler, vol. IV, Landau/Pfalz 1983, p. 2881, footnote 1, no. 13 (as Rembrandt School, circa 1650, , with vague reminiscences of (‘mit Anklänge an’) Karel van der Pluym.  Sumowski was apparently thinking here of the Old woman with a book, Sumowski Vol. IV, nr. 1593. A comparison between those two paintings shows however how far-fetched is Sumowski’s reference to Van der Pluym);
E. van der Wetering, Rembrandt’s Oil Studies: new light on an old problem, unpublished text (the Dutch text may be viewed on the Rembrandthuis website), September 2005. 

Catalogue Note

This is an important re-discovered work by Rembrandt, painted in Amsterdam around 1640, which came to light in 2003, and upon completion of detailed examination and research by the Rembrandt Research Project and painstaking cleaning by Martin Bijl, was publicly unveiled in the building where it was probably painted – the Rembrandthuis in Amsterdam – in October 2005.  In the 19th and early 20th Centuries it had been known and admired as an authentic work by the artist, and repeatedly published as such by distinguished scholars such as Cornelis Hofstede de Groot and Walter Valentiner, but following Valentiner’s last publication of it in 1931, it was completely ignored by all Rembrandt scholars until 2005, with the exception of Werner Sumowski, who knew it only from an old photograph, and who made passing reference to it in a footnote.

The rediscovery of an authentic painting by Rembrandt is an exciting event in itself, but as Ernst van de Wetering has shown, this picture has furthermore revealed a hitherto unexamined aspect of Rembrandt’s activity: his use of oil studies.  This picture used to be thought of as a portrait, but as Ernst van de Wetering has shown (see Literature below), it is rather a study to investigate the effects of light and shadow upon a woman’s head for which an elderly woman posed.  Martin Bijl’s careful cleaning and removal of overpaintings, thus revealing its original appearance, makes it clear that it is an oil sketch or study, and not a portrait, nor a fanciful imaginary portrait or trony, as it appeared before cleaning.

The following catalogue entry is heavily dependent on Ernst van de Wetering’s unpublished article on this painting and Rembrandt’s oil studies, a copy of which is available on request.  We are most grateful to Professor van de Wetering for granting us permission to use it, and for his help in preparing this text.

In view of the lighting which leaves the woman’s face largely in shadow, it would be difficult to visualize it as a formal portrait.  Moreover, the woman’s clothing betrays her social class: she is a servant and does not belong to the class that customarily had its portraits painted.  She wears a bonnet whose point, normally stretched over the cheek, is folded back in an informal fashion, partially exposing the ear and the metal structure that fixes the bonnet on the head - the  oorijzer, or ‘ear iron’.  Furthermore, the painting has in parts been summarily executed, with sketch-like brushwork.  It would seem, therefore, that it is a study, for which, perhaps, one of the painter’s domestic servants sat as a model.  It is furthermore clear that Rembrandt’s particular interest when painting it was the effect of strong diagonal lighting on his subject, such that it may be considered a study of lighting.  The light falls obliquely from behind, so that the face remains largely in shadow, and is modelled from half-tones. The bonnet, by contrast, catches full light, as does also part of the woman’s ear, which is painted in detail, and also the neck and jaw.

While it is unquestionably entirely Rembrandtesque in character, this painting does not at all resemble Rembrandt’s portraits of women.  Seen in the context of Rembrandt’s oil studies however, it is clear that it finds a natural place in Rembrandt’s oeuvre: stylistic and technical characteristics as well as its masterly quality and brilliancy of execution are decisive in confirming this.

The panel
A dendrochronological examination conducted by Dr. Peter Klein has confirmed that the oak panel on which it is painted was taken from the same tree as the panel for the Self-portrait à la toque from 1633 in the Louvre (Bredius 19) and the panel on which the so-called Portrait of Willem Burggraaff in Dresden (Bredius 175) is painted, a portrait that was likewise produced in Rembrandt’s workshop in 1633, whether or not with Rembrandt’s direct involvement. Similarly, the panel on which the Landscape in the Wallace Collection (Bredius 451) was painted also derives from the same tree. This landscape is considered to be a work by one of Rembrandt’s pupils which, given its affinities with Rembrandt’s landscapes from between 1638 and 1640, must have originated around 1640.

Cleaning and restoration
It was noted that the panel, rounded above, had been in-filled to make it rectangular and both the bottom and left side extended.  The added piece of wood (bearing the Kaunitz stencil; see Provenance below) was removed, but is part of this lot.  During the subsequent cleaning, it was found that the part around and below the head had been completely over-painted: a fur collar had been painted over a wide, white collar, and standing above it a decorated shirt collar, also introduced later. The fur collar had been painted over a black layer covering the original white collar. The background too was almost completely over-painted in order to hide the joins with the added bits of panel (see Fig. 3, showing the present picture before cleaning).  These changes to Rembrandt’s original work reveal a likely intention in over-painting it to create a picture that is more like a formal portrait or a trony.

The cleaning was undertaken in small stages by Martin Bijl, with each stage reviewed carefully in conjunction with Professor van de Wetering, and where needed fellow members of the Rembrandt Research Project, in particular the chemist Karin Groen and the specialist on dress and costume Marieke de Winkel.  Many parts of the picture required little intervention: the white bonnet and the lit parts of the face and neck had survived in largely original condition.  While it was clear that the coarse fur collar was not original, and that evidence for a white collar was provided by X-Rays (see Fig. 7) and by its reflection on the skin of her jaw, it was by no means clear how well the parts underneath had survived, nor how easily the collar could be removed, so the cleaning of it was done in an exploratory way: opening a hole, reviewing the results, and then progressively widening it. Proof that the fur collar was a much later addition was provided by the observation that the underlying white collar had been aged to the extent that a pattern of craquelure had developed before the collar was added. The wet paint of the black underlayer of the fur collar had filled this craquelure.  

Rembrandt’s style and handling of paint
The ear, the flesh of the neck and the cheek and the forehead with its subtle transition from skin to hair, show that wealth of nuances, with an amazingly refined differentiation between warm and cool flesh-tones that are so typical of Rembrandt.   These nuances are introduced with visible brushstrokes that possess in parts - for example in places in the ear or along the jaw - an autonomy that is also characteristic of Rembrandt.   More visual information concerning style and the quality of the work, which are key to the attribution, can be derived from the method of painting applied to the bonnet (Fig. 3, detail of bonnet).  Evident at first glance, in a detail such as the bonnet’s folded back flap, is an interplay that is highly specific to Rembrandt between a loose, graphic handling of the brush (for instance, in the edge of the flap) and a rhythmic, painterly way of working in the folds. The predominantly grainy handling of the paint in the linear passages contributes to the impression of such part being detached from the part lying behind it. The overwhelming impression of such a passage is determined by great pictorial freedom, virtuosity and effectiveness.

In other respects too the painting shows this mixture of extreme refinement coupled with a freedom of execution that is highly characteristic of Rembrandt. For example, the shading of tonal values within the white bonnet is remarkably subtle, paralleled by the differentiation in the substance of the paint and the degree of visibility of the brushstrokes.  In this context, when the back of the bonnet is compared with the bands that run transversely over the woman’s cranium, one is struck by the way that the more distant part has been executed with calm, long brushstrokes in a white which, tonally, almost imperceptibly graduates away from the lighter white applied with a fuller and more agitated brushstroke, with which the bands and the front frill of the bonnet are painted.  The result of all this is a strongly spatial effect, without the viewer being consciously aware of it. What we may observe here is in fact a striking example of Rembrandt’s ability to suggest atmospheric perspective over a very short distance. This same subtle play can also be seen, for example, in the white shirt in the Girl Leaning on a Window Sill from 1645 in the Dulwich Picture Gallery (Bredius 368; see fig. 6), where the sleeve of the arm, situated a little deeper in the space, has a slightly more muted tone in the white and is more ‘calmly’ executed than the sleeve more to the foreground. Rembrandt’s pupil, Samuel van Hoogstraeten, spoke in this connection of the  ‘density of the air that changes the colours’ (Hoogstraeten, p. 264) to which he adds the significant remark that this change  -and thus the difference in colour and tone between two forms, one close behind the other – occurs because ‘even over a slight distance, the air itself has a certain body’ (idem, p. 265).  The interplay between the ‘graphic’ and painterly approaches in the  service of spatial suggestion, so characteristic of the painting discussed here and especially evident in the bonnet, is typical of  Rembrandt’s painting during the period around 1640.  It occurs in such paintings as the Still life with peacocks (Bredius 456), for example, or the Hunter with a dead bittern (Bredius 31),  both from 1639, or in the so-called Saskia in Washington (Bredius 96), a painting which,  because it was done on a poplar panel, must also have been painted in the period around 1640, the only period when Rembrandt apparently received a consignment of poplar panels. The Landscape with a stone bridge (Bredius 440) displays the same mixture of painterly and more graphic techniques. 

The fact that the work is painted entirely `wet in wet’ –  that is to say in paint applied on top of paint that is still wet, the one layer impacting the other - without any trace of correction during its execution, witnesses to the mastery of the painter. This painting could well have been completed in a single day, which would mean that all the extremely fine gradations in white, mixed on the palette, had to fit immediately into the image’s suggested spatiality and plasticity, without correction or revision.  At the back of the woman’s head, Rembrandt has even introduced yellowish tints on the folds of the bonnet, the result of light reflected from the yellow-white collar. Certain minimal details are so typical for him that, for all their modesty they play a role in the attribution to Rembrandt: for instance, the few loose hairs that give the viewer the feeling that he can see all the woman’s hairs separately. Another detail, even more modest, is the spot on the woman’s cheek. While Rembrandt accurately represents his own facial spots - probably pimples – in a number of self-portraits, one does not expect this is in his formal portraits in which, naturally, he must have had to omit any such blemishes.  It is therefore all the more interesting that in this painting, which, as already argued, is not a portrait, he took the trouble to paint this spot.

As noted above, the woman posing for this study must have been someone from a lower class, quite possibly one of Rembrandt’s household servants. The impression of domesticity and casualness is partly created by the woman’s exposed dark oorijzer – the ear iron – clasped to her cheek. From a number of works with women wearing similar bonnets, one can infer that the bonnet’s side-flap, folded up above the ear in the present painting, was normally folded down and stretched over the end of the oorijzer, with only the decorated button remaining visible. There is another painting in which an almost identical oorijzer can be seen, with the point of such a bonnet raised in identical fashion. This is a painting by Philips Koninck, in which a young girl, clearly shown as a servant-girl, is busy with some domestic task (see Fig. 5).  Quite a number of such oorijzers, or fragments of them, have been found during excavations in Amsterdam, often in old cesspits. In every case they turn out to be made of copper, despite their name. The relatively wide metal band, always with some simple decoration at the end, was bent backwards above the ear to run as a thinner, springier band beneath the bonnet, around the back of the head below the obligatory hair knot, to the other ear.  Where the oorijzer was bent at the temples, there were elongated holes through which apparently ran a ribbon that passed under the front of the bonnet. The oorijzers worn by well-to-do women who sat to Rembrandt, such as Mrs. Anslo (Bredius 409)  and Catharina Hoogsaet (Bredius 391) were probably made of silver or gold and expensively decorated with a pearl or a gold button. Where the dress of the young girl in Philips Koninck’s painting is clearly described, the painting under discussion here ends below in an extremely summary indication of the woman’s torso.

Rembrandt and oil sketches or studies
We are used to thinking of Rembrandt as an exceptionally free artistic personality, and consequently almost modern in this regard, or at any rate as an early Romantic. Who else was so self-willed, who left so many works almost nonchalantly unfinished, who threw himself with such abandon into what we long took to be highly personal projects: for example, his many self-portraits?  With patient study of his oeuvre, however, a whole range of different functions emerges for different categories of works, not least the many, ostensibly freely sketched works. For example, Rembrandt painted a number of oil sketches  preparatory to undertaking ambitious etchings, partly, between 1632 and c.1635, for a Passion series and later, in three cases, with an eye to etched portraits. In these often remarkably sketchy works the composition as a whole was executed in oils, in a number of cases en grisaille, on the basis of which Rembrandt – or, more likely, others – could subsequently work up etching plates with the ultimate aim of producing prints.  Rembrandt also painted other preparatory oil-sketches, or more properly, oil studies. These concern heads or half-figures which, may well have served as preparatory exercises for figures in a larger composition. This category of works has been rather neglected in the research of Rembrandt’s oeuvre.

Lighting Studies
As Ernst van de Wetering has related (see Literature below), investigation of Rembrandt’s self-portraits published in Volume IV of A Corpus of Rembrandt Paintings led him and his colleagues to the suspicion that some of Rembrandt’s paintings could be considered as studies for figures in larger compositions.  For example, the hypothesis arose that Rembrandt’s head in the Amsterdam Rijksmuseum could have been a preparatory study, painted in front of the mirror, in connection with the Repentant Judas (Bredius/Gerson 539A) dated 1629. In that painting two figures are similarly lit obliquely from behind. The fact that the small Amsterdam painting was not signed and was copied remarkably often (in Rembrandt’s workshop, it would seem), suggests that it was a studio prop and was never intended for sale. Rembrandt evidently sometimes used himself as a model to try to solve particular pictorial problems in front of the mirror.  In this case, it was the problem of the figure lit obliquely from behind, in fact a problem that continued to engage Rembrandt in one form or another throughout his life.  That problem undoubtedly arose as a consequence of Rembrandt’s need to convey as lifelike as possible interactions between the figures in his history pieces and in his double- or group-portraits; this necessarily entailed the actors in such scenes being placed opposite each other, which meant that one or more figures had to be situated almost with their back to the incident light. Besides the Judas-painting mentioned above, for which Rembrandt must have made the study in front of the mirror, the painting with Two old men disputing (Peter and Paul) in Melbourne is also a striking example of this (Bredius 423). Rembrandt was a master in introducing reflected light in the shaded areas. A striking example of such treatment of light with figures lit from an oblique light source is the girl to the left in the Braunschweig family portrait from Rembrandt’s late period.  Rembrandt’s pupil Samuel van Hoogstraten defined the phenomenon of light reflected into shadow as follows:  Reflection is actually a casting back of light from all lit objects, but in art we only use the term  reflection to refer to the secondary illumination that falls in a shadow, and elsewhere he writes:  …our Rembrandt has acquitted himself wonderfully in reflections, yes, it seems as though this choice for the secondarily thrown, faint light was his proper element… (Hoogstraeten, p. 273)   In the course of his career, Rembrandt would increasingly use the reflection of light into passages in shadow where figures were lit from behind.  One could speculate about the possibility that the Study of a woman in a white cap was made in preparation of a double- or group portrait.

Ever since Hoogstraeten wrote these words, Rembrandt has been admired for his handling of half-tones, and the face of the woman in the present painting, partly in deep shadow (for example the base of her nose), partly lit by reflected light (for example the lower part of her nose), with gradations of shadow and indirect light in between, is an excellent example of his mastery of this.  In the present study, however, these half-tones are there because the side and back of the woman’s head as well as her cheek and the side of her neck are in strong diagonal light, as in the Melbourne figures and in many subsequent paintings by Rembrandt.  It may seem to us that Rembrandt achieved these effects effortlessly, but he must have had to work them out, and the present work, albeit executed with an unhesitating fluency that belies the mental process of working out an idea, is nonetheless an exercise in the understanding of light and its handling in paint.

Study of an Elderly Woman in a White Cap and Rembrandt's Anslo portrait
We would not expect Rembrandt to have painted anything without a reason, and it is a reasonable assumption that the present picture had a particular purpose.  Professor van de Wetering has suggested that it might have been made in connection with one of Rembrandt’s most unusual projects from this period, the highly ambitious Double portrait of Dominee Anslo and his wife in Berlin (Bredius 409; R.R.P. vol 3, pp. 403-415, no. A 143, reproduced; here reproduced fig. 9). This monumental painting is unusual in its composition as a whole, and in particular in the placing of the two figures within the pictorial space and in relation to each other. The entire left half of the composition is taken up by a highly complex still-life with books and a candelabra on a table covered with a Persian rug.  Rembrandt’s surprising invention is to place the two portrayed subjects, the famous preacher, Cornelis Claesz. Anslo and his wife, Aeltje Gerritsdochter Schouten, close together in the other half of the composition. However, this particular and unusual composition is not unique in Rembrandt’s oeuvre: it was pointed out in the R.R.P. vol. III that there is a striking similarity in compositional terms between the Anslo-double portrait and one of the paintings mentioned above, the Old men disputing (Peter and Paul) in Melbourne from c. 1628.  They are in broad compositional terms mirror images of each other: in the painting in Melbourne it is the left half of the composition that is taken up by the two figures, while in the right half a complex still-life with books and with a candelabra can be seen.  In both, however, the incident light comes from the left.

It is in the lighting of the figures that the dilemma that presented itself to Rembrandt with the Anslo-double portrait becomes clear. In the painting in Melbourne, the figure of Paul – who is apparently speaking and indicating with his forefinger a passage in the book on Peter’s lap  - is placed in full lighting, and only the right side of his face is slightly in shadow. Peter, on the other hand, is lit from behind, so that his face remains entirely in shadow. In the Anslo painting the situation is reversed. The face of Aeltje is placed in full light, whereas by contrast  the face of the main personage in the painting, Cornelis Anslo, is very largely cloaked in shadow, although the right side of his face and his gesturing hand a re strongly lit from the side.

Profesor van de Wetering has speculated that Rembrandt may have based the composition of the Anslo double portrait on the scheme of the Peter and Paul.   Anslo, the chief personage, would then, like Paul, have been most favourably placed with regard to the incident light, while his wife would have been partially lit from behind. In the final version so familiar to us, the reverse is of course the case.  Had Rembrandt intitally approached the composition the other, more conventional and more logical way round however, the head of Aeltje would plausibly have been lit in the same way as in the present sketch, in which case its peparatory role would appear much clearer. It is in any event certainly possible that the present oil study was painted in connection with the Anslo portrait, and may have had a role in resolving Rembrandt's dilemma over the placing of the two figures in relation to each other and to the source of light.  In the eventual choice of the placing of the wife, the commissioning patron would undoubtedly have played a rôle, and the present study may have played a role in reaching this decision.

Only one has come to light: in the Hermitage, St. Petersburg (oil on canvas, 53 by 44 cm.) for this information we are indebted to Irina Sokolowa.  It replicates the present work in its enlarged and fully overpainted phase, and may perhaps date from Prince von Kaunitz' ownership.  The Hermitage picture was in the collection of Baron Alexander Stieglitz, St. Petersburg, 1869, when it appears in a watercolour by Premazzi of the interior of his St. Petersburg mansion.

Fürst Wenzel Anton Kaunitz was one of the principal figures of the Austrian Enlightenment.  He may perhaps have started to collect pictures during his diplomatic mission to Brussels between 1744 and 1746, but visits to Brussels and Antwerp in 1776 and 1777 confirmed his passion for Rubens, and in these years he bought works by that Master for the Imperial collection.  He had between 1770 and 1773 supervised the newly organised Akademie der bildenden Künste, and as well as a notable collection of pictures, amassed an enormous cabinet of prints.  After his death his collection remained in his family in the Kaunitz Stadtpalais in Dorotheergasse until dispersed by his nephew Fürst Aloys Kaunitz in a sale held by Artaria in Vienna on 13th March 1820.  The present picture, lot 132, was one of two Rembrandts, and several Rembrandt School pictures, in the sale.

For a condition report on this painting please see back of this catalogue.

Hoogstraeten:            S. van Hoogstraeten, Inleyding tot de hooge schoole der schilderkonst, Rotterdam 1678
Bredius:                      A, Bredius, The paintings of Rembrandt, first edition in German, Vienna 1935, revised by H. Gerson, London 1969
R.R.P. Vol. III:             J. Bruyn et al, (Stichting Foundation Rembrandt Research Project),  A Corpus of Rembrandt Paintings, Vol. III, Dordrecht/Boston/London 1989.