Woman Bathing: the lost Van Eyck image and its surviving witnesses

Woman Bathing: the lost Van Eyck image and its surviving witnesses

Woman Bathing is the conventional name for a lost composition associated with Jan van Eyck. Its two principal visual witnesses are a small later panel at Harvard and a picture shown on the wall of Willem van Haecht’s 1628 Gallery of Cornelis van der Geest. KIK-IRPA identifies the image in that Antwerp collection scene as a missing Van Eyck. Neither surviving witness should be presented as the original painting.^[2]

Harvard now calls its panel Woman at her Basin, retaining Woman at Her Toilet as a former title. The attribution remains “(?) After Jan van Eyck”, and the museum dates it to the early sixteenth century. The qualification is essential: an unknown artist transmitted a composition thought to derive from Van Eyck, whose own version is unavailable for comparison. The lost design and the surviving copy have different histories.^[1]

Surviving witnesses

What the surviving copy shows

The Harvard panel brings two women close together in a narrow domestic room. The nude woman stands beside a basin and holds a small light cloth before her body. Her companion wears a long red dress and a white head covering. A window lights the scene from the left; a convex mirror hangs nearby. The room offers enough furniture and utensils to make washing a credible activity, while the unequal states of dress give the encounter an unusual intimacy.^[1]

The composition directs attention through a sequence of surfaces. Pale flesh stands out against the dark background, the companion’s red dress forms a broad vertical, and the small basin supplies a dark oval near the nude woman’s hand. The mirror contributes a second, compressed arrangement of the figures. It makes the room appear larger than the portion directly visible and brings a different angle of looking into the same narrow panel. These are observations about the surviving image, before any decision about a biblical identity.

Small objects keep the scene from becoming a study of bodies alone. The companion carries a round object and a transparent bottle; other items occupy the window area and the furniture beside the basin. Their closeness to the women makes them part of the action of preparation. A bottle can establish a practical domestic activity without proving what liquid it contained. The visible container supplies evidence; a named perfume, medicine or love potion would be a further hypothesis.

The contrast between nakedness and elaborate clothing has consequences for the viewer. One woman is physically exposed, while the other remains enclosed in a full dress and headdress. Their reflected images introduce another way of seeing them, but do not eliminate that contrast. The companion’s presence also makes the scene a social encounter rather than an isolated nude posing in an empty room. The identities of the women and the relationship between them remain unsettled.

The copy’s date and material history

The surviving panel is oil on oak, measuring 27.2 × 16.3 × 0.7 centimetres, and bears Harvard’s accession number 1969.83. Those measurements describe the Cambridge object. They cannot be transferred to the vanished prototype, nor can its early sixteenth-century date become the date of a painting by an artist who died in 1441. A small copy might preserve a much larger composition, but the scale of that reduction needs independent evidence.^[1]

Technical study has helped separate the objects. Hélène Verougstraete’s review of Teri Hensick’s conservation study reports that Peter Klein’s tree-ring analysis placed the Harvard painting after 1511, replacing an earlier fifteenth-century dating. This is a reported conclusion of the technical literature, rather than a newly examined tree-ring sequence here. It gives a useful boundary: a sixteenth-century copy can preserve an earlier design, but cannot be an autograph panel executed during Van Eyck’s lifetime.^[4]

Noëlle Streeton likewise includes Woman at Her Toilet among compositions that studies have shown were copied from earlier models. The distinction between a contemporary workshop repetition and a later copy matters. Similarity of subject and arrangement does not make every surviving version the product of the master’s own studio. The evidence supports transmission of a design; it does not recover Jan’s individual brushwork from the Harvard paint surface.^[5]

The image in an Antwerp collection

Van Haecht’s Gallery of Cornelis van der Geest puts the lost composition into a different kind of record. On the right wall, above the classical figures and near the doorway, a small framed interior contains the nude woman and her red-clothed companion. The image belongs among dozens of pictures, sculptures and objects assembled to celebrate a collector. Its importance lies partly in its company: an early Netherlandish painting is presented within a seventeenth-century culture of connoisseurship.^[3]

The owning museum identifies 1628 as the gallery picture’s date. The princely visit it represents occurred thirteen years earlier, in 1615. These are different moments. The painting arranges people and artworks into a commemorative whole; it was not painted as an instantaneous visual inventory during the visit. Its inset image can preserve a composition without proving precisely where the lost panel hung, how large it was, or which details were legible to a visitor in 1615.^[2][3]

Recent conservation makes that mediation especially clear. KIK-IRPA’s 2019–2021 investigation found evidence that Van Haecht used a grid when scaling small details. It also recovered information about changes to the gallery painting’s support and composition. These discoveries concern the artist who painted the collection scene. They explain how a later witness was made and altered, while offering no direct analysis of the missing Van Eyck’s pigments or underdrawing.^[2]

Watch: Jan van Eyck in context

Choose a Smarthistory close reading of the Arnolfini Portrait or the closed and opened Ghent Altarpiece.

Jan van Eyck, The Arnolfini Portrait

Smarthistory

Look closely at the room, mirror and gestures in the London double portrait.

Watch on YouTube

▶01 · PortraitThe Arnolfini PortraitSmarthistory▶02 · ExteriorThe Ghent Altarpiece, closedSmarthistory▶03 · InteriorThe Ghent Altarpiece, openSmarthistory

The two visual witnesses therefore offer complementary evidence. The Harvard panel permits close inspection of a freestanding version. Van Haecht places the composition within a collector’s display. Their shared arrangement is significant, yet neither supplies the lost original’s undamaged surface. Reading them together allows more confidence about the broad design than about every small object, colour or reflection within it.

The 1668 sale and the claim about Van Eyck’s wife

A French-language study by S. Speth-Holterhoff reproduces a particularly revealing entry from the 13 August 1668 sale of the Antwerp collector Pierre Stevens. Number 3 describes a celebrated bath picture in which Van Eyck supposedly represented his wife naked and clothed. The entry also claims an earlier connection with the Duke of Urbino. This explains the historical basis of an identification that can otherwise sound like a modern guess.^[6]

Speth-Holterhoff questioned both the sitter claim and the proposed Italian provenance, suggesting that different bath pictures had become confused. A sales catalogue records what was said about an object at the time of sale; its wording does not establish that a fifteenth-century painter intended two figures to be portraits of one woman. The claim about Van Eyck’s wife consequently belongs in the history of interpretation, alongside the record of Stevens’s collection, rather than functioning as a secure identification of Margaret van Eyck.^[6]

Harvard’s own modern provenance is a separate history. The panel passed through the collection of Augustus John and subsequently through dealers in London and Boston before its 1969 acquisition by the Fogg. The museum qualifies an intermediate ownership date in its notes. This chain describes the small surviving panel, and does not connect it continuously with Stevens or with Van der Geest’s collection.^[1]

Does Facio describe the same painting?

The earliest literary evidence takes us back to the 1450s. In his De viris illustribus, Bartolomeo Facio, also called Fazio, praised a Van Eyck painting of women emerging from a bath. The passage describes partial coverings, a mirror that reveals a woman’s otherwise hidden back, an apparently burning lamp, a perspiring older woman and a dog drinking water. It then opens into a landscape of small people, horses, mountains, woods and distant settlements. This is a much more expansive description than a generic reference to a nude.^[7]

The account’s geography creates the main difficulty. The Harvard version presents a close interior, whereas Facio praises extensive distance. The number and activities of the figures also do not fit neatly. A later copy could omit something, and a written description could emphasise one part of a composition, but neither possibility proves a match. The two kinds of evidence must be tested against one another rather than joined because both involve a bath and a mirror.^[7]

France’s Institut national d’histoire de l’art usefully distinguishes the two records in its artist entry: Femmes au bain, associated with Ottaviano della Carda and Facio’s testimony, and Femme à sa toilette, represented in Van Haecht’s gallery. This does not settle every historical question, but it prevents an attractive title from doing the work of a provenance document. An Italian literary witness and an Antwerp visual witness need not describe one travelling panel.^[9]

Facio’s praise also has a literary dimension. In her French-language study, Mathilde Bert relates his descriptions of bodily modesty, sweating and landscape to ancient ways of praising art, particularly the language of Pliny. She also discusses the mirror in relation to the later comparison between painting and sculpture: a painted reflection can offer several views at once. This reading makes the text valuable for the history of artistic admiration, while reminding us that it is an eloquent humanist account rather than a neutral condition report.^[8]

Judith, Bathsheba or a scene of personal care?

The descriptive titles leave the subject open. Peter Schabacker proposed Judith preparing herself before her encounter with Holofernes, following the account in Judith 10. His published abstract also suggests the bathing apartments of the Prinsenhof in Bruges as a possible original setting. Both are proposals. An association between washing, adornment and Judith’s preparation can make sense without demonstrating that a particular panel decorated those apartments.^[11]

The alternatives change how the same visible details are understood. A biblical reading makes the washing part of a particular woman’s story; a moral interpretation can turn the mirror and personal care towards Vanity. A secular reading can attend to beauty, intimacy and the painter’s skill at representing flesh and reflection. The image’s ambiguity is therefore substantive, rather than an inconvenience to be removed by choosing the most familiar name.

Judith’s maid and bathing provide a plausible comparison, but the composition does not show the later beheading of Holofernes. Bathsheba’s story likewise involves washing, yet naming her would require a persuasive connection beyond that shared activity. A title such as Woman at her Basin preserves what is visible without assigning the action to an episode the surviving image cannot conclusively identify. This makes it particularly appropriate for a later witness to a lost painting.

The comparison with the Arnolfini Portrait

The domestic interior naturally invites comparison with the Arnolfini Portrait. Hugh Hudson draws attention to specific correspondences and differences: the placement of the mirror beside the window, the folding chair, and vessels associated with washing. He also records disagreement about whether the similarities prove a deliberate relationship. A visual vocabulary can recur without establishing that two paintings were made for the same patron.^[10]

Some interpretations go further, proposing that the bathing picture served as a cover or companion to the London portrait. Hudson discusses such suggestions while noting objections. Their practical implications matter: a cover would have to relate physically to another panel, not merely share an atmosphere or motif. No surviving joinery or securely measured original is available to demonstrate that relationship. The Harvard copy cannot by itself supply the dimensions and construction of a missing cover.^[10]

The lost work’s value extends beyond resolving this comparison. A room containing a nude woman, a companion and an active mirror broadens the kinds of imagery associated with Van Eyck. Its later copies and written traces show how such an image could attract curiosity long after its first setting had disappeared. The strongest account keeps each witness recognisable: a surviving sixteenth-century panel, a seventeenth-century gallery image, a sale description and an earlier literary celebration whose exact connection remains uncertain.

Related reading

The Arnolfini Portrait

Ghent Altarpiece

The Annunciation in Washington

References

1. Harvard Art Museums/Fogg Museum, Woman at her Basin, object 1969.83, formerly Woman at Her Toilet; current title, qualified attribution, materials, dimensions and provenance, record refreshed 28 October 2026. View Source
2. KIK-IRPA, The Gallery of Cornelis van der Geest Gets a New Lease of Life; Van Haecht’s 1628 painting, the visit thirteen years earlier, the missing Van Eyck and 2019–2021 investigation. View Source
3. Rubenshuis, Willem van Haecht, The Gallery of Cornelis van der Geest; owning collection, display context and public-domain whole-composition photograph. View Source
4. Hélène Verougstraete, review of Faries and Spronk, Recent Developments in the Technical Examination of Early Netherlandish Painting, sehepunkte 5:2 (2005), p. 3 of PDF; reports Teri Hensick’s study, pp. 83–95, and Peter Klein’s post-1511 dating. The review, not the full technical chapter, was consulted. View Source
5. Noëlle L. W. Streeton, “Emulating Van Eyck: The Significance of Grisaille,” CATS Proceedings I (2012), p. 29 and note 5; distinction between contemporary replication and later copying. View Source
6. S. Speth-Holterhoff, “Trois amateurs d’art flamands au XVIIe siècle,” Revue belge d’archéologie et d’histoire de l’art 27 (1958), especially pp. 53–56; transcription of the 1668 sale entry and its critical interpretation. View Source
7. Bartolomeo Facio, De viris illustribus, Van Eyck passage, Latin transcription in W. H. James Weale, Hubert and John Van Eyck: Their Life and Work (1908), pp. lxxiii–lxxiv; compare Weale’s English account pp. 175–176. Obsolete owner identifications in that commentary are not adopted. View Source
8. Mathilde Bert, “Pline l’Ancien et l’art de la Renaissance: balises pour une étude de réception entre le Nord et le Sud,” Revue belge d’archéologie et d’histoire de l’art 75 (2006), pp. 34–36 and notes 96–103; Facio’s rhetoric and separate bathing-picture witnesses. View Source
9. Institut national d’histoire de l’art, AGORHA, Jan van Eyck artist record, entries for the lost Femmes au bain and Femme à sa toilette. View Source
10. Hugh Hudson, Jan van Eyck, The Ince Hall Virgin and Child, and the Scientific Examination of Early Netherlandish Paintings, revised 2009, pp. 180–182 and notes 138–144, pp. 203–204; RKD-hosted study, comparisons and disputed links. View Source
11. Peter H. Schabacker, author’s abstract of “Jan van Eyck’s Woman at Her Toilet: Proposals Concerning Its Subject and Context,” Getty Research Institute RILA/OpenBibArt; Judith and Prinsenhof hypotheses. The 1974–1976 paper with Elizabeth Jones was not available in full. View Source