The Wild Sea

Margaret van Eyck, a house called ‘The Wild Sea’ and Jan van Eyck’s posthumous workshop Understanding of the transition in painting in Bruges between the death of Jan van Eyck in 1441 and the early career of Petrus Christus has been hampered by uncertainty about the size, location and fate of Van Eyck’s workshop, which some scholars argue was continued by his widow, Margaret. Reassessment of the evidence suggests a new terminus ante quem for its closure. by jan dumolyn, susan frances jones, ward leloup, toon de meester and mathijs speecke emarkably little is known about Jan van Eyck’s workshop, including its size, location and date of closure. It is widely accepted that the workshop would not have stopped trading immediately on Van Eyck’s death, which seems to have occurred towards the end of June 1441; however, it is not clear how long it remained operative or who took over the workshop’s management, whether his widow Margaret, his surviving brother, Lambert, or a trusted workshop assistant. Before the 1990s it was believed that Petrus Christus (d.1475/76), Van Eyck’s principal follower in Bruges, had been a member of his workshop, but this idea was later rejected, and the nature of their relationship remains obscure, although it was clearly fundamental to Christus’s artistic identity. All these problems impede proper understanding of the transition in Bruges painting between Van Eyck’s workshop of the 1430s and the subsequent R This article is a collaborative effort: Jan Dumolyn, Ward Leloup, Mathijs Speecke and Toon De Meester conducted new archival research on The Wild Sea house (De Wilde Zee), Susan Frances Jones addressed the evidence for the various paintings referred to in the article and both Jones and Dumolyn re-examined the evidence for Jan van Eyck’s workshop and family. The authors are very grateful to Lorne Campbell and Catherine Reynolds for reading a draft of this article and making improvements to the text. 1 For a useful overview of methodological approaches to late medieval and early modern workshops, see S. Cassagnes-Brouquet: ‘Les ateliers d’artistes au Moyen Âge: entre théorie et pratiques’, Perspective: Actualité en histoire de l’art 1 (2014), pp.83–98. A foundation for a complete visual and technical reassessment of Jan van Eyck’s œuvre has been created by the VERONA project (Van Eyck Research in OpeN Access), based at the KIK-IRPA (The Royal Institute for Cultural Heritage) in Brussels, for which, see http://closertovaneyck.kikirpa.be/, accessed 11th January 2022. 2 The documents are in Stadsarchief, 120 Bruges (hereafter cited as SAB), Oud Archief (OA), 216, Rekeninghe van de tresoriers van der stede van Brugge van 2 September, 1431, tot 1 September, 1432, fol.78r; and Archives Départementales du Nord, Lille, Série B, 1948: Compte de la recette générale des finances du 1 Janvier au 31 Décembre, 1433, fol.viijxxviijv. They are published in W.H.J. Weale: Hubert and John van Eyck: Their Life and Work, London 1908, pp.xxxviii, no.18 and xxxix, no.20; and J. Paviot: ‘La vie de Jan van Eyck selon les documents écrits’, Revue des archéologues et historiens d’art de Louvain 23 (1990), pp.83–93, p.88 and notes 36 and 37. Since the amounts of both gratuities are divisible by five, Campbell suggested that Van Eyck had five assistants at that time, see L. Campbell: National Gallery Catalogues: The Fifteenth Century Netherlandish Schools, London 1998, p.23. For the idea that Van Eyck had twelve assistants, see, for example, T.-H. Borchert: Jan Van Eyck, Hong Kong, Cologne, London, Los Angeles, Madrid, Paris and Tokyo 2008, p.69; idem: ‘The Ghent Altarpiece and the workshop of the Van Eyck brothers’, in S. Kemperdick and J. Rößler, eds: Der Genter Altar – the burlington magazine | 164 | february 2022 phase, dominated by Christus. This article seeks to clarify this short but complex period. It begins by reassessing documentary evidence for the number of assistants in Van Eyck’s workshop and proceeds to reconsider the question of the posthumous workshop, presenting new research on Jan’s widow, Margaret, and a house in Bruges she is reputed to have owned, called The Wild Sea (De Wilde Zee). As such, it addresses the topic of Van Eyck’s workshop partly from the point of view of its material structure and physical location.1 Attempts have been made to extrapolate the number of assistants in Van Eyck’s workshop from records in account books of gratuities given to them in Bruges on two separate occasions in 1432, one by the two burgomasters of Bruges along with members of the city council and the other by Van Eyck’s patron Philip the Good, Duke of Burgundy on a visit to Van Eyck’s ‘hostel’, most probably the workshop at his house.2 Reproduktionen, Deutungen, Forschungskontroversen, Petersberg 2017, pp.158–61, p.158 and note 4. 3 For a useful look at categories of assistant and relevant terminology, see H. Deceulaer and A. Diels: ‘Artists, artisans, workshop practices and assistants in the Low Countries (fifteenth to seventeenth centuries)’, in N. Peeters: Invisible Hands?: The Role and Status of the Painter’s Journeyman in the Low Countries c.1450–c.1650, Leuven 2007, pp.1–32, at pp.13–22. 4 For the document, see Weale, op. cit. (note 2), p.xxxix, no.20; and Paviot, op. cit. (note 2), p.88, notes 36 and 37. The visit probably took place late in 1432, between September and December. There has been some debate about whether the duke went to Van Eyck’s premises or vice versa. J. Duverger: ‘Jan van Eyck as court painter’, The Connoisseur 194 (March 1977), pp.172– 79, at pp.176–77, thought that the term ‘son hostel’ implied the duke’s palace in Brussels, but Paviot has shown that the painter’s workshop at his house is more likely. For the money of account used in these documents, see P. Spufford: Monetary Problems and Policies in the Burgundian Netherlands, 1433– 1496, Leiden 1970, p.17. 5 For the document on Van der Weyden, see J. Guillouet: ‘Deux volets peints par van der Weyden pour l’abbaye Saint-Aubert de Cambrai’, Bulletin de la Commission historique du Département du Nord 47 (1993), pp.9–17, p.14. 6 The authors are grateful to those who generously lent their expertise on this question: Erik Aerts, Professor Emeritus at the KU Leuven; and the late Peter Spufford, formerly Professor Emeritus of European History at the University of Cambridge, who both favoured the idea that standard silver coins would have been used. 7 The kromstaart was valued at 2 groats in the pound Flemish, or pound groat; in the pound of 40 groats, it was valued at 1 shilling. See E. Aerts and H. Van der Wee: Vlaams-Brabantse muntstatistieken 1300–1506. Deel 1. De aanmuntingsgegevens van de zilvermunten, Leuven 1980, pp.58–59. For further details regarding these documents, see J. Dumolyn, N. Geirnaert and M. Speecke: ‘“Giovanni da Bruggia”: Jan van Eyck in Brugge (1425– 1441)’, in T.-H. Borchert, ed.: Jan van Eyck in Brugge, London forthcoming. According to the accounts, the tips were given to Van Eyck’s ‘cnapen’ (in Dutch) or ‘varlets’ (in French). A systematic linguistic study of these terms and their usage in guild records shows that they are generic terms that can mean either apprentice boy or journeyman. In account books, where the precise status of these persons of lesser importance was no issue for the clerks registering such gratuities at the end of a year by copying a pile of receipts, the distinction would not have been made. This article therefore will use the anachronistic but pragmatic translation ‘workshop assistants’ to include the two possible categories.3 The documents in each instance do not record the denomination or number of the actual physical coins that were distributed but rather the equivalent value in money of account: the clerk who recorded the expenses in the city accounts, which use the pound Flemish or pound groat, valued the tip given by the city officials at 5 shillings of Flemish groats (or 60 groats); the accounts of the Burgundian recette générale, which for the most part use the pound of 40 groats, valued the duke’s tip at 25 sols (shillings).4 Gratuities might sometimes be given in gold coins – Rogier van der Weyden’s wife and his ‘ouvriers’, for example, were given a tip in écus d’or when they delivered a retable to Cambrai in 1459 – but gold coins are usually explicitly recorded in accounts, specifying both the type of coin and its value.5 Where such information is not provided, it is reasonable to assume that the tip was made in silver coins, which were in common use. It is most likely that the coins used for these tips to the assistants in 1432 – the year before Philip the Good introduced a unified system of coinage in his disparate territories – were the silver double groats known as kromstaarten (because the obverse showed a lion rampant with a curved tail, or ‘gekromde staart’).6 The number of coins that were distributed is clear from the value of the kromstaart in the two different monies of account: in short, the city government would have given out 30 kromstaarten, and the duke of Burgundy a total of 25.7 If, as appears likely, ‘round numbers’ of coins such as five or ten would have been distributed, this could give us five assistants (each of whom received five coins from the Duke), but it also allows for six (each of whom received five coins from the city government), or even only three (who would have been handed ten coins each from the city). The number of workshop assistants could also have changed between the first occasion and the second. The question of the workshop’s size or organisation can also be investigated through other kinds of document. In a decree of 1441 and a related lawsuit still ongoing in 1487, the master painters of Bruges sought to reduce competition from those in Sluis by regulating the number of ‘dieninghe’ – the latter could use to two or three at most – with ‘dieninghe’, in 1. The Virgin and Child with St Barbara, St Elizabeth and Jan Vos, by Jan van Eyck, possibly with workshop. c.1441–43. Oil on masonite, transferred from panel, 47.3 by 61.3 cm. (Frick Collection, New York). the burlington magazine | 164 | february 2022 121 Jan van Eyck’s posthumous workshop 2. St Jerome in his study, by the workshop of Jan van Eyck. c.1442. Oil on linen paper on panel, unframed 20.6 by 13.3 cm. (Detroit Institute of Arts; © KIK-IRPA, Brussels). this context, apparently signifying two journeymen and one apprentice.8 From this we can infer that Bruges masters regularly employed two or more journeymen themselves: a reasonable estimate might be a workshop of three or four journeymen and an apprentice. Other South Netherlandish masters are recorded using similar numbers of journeymen: in 1454, for example, the painters Jacques Daret (c.1404–c.1470) and Daniel de Rijke (active 1440–82) worked on preparations for court festivities with four and three journeymen respectively.9 Furthermore, the complex 8 The authors thank Catherine Reynolds for bringing this lawsuit to their attention. For the decree of 5th November 1441, see L. Gilliodts-Van Severen: Inventaire des archives de la ville de Bruges, Section Première, Inventaire des chartes, Bruges 1871– 85, V, pp.231–51, at p.248; for the arbitral proceedings, W.H.J. Weale: ‘Inventaire des chartes et documents appartenant aux archives de la corporation de S. Luc et S. Eloi à Bruges’, Le Beffroi: Arts, Héraldique, Archéologie 1 (1863), pp.214–20. On the dispute, see also C. Reynolds: ‘Illuminators and the painters’ guilds’, in T. Kren and S. McKendrick, eds: exh. 122 cat. Illuminating the Renaissance: The Triumph of Flemish Manuscript Painting in Europe, Los Angeles (J. Paul Getty Museum) 2003, pp.15–33, at p.17. 9 L. Campbell: ‘The early Netherlandish painters and their workshops’, in D. Hollanders-Favart and R. Van Schoute, eds: Le Dessin sous-jacent dans la peinture, Colloque III, 6–7–8 septembre 1979, Le problème Maître de Flémalle-Van der Weyden, Louvain-la-Neuve 1981, pp.43–61, at pp.49–50. 10 Only days after the sentence two members of Campin’s workforce (including Rogier van der Weyden) the burlington magazine | 164 | february 2022 documentation concerning the Tournai painter Robert Campin indicates that when he was sentenced to a year’s banishment on 30th July 1432 he was working with three or four journeymen.10 These sources on painters are broadly in keeping with studies by socio-economic historians on workshop sizes in the artisanal world of the Southern Low Countries of the period, as well as other trades, which have always emphasised the importance of small and medium-size workshop.11 In this context, it is also worth remembering that earlier in his career, when he had a permanent position in the service of John of Bavaria, Count of Hainaut-Holland, Van Eyck was also working with only a few assistants, as was common in his day. Accounts of the comital treasury running from 1422 to 1424 show regular payments to Van Eyck that mostly appear under the rubric ‘Pantgelt’.12 In 1422/23 the accounts record a payment to Van Eyck and one assistant (‘Jan den maelre [. . .] ende sijnen knecht’); in 1423 there are payments to ‘Johannes die scilder ende sijnre knechten’, indicating more than one assistant, and in 1423/24 two assistants are specified: one ‘who works with him’ (‘sinen knecht die met hem werct’) and ‘another assistant’ (‘een ander knecht’).13 Whether these documents record all of the assistants Van Eyck employed at the time is open to question, but whether or not that is the case, they suggest a relatively small-scale operation. Taking all this into account, it is reasonable to assume that in the 1430s Van Eyck had somewhere between two and six assistants, perhaps changing according to the circumstances of production or the kinds of work executed. All this evidence makes it unlikely that Van Eyck’s workshop exceeded the average size of a master painter’s workshop of the period by any great degree. It is true that his position of valet de chambre to the duke of Burgundy exempted him from guild regulation, but this did not mean that he had special freedom to employ more journeymen than did guild members, as the Bruges guild of image-makers and saddlers, to which the city’s painters belonged, set no limit on the number of journeymen that a master painter might employ (although it did restrict master painters to training only one apprentice at a time).14 In addition to his workforce, Van Eyck may also have received visiting painters who came to learn, potentially including Southern European painters who required training in the oil medium. One possible candidate is Lluís Dalmau, a painter active in Valencia who was in Flanders between 1431 and 1436.15 The evidence that Van Eyck’s workshop continued to function after his death relies on the dating of just two or three paintings and is more fragile than often assumed. Documents published by Hendrik Jan Joseph Scholtens in the 1930s allow for the possibility that the Virgin and Child with St Barbara, St Elizabeth and Jan Vos (Fig.1) was painted or completed between 30th March 1441 and 3rd September 1443 and thus almost entirely after Van Eyck’s death; however, they do not in themselves exclude production enrolled in the guild as independent masters; a third did so on 18th October. All three had registered in Campin’s shop as ‘apprentices’ in 1427, but in actuality they were probably journeymen. A fourth member of the workforce, who registered as an apprentice in 1431, never became a free master at Tournai, see A. Châtelet: Robert Campin: Le Maître de Flémalle. La fascination du quotidien, Antwerp 1996, pp.26–29. 11 See for example Deceulaer and Diels, op. cit. (note 3), pp.9–13. See also N. Peeters and J. Dambruyne: ‘Some introductory remarks on journeymen in painters’ workshops in the Southern Netherlands c.1450–c.1650’, in Peeters, op. cit. (note 3), pp.ix–xxiv, at p.xviii, citing a hypothesis of M. Martens and N. Peeters that the average workshop in Antwerp between 1500 and 1579 had between five and seven people, comprising the master, apprentices and journeymen. 12 A.-M.J. van Egmond: ‘Materiële representatie aan het Haagse hof 1345–1425’, unpublished PhD thesis (University of Amsterdam, 2019), pp.235–38. This is an elaboration of idem: ‘Dirc die maelre en Jan van Eyck: Een ambachtsman en een kunstenaar Jan van Eyck’s posthumous workshop themselves exclude production in the 1430s.16 The date ‘1442’ inside the pictorial field of St Jerome in his study (Fig.2) may well be the date either of the painting’s execution or of its completion. It does not, at any rate, show that the painting is a forgery, an idea disproved by recent technical analysis.17 A final work, now lost, which has traditionally been regarded as an unfinished work by Van Eyck is the now lost Virgin of Niklaas van Maelbeke, destined for St Martin’s church in Ypres, the composition of which is preserved in two fifteenth-century silverpoint drawings (Figs.3 and 4). The hypothesis that the painting had been left unfinished by Van Eyck at his death arose in the early nineteenth century and has since become entrenched in the literature, but the in Den Haag’, Geschiedkundige Vereniging Die Haghe, Jaarboek (2014), pp.11–28. The present authors are most grateful to A.-M.J. van Egmond for providing access to her most recent analysis of the documents. The documents are also discussed in Weale, op. cit. (note 2), pp.xxvii–xxviii, no.1; and Paviot, op. cit. (note 2), pp.83–84. 13 The documents are transcribed in full in Van Egmond 2019, op. cit. (note 12), pp.299–300, appendix 10. 14 For sources on journeymen and their regulation, see N. Peeters with the collaboration of M. Martens: ‘Assistants in artists’ workshops in the Southern Netherlands (fifteenth and sixteenth centuries)’, in Peeters, op. cit. (note 3), pp.33–48, at p.35 and note 10. For restrictions on journeymen numbers in other crafts, see G. Des Marez: L’Organisation du travail à Bruxelles au XVe siècle, Brussels 1904, pp.64, 70–71 and 212– 13. On the regulation of Bruges apprenticeships, see Campbell, op. cit. (note 9), pp.47; and D. van de Casteele: Keuren 1441–1774, Livre d’admission 1453–1574, et autres documents inédits concernant la Ghilde de StLuc, de Bruges suivis des Keuren de la corporation des peintres, sculpteurs et verriers de Gand 3. Virgin and Child with a kneeling cleric. South Netherlandish, 15th century. Silverpoint on prepared paper with areas of light yellow wash, 13.4 by 10.2 cm. (Germanisches Nationalmuseum, Nuremberg). 4. Virgin and Child with a kneeling cleric. South Netherlandish, 15th century. Silverpoint on prepared paper with areas of light yellow wash, 27.8 by 18 cm. (Graphische Sammlung Albertina, Vienna). evidence is questionable. When the sixteenth-century writer Marcus van Vaernewijck described the painting as incomplete (‘onvuldaen’), he referred specifically to wings then attached to the centre panel – and these were not necessarily by Van Eyck. The only other evidence for the painting’s state rests in the silverpoint drawings, which omit exactly 1541–1575, Bruges 1867, p.19. 15 N. Salvadó et al.: ‘Mare de Déu dels Consellers, de Lluís Dalmau. Una nova tècnica per a una obra singular’, Butlletí del Museu Nacional d’Art de Catalunya 9 (2008), pp.43–61; S.F. Jones: ‘Jan van Eyck and Spain’, Boletín del Museo del Prado 32 (2014), pp.30–49, pp.32–33; and B. Fransen: ‘Van Eyck in Valencia’, in C. Currie, L. Preedy et al., eds: Van Eyck Studies, Papers presented at the Eighteenth Symposium for the Study of Underdrawing and Technology in Painting, Brussels, 19–21 September 2012, Paris, Leuven and Walpole MA 2016, pp.469–78, at pp.476–77. 16 H.J.J. Scholtens: ‘Jan van Eyck’s “H. Maagd met den kartuizer” en de Exeter-Madonna te Berlijn’, Oud Holland 55 (1938), pp.49–62. 17 For the scientific and technical evidence for the painting’s authenticity, see B. Heller and L.P. Stodulski: ‘“Saint Jerome” in the laboratory: scientific evidence and the enigmas of an Eyckian panel’, Bulletin of the Detroit Institute of Arts 72 (1998), pp.38–55; the study did not attempt ‘to verify the date’s age or originality by invasive analytical means’, but its authors observed that it predated the development of the ‘crackle pattern’, for which, see ibid., p.49. the burlington magazine | 164 | february 2022 123 Jan van Eyck’s posthumous workshop B A the same parts of the design and must therefore go back to a single model. Most scholars have concluded that the drawings are copies made from the surface of the Van Maelbeke Virgin, which had been kept in the workshop in an unfinished state. On these grounds, it has been proposed that the painting was completed by Van Eyck’s journeymen after Jan’s death and that this occurred as late as 1445, the year of Van Maelbeke’s death;18 it has also been argued, however, that the drawings are not copies of an unfinished painting at all but rather deliberately omit select features from a finished model.19 Further, an eighteenthcentury transcription of Van Eyck’s name and the date on the painting, although problematic in its own right, permits the argument that the painting was designed, painted and delivered to Ypres all within Van Eyck’s lifetime.20 18 See, for example, Borchert, op. cit. (note 2), p.72 and figs at pp.72 and 75– 76; idem: ‘Petrus Christus after Jan van Eyck and workshop. “The Virgin and Child with a donor” (copy of the “Maelbeke Madonna”) c.1445–50’, in idem, ed. with contributions by J. Chapuis et al.: exh. cat. Van Eyck to Dürer: Early Netherlandish Paintings and Central Europe 1430–1530, Bruges (Groeningemuseum) 2010, pp.153–55, at p.153. 19 S.F. Jones: ‘The use of patterns by Jan van Eyck’s assistants and followers’, in S. Foister, S.F. Jones and D. Cool, eds: Investigating Jan van Eyck, Turnhout 2000, pp.197–207, p.197–98; and S.F. Jones: ‘New evidence for the date, function and historical significance of Jan van Eyck’s “Van Maelbeke Virgin”’, THE BURLINGTON MAGAZINE 148 (2006), pp.73–81, at p.73. 124 20 Jones 2006, op. cit. (note 19), pp.73–81. 21 J. Bruyn: Van Eyck Problemen. De Levensbron: Het werk van een leerling van Jan van Eyck, Utrecht 1957. 22 M.J. Friedländer: Early Netherlandish Painting. I. The Van Eycks and Petrus Christus [1924], transl. H. Norden, comments and notes by N. Veronée-Verhaegen, repr. New York and Washington 1967, p.81; E. Panofsky: Early Netherlandish Painting, its Origins and Character, New York, Hagerstown, San Francisco and London 1971, pp.187–90; and Bruyn, op. cit. (note 19), pp.100 and 115–21. Bruyn argued instead that two other paintings, the Lamentation in the Musée du Louvre, Paris, and the lost Crucifixion formerly in the Anhaltinische Gemäldegalerie, Dessau, the burlington magazine | 164 | february 2022 5. Detail of Panoramic map of Bruges, by Marcus Gerards, showing the city centre. ‘A’ marks the location of the Sint Gillisnieuwstraat (now called the Gouden-Handstraat) and ‘B’ the location of the Oostmeers. North is towards the lower left. 1562. Modern coloured offset of a copper engraving assembled from ten sheets, 177 by 100 cm. (www. kaartenhuisbrugge.be/Kaart). As with other master painters, the historiography of Jan van Eyck’s workshop in the nineteenth and early twentieth centuries attributed most of the extant paintings to the hand of the master. Systematic study of anonymous ‘pupils’ began only in the 1950s, most notably in Josua Bruyn’s Van Eyck Problemen, published in 1957.21 The book re-examined the influential hypothesis put forward by both Max J. Friedländer and Erwin Panofsky that Petrus Christus had been one of Van Eyck’s ‘pupils’. Bruyn broadly supported these views but, interestingly, showed a direct relationship between Christus and Van Eyck. 23 Friedländer, op. cit. (note 22), p.81; M.P.J. Martens: ‘Petrus Christus: a cultural biography’, in M. Ainsworth, ed. with contributions by M.P.J. Martens: Petrus Christus: Renaissance Master of Bruges, New York 1994, pp.15–23, at p.15; and J.M. Upton: Petrus Christus: His Place in Fifteenth-Century Flemish Painting, University Park and London 1990, p.7. 24 Panofsky, op. cit. (note 22), p.188. 25 For Van Eyck’s children, see Weale, op. cit. (note 2), pp.xl, no.22, xlvii, no.31 and xlix, no.36; and Paviot, op. cit. (note 2), p.90. 26 J.K. Steppe: ‘Lambert van Eyck en het portret van Jacoba van Beieren’, Mededelingen van de Koninklijke Academie voor Wetenschappen, Letteren en Schone Kunsten van België, Klasse der Schone Kunsten 44 (1983), pp.53–86. The document is an early seventeenth-century inventory of the castle of Arenberg at Heverlee, near Leuven, drawn up for Karel van Croÿ (Charles de Croÿ), the fourth Duke of Aarschot (d. 1612), entitled Recueil et Registre du Chateau D’Heverlé de toutes les places et Chambres étans en icelui, Leuven, KU Leuven, University Archives, B.H. 76.2. 27 Steppe, op. cit. (note 26), pp.60, 62 and 85. 28 T.-H. Borchert: ‘Introduction, Jan van Eyck’s workshop’, in idem, ed.: exh. cat. The Age of Van Eyck: The Mediterranean World and Early Netherlandish Painting 1430–1530, Bruges (Groeningemuseum) 2002, p.25; idem: ‘Being there: Jan van Eyck and Petrus Jan van Eyck’s posthumous workshop he disagreed with Panofsky’s attributions to Christus, dismissing Panofsky’s view that Christus collaborated on the Detroit Jerome and painted the Frick Virgin.22 Christus is first documented in Bruges on 6th July 1444, some three years after van Eyck’s death, when he paid to become a burgher (poorter) of the city. It is recorded in the register in which new burghers were inscribed (‘Poorterboek’) that he took this step ‘om[m]e scilde[re] te zine’ – in order to set himself up as the master of an independent workshop – which necessitated becoming a burgess of the city.23 Panofsky placed Christus in the posthumous workshop, arguing that he oversaw production in the workshop’s last years ‘in the name of the widow until such time as he established himself in his own right’.24 The posthumous management of the business has always been at issue because Van Eyck’s children were too young to inherit it: one of them (of unknown gender) was born in 1434 and was thus no more than seven years old when Van Eyck died.25 In 1983 the terms of the debate changed when Jan Steppe published a description in an early seventeenth-century inventory of a lost portrait of Jacqueline of Bavaria by Lambert van Eyck, which suggested that Lambert too was a painter.26 Since Lambert was certainly in Bruges between March and June of 1442, Steppe proposed that he was one of Van Eyck’s executors and that he took over his brother’s workshop, running it until around 1450.27 A date around 1450 for the workshop’s closure has sometimes been accepted.28 Susan Jones, however, raised the possibility that the workshop closed earlier, since Van Eyck’s house in Bruges changed hands between 24th June 1443 and 24th June 1444.29 Her proposal was that the workshop was wound down in the early 1440s, possibly by Lambert or alternatively by Jan’s widow, Margaret, as the guild regulations allowed for the widow of a master painter to inherit his rights and privileges.30 As scholarship moved away from the figure of Petrus Christus and towards Van Eyck’s surviving family members in the course of the 1980s and 1990s, Christus’s own career was also reassessed. In the mid-1990s, it was argued – and widely accepted – that Christus could not have been one of Van Eyck’s journeymen because he acquired the status of burgess by payment on 6th July 1444, whereas if he had already lived in Bruges for more than one year and one day he would have become a burgess automatically.31 Art historians during this period therefore argued that Christus’s career in Bruges probably began in July 1444, or only shortly beforehand. Only if Van Eyck’s posthumous workshop continued to operate after that date might Christus have been active there – and indeed, some scholars argued that Christus must have worked in a posthumous workshop still functioning in 1444–45.32 The theory that the workshop closed somewhat earlier, in the course of 1443–44, however, required an alternative explanation, such as that Christus in Bruges’, in E. Capron, ed. with M. Ainsworth and T.-H. Borchert: exh. cat. The Charter-house of Bruges: Jan van Eyck, Petrus Christus and Jan Vos, New York (Frick Collection) 2018, p.102; M.W. Ainsworth: ‘Attribution mysteries of the “Virgin and Child with St Barbara, St Elizabeth, and Jan Vos”’, in ibid. 2018, p.87. Till-Holger Borchert speculated that production took place in two successive locations: Margaret (probably in collaboration with Lambert) would have sold the house on the Sint Gillisnieuwstraat in 1444 and potentially transferred the business to new premises, remaining in business until 1450, see Borchert, in ibid. 2018, pp.101–02. 29 S.F. Jones: ‘The workshop and followers of Jan van Eyck’, unpublished PhD thesis (Courtauld Institute of Art, London, 1998), pp.17–20 and 184–86 (distinguishing ‘a period of about two to three years during which Jan van Eyck’s workshop may have remained operative after his death’); and idem: ‘Jan van Eyck and Spain’, Boletín del Museo del Prado 32 (2014), pp.30–49, at p.45. 30 On this possible role for Jan’s widow, see, for example, Jones 2000, op. cit. (note 19), p.197; T.-H. Borchert, ‘Introduction, Jan van Eyck’s Workshop’, in idem, ed.: exh. cat. The Age of Van Eyck: The Mediterranean World and Early Netherlandish Painting 1430– 1530, Bruges (Groeningemuseum) 2002, pp.9–32, pp.15 and 25. 31 Upton, op. cit. (note 23), pp.8–19, and in particular p.9, note 9; and Martens, op. cit. (note 23), p.15. 6. Detail of the map in Fig.5, showing Van Eyck’s house on the Sint Gillisnieuwstraat, now called the Gouden-Handstraat. Christus was in contact with former assistants of Van Eyck who had subsequently opened their own workshops.33 Resolving when the workshop closed necessitates revisiting the question of its location, and in particular whether it was in Jan van Eyck’s house on the Gouden-Handstraat in Bruges (then called the Sint Gillisnieuwstraat; see Fig.5, A). Between 1432 and 1441, he paid an annuity of 30 shillings parisis on a property on that street, due on 24th June, and his widow Margaret continued to make the payment subsequent to his death, in 1442 and 1443.34 Yet the evidence for the workshop’s location and posthumous continuity has always been complicated by a statement made by W.H. James Weale (1832–1917) that in 1444 Margaret sold the house 32 For the idea that Christus ‘participated in’ a posthumous workshop, based on the attribution to him of the Vienna drawing after the Van Maelbeke Virgin, see M.W. Ainsworth: ‘Madonna and Child with a donor’, in M.W. Ainsworth, ed. with contributions by M.P.J. Martens: exh. cat. Petrus Christus. Renaissance Master of Bruges, New York (Metropolitan Museum of Art) 1994, pp.182–84. Till-Holger Borchert has also argued that Christus made the Vienna drawing c.1445 as a member of Van Eyck’s posthumous workshop, see Borchert 2008, op. cit. (note 2), p.72 and figures at pp.75–76; and idem 2010, op. cit. (note 18), pp.153–55. More recently, Maryan W. Ainsworth has spoken of a workshop that was ‘perhaps’ still in operation when Christus arrived in Bruges, see Ainsworth, op. cit. (note 28), pp.71–89, at p.87. 33 Jones adopted the viewpoint that since Van Eyck’s workshop likely closed before 24th June 1444, Christus probably obtained knowledge of Van Eyck’s working practice through contact with former journeymen still active in Bruges, see Jones 1998, op. cit. (note 29), pp.185–86; and Jones 2000, op. cit. (note 19), p.204. 34 See T. De Meester et al.: ‘“Meester Jans huus van Eicke”. The house, workshop and environment of Jan van Eyck in Bruges: new evidence from the archives’, in T.-H. Borchert, M. Martens and J. Dumolyn, eds: exh. cat. Van Eyck: An Optical Revolution, Ghent (Museum voor Schone Kunsten) 2020, pp.127–37, at p.130. the burlington magazine | 164 | february 2022 125 Jan van Eyck’s posthumous workshop in the Gouden-Handstraat and moved to a dwelling on the Oostmeers named The Wild Sea (see Fig.5, B).35 Weale claimed that she lived there until at least 1456. Weale was one of the pioneers of archival research in the study of the so-called ‘Flemish Primitives’ and his observations must be taken seriously. Uncharacteristically, however, he did not provide any reference for this information, making the evidence regarding The Wild Sea distinctly problematic: it could not be excluded that Margaret set up a posthumous workshop in a new location after selling the house on the Gouden-Handstraat, or even that the house on the Oostmeers was already in Van Eyck’s possession in the 1430s, giving him two possible workspaces. New archival research on The Wild Sea, however, has resolved this matter, and indicates that Weale, who was otherwise a meticulous Van Eyck biographer – and someone who did not shy away from castigating other scholars for their mistakes – drew some hasty and inaccurate conclusions in this instance. Although he neglected to provide his source, it is clear that he relied on the accounts of the Obediëntie of the Chapter of the Collegiate Church of St Donatian in Bruges.36 These are the same accounts that record Margaret paying the annuity on the house in the Gouden-Handstraat. A first mistake made by Weale is his claim that Margaret ‘sold’ the house: the accounts do not explicitly mention a sale, and although the house must have changed ownership between the summer of 1443 and the spring of 1444, this could also have happened by way of gift or inheritance. More seriously, Weale also appears to have jumped to conclusions regarding his principal claim that The Wild Sea house was owned by Margaret. From 1438 onwards, a certain ‘Margriete sHeex beghina’ appears in the accounts of the Obediëntie, to which she owed an annuity of 6 shillings, 9 pence and 1 poitevin on a property in one of the city’s six administrative districts (‘zestendelen’), that of Our Lady (Onze-LieveVrouw).37 This Margaret sHeex (also sHeics, sHeycx or sHeyx, the female genitive of ‘Van Eyck’) remained the owner of this estate until at least 1459–60, after which the accounts are missing.38 In later life, this beguine must have enjoyed high status, since from the mid-1440s she was referred to in records as ‘domicella’. What seems to have happened is that Weale identified this Margaret as Van Eyck’s wife and, later, widow – despite the fact that the woman in question was described in documents as a beguine even during Van Eyck’s lifetime. As beguines by definition lived a single life and took a vow of chastity, this would have entailed the couple living separately from at least 1438 onwards, but Van Eyck’s portrait of 35 Weale, op. cit. (note 2), p.26; and W.H.J. Weale and M.W. Brockwell: The Van Eycks and their Art, London 1912, p.22. 36 Without giving an explicit source, Weale stated in an explanatory footnote that ‘the accounts for the following years [post 1456] are wanting’, Weale, op. cit. (note 2), p.26. The accounts were likely those of the Obediëntie of St Donatian’s – Bisschoppelijk Archief, Bruges (herafter cited as BAB), Obediëntie van Sint-Donaas, G130-131 – which Weale had used for his earlier research on Van Eyck’s house, for which, see W.H.J. Weale: Notes sur Jean van Eyck: Réfutation des erreurs de M. l’Abbé Carton et des théories de M. le Comte de Laborde suivie de nouveaux documents découverts dans les archives de Bruges, London, Brussels and Leipzig 1861, pp.6–14. These are indeed wanting, although not after 1456 as Weale claimed, but rather after 1459–60. 37 BAB, Obediëntie van Sint-Donaas, G128: rekening 1438, fol.14r. Before that date, the annual charge was paid by a certain Johannes Monachi (‘Jan de 126 Muenc’): ibid., account 1435, fol.14r. A painter in Bruges called Jan de Meunic is recorded between 1448 and 1470, however, there was also an important ‘de Muenc’ family of carpenters in the beginning of the fifteenth century (Jan and Jacop de Muenc, among others) with properties in the nearby Goezenputstraat and Zonnekemeers. For the painter, see A. Schouteet: De Vlaamse primitieven te Brugge: bronnen voor de schilderkunst te Brugge tot de dood van Gerard David, Brussels 2004, II, pp.77–78. 38 BAB, Obediëntie van SintDonaas, G129 and G130, passim. 39 BAB, Obediëntie van SintDonaas, G129: account 1441–43, fols.14r and 15v. 40 BAB, Obediëntie van SintDonaas, A191, fol.224v. 41 BAB, Obediëntie van Sint-Donaas, A194, fol.58v. The annuity is here described as mortgaged on ‘two stone dwellings parallel to the street with some open space at the street side and a stone wall at the south side. Next, on a shelter and a small house with a small wooden façade, north of the burlington magazine | 164 | february 2022 7. Margaret van Eyck, by Jan van Eyck. 1439. Oil on oak panel, including the frame 41.3 by 34.5 cm. (Groeningemuseum, Bruges; © KIK-IRPA, Brussels). Margaret, finished in 1439, makes this impossible, as an inscription on the frame refers to ‘my husband Jan’ (‘co[n]iu[n]x m[eu]s Joh[ann]es’) (Fig.7). Additionally, the fact that Van Eyck’s widow (‘relicta Johannis de Eyke’) and the beguine Margaret sHeex were referred to with different appellations in the same source contradicts Weale’s implicit supposition that they were one and the same person.39 The accounts of the Obediëntie provide no additional information on the house of Margaret sHeex aside from the fact that it was situated in the district of Our Lady. Luckily, however, the annuity mortgaged on the house can be traced in subsequent rent registers that yield further details. Thus, it appears from a register drawn up in 1451 that the annuity Margaret paid every year on St Bavo’s Day (1st October), was mortgaged on two singleroom houses (‘ex fundo duarum camerarum’) situated on a plot on the west side of the Oostmeers.40 The clerk compiling the register probably made a mistake in locating the houses, as a later register, dated 1543, situated the same annuity on the east side of the Oostmeers.41 At that time, the plot on which the houses were located was owned by a tanner named Pieter van den Berghe. However, none of these fifteenth- and sixteenth-century rent registers gives the house a name – or any house for that matter, as the registers of St Donatian’s merely record individual rents to be paid on each street without providing details on the individual plots of land or houses. Why then did Weale think the house was called The Wild Sea? As far as we know, the house name is first recorded in a deed of sale dated 1578. The deed concerns a house ‘formerly named De Wilde Zee’ and an adjacent dwelling whose roof ridge runs parallel to the street (‘dweersloove’) on the east side of the Oostmeers, which had until recently been owned by a certain Jozef de Roo.42 The proto-cadastral registers of the zestendelen, which start in 1580, indicate that The Wild Sea was located in the row of houses between the Zonnekemeers and the Wijngaard (Fig.8). As is the case in the deed of sale of 1578, the registers record two houses: a northern house ‘formerly joined to the adjacent property’ and a southern house, designated as The Wild Sea.43 In 1580 both houses belonged to Joos de Muelenaere; two years later his heirs sold the northern house,44 and in 1585 an outstanding debt appears to have been mortgaged on The Wild Sea in favour of a Pieter van den Berghe – the first stone house, all next to one another, located at the east side of the Oostmeers, between a house with a thatched roof and a stone façade belonging to the carpenter Pieter Eenhoghe to the south, and the premises of Pieter Aert, stonemason, with a stone wall at the street side, to the north’ (‘twee steenen woensten loofwijs staende met wat aerve ende plaetse van lande ter strate met eenen muere ande zuudtside daer an. Item noch up een love ende een cleen huusekin booven de eerste steede die van steenen es met een cleen houten ghevelken ande noordtside daer an al neffens malcan-dren, staende ande oostside vander strate ghenaemt de Oostmeersch tusschen een huus met stroo ghedect, hebbende eenen steenen ghevele, toebehoorende Pieter Eenhoghe, temmerman, ande zuudtside ende de aerve ende plaetse van lande hebbende een muur ter straten waert toebehoorende Pieter Aert, steenhauwere, ande noordtside’). 42 SAB, OA, 182: Verkopingen bij decreet 1561–81, fols.143v–144r. The parcel sold was described as ‘a house with its belongings formerly known as The Wild Sea, with a house to the south of it, lying next to each other in the Oostmeers on the east side of the street’ (‘een huus met zijnen toebehoorten wijlen gheheeten De Wilde Zee, met een dweersloove ter zuudtzijde daerneffens staende, ten voorhoofde neffens elcanderen in de Oostmeersch an de oostzijde van der strate’). 43 SAB, OA, 138: Zestendelen, OnzeLieve-Vrouw, fols.596 bis and 597. The deed of sale designated the northern house as the Wild Sea while the 1580 register assigns this name to the southern house, a contradiction that must stem from the fact that both houses previously formed a single unit called the Wild Sea. 44 The buyer was an Inghelbrecht Vlamync, see SAB, OA, 182, Verkopingen bij decreet 1581–87, fol.151r. Earlier, Pieter van den Berghe was the owner of the houses to the north of the Wild Sea, which he sold to the same Inghelbrecht Vlamync in 1580, see SAB, OA, 138: Zestendelen, OnzeLieve-Vrouw, fols.592bis and 596. Jan van Eyck’s posthumous workshop the burlington magazine | 164 | february 2022 127 Jan van Eyck’s posthumous workshop apparently the man who owned the house of Margaret sHeex in 1543.45 These must be the same pieces of evidence that Weale uncovered in the nineteenth century and linked together to arrive at the conclusion that Margaret sHeex’s house was called The Wild Sea. It was a premature conclusion. Among other reasons, the outstanding debt owed to Pieter and mortgaged on The Wild Sea does not necessarily prove that Pieter was once the owner of the house: it is also possible that the owner of The Wild Sea mortgaged a debt to Pieter on his property without the latter having had anything to do with the house. It is even debatable whether this Pieter van den Berghe documented around 1580 was the same man as the tanner recorded in 1543.46 All in all, the hypothesis that the house of Margaret sHeex was named The Wild Sea rests on decidedly slender foundations – and the fact that Weale, unusually, did not cite his sources may mean that he was conscious of this at some level. The idea should probably be rejected. It is impossible to identify the beguine Margaret sHeex as the wife of the master painter Jan van Eyck, but she could still have been a relative of the Van Eyck family of painters, as the sixteenth-century writers and humanists Marcus van Vaernewijck and Lucas de Heere believed that Hubert and Jan had a sister called Margaret. This sister was proficient in ‘the noble art of pictoria or painting’ and, according to Van Vaernewijck, remained a virgin for her entire life.47 The question whether this beguine Margaret sHeex was Van Eyck’s sister was raised as early as 1847 in a publication by the priest and scholar Charles Carton, who had seen the name in the rent registers.48 But this too is very unlikely, as Margaret sHeex was a well-known figure in the beguinage of the Wijngaard (St Elizabeth) in Bruges around the mid-fifteenth century. The daughter of ‘Maergrieten’ and ‘Jooris sHeics’ (who died before 1420), she rose up the career ladder at the beguinage from 1425 to 1467, and was grand mistress from 1429/30 onwards.49 These findings have several important implications. The overturning of Weale’s speculations about The Wild Sea makes it almost definite that Van Eyck’s workshop was in his house on the Gouden-Handstraat – which is now the only house in Bruges recorded in the possession of the Van Eyck family. This is not excluded by the physical characteristics of the house: it was two storeys high with a peaked gable, and was set on a fairly large plot of land on which stood a second two-storey building, likewise with a peaked gable, which backed directly onto the canal now called the Gouden-Handrei (Fig.6).50 The second house at the rear of the property could even have been Van Eyck’s main working space, or at least an additional one. That Van Eyck actually resided there is supported by the fact that Margaret continued to pay the annuity on the house after his death, the last payment dating from June 1443.51 Rather than a smooth, decade-long continuation of Jan’s business, therefore, we propose the hypothesis that there was a particular moment, in spring 1444, when Jan’s workshop closed for good. The stock of 45 SAB, OA, 138: Zestendelen, OnzeLieve-Vrouw, fols.596bis and 597. 46 The latter, born around 1515, is recorded in the archives of the guild of the Bruges tanners up to 1559, but not subsequently, not even in a list of all of the guild members compiled in 1568, see Rijksarchief, Ambachten Brugge, Bruges, 71, fols.103r and 116r. 47 For transcriptions of the texts by De Heere and Van Vaernewijck, see Weale, op. cit. (note 2), pp.lxxix, lxxxvi and xc. There are no contemporary sources for the sister of the Van Eycks. In an article of 1847, Charles Carton cited the notes of the architect and collector Pierre-Jacques Goetghebuer 128 (1789–1866), who found in the written subscription to the guild of Onze-LieveVrouw-op-de-Rade in the church of St John in Ghent a master Hubrech Van Hyke in 1412 and his sister Mergriete van Hyke in 1418, for which, see C.L. Carton: ‘Les trois frères van Eyck’, Annales de la Société d’Émulation pour l’Étude de l’Histoire et les Antiquités de la Flandre 9 (1847), pp.237–326 and 324–25. The guild register is now missing, see D. Lievois: ‘Het archiefmateriaal over Hubert van Eyck in Gent’, Handelingen der Maatschappij voor Geschiedenis en Oudheidkunde te Gent 67 (2013), pp.59–67, at pp.64–65. 48 The unreliable Carton seems to the burlington magazine | 164 | february 2022 C A B 8. Detail of the map in Fig.5, showing the Oostmeers. ‘A’ marks the row of houses between the Zonnekemeers; ‘B’ marks the Cowgate; ‘C’ marks the Wijngaard. unfinished panels it had contained must have been liquidated by then. This is not contradicted by the dates of the three works most often regarded as ones left incomplete by Van Eyck: the Van Maelbeke Virgin, the Frick Virgin and the Detroit Jerome. The first two works were certainly finished before 1444, and this may well be the case for the Jerome, which is inscribed with the date 1442. In the case of the Van Maelbeke Virgin, numerous scholars have supported the idea that the painting was still preserved in the workshop in an unfinished state as late as 1445; however, that year was only ever associated with the painting in an unreliable late eighteenth-century town chronicle of Ypres by Petrus Martinus Ramaut (1719–83), published in 1825 by the archivist Liévin de Bast.52 Whatever the explanation for the date 1445 in the chronicle, it is not solid evidence for the date of the work’s completion.53 At least some of these paintings could have belonged to a final, posthumous phase of production in the early 1440s, when unfinished paintings were completed and potentially new ones made. This relatively short phase is probably best understood not as one of continuity but as a gradual winding down. Evidence arising from other works made in the early to mid-1440s does not contradict this argument. A lost painting of St George and the dragon attributed to ‘Johannes’ and acquired in Bruges by the merchant have misstated his own source, as he wrote that the name was listed in the accounts of the chapter of St Donatian between 1438 and 1465, see Carton, op. cit. (note 47), p.265, whereas it appears in the accounts of the Obediëntie of St Donatian. His assertion that the name could be found in the accounts until 1465 cannot be verified since the accounts of the Obediëntie are missing between 1460 and 1479. 49 For this document and the career of Margaret sHeex, see D. Desmet: ‘Het begijnhof ‘De Wijngaerd’ te Brugge. Onderzoek naar het dagelijks leven rond het midden van de vijftiende eeuw’, unpublished MA thesis (Katholieke Universiteit Leuven, 1979), pp.217–19. 50 For further discussion of the material aspects of the building, see De Meester et al., op. cit. (note 34), p.130. 51 Ibid. 52 L. De Bast: ‘Notes on the “Série inférieur de la grande composition peinte, pour l’église de S. Jean à Gand, par Jean van Eyck” by G.F. Waagen’, Messager des Sciences et des Arts (1825), p.168, notes 1 and 2, and p.169, note 1. For a transcription of two known versions of the chronicle, only one of which is now extant, see Jones 2006, op. cit. (note 19), pp.73–81, at p.81. 53 For the idea that the chronicle Jan van Eyck’s posthumous workshop Joan Gregori for Alfonso V of Aragon for the substantial sum of 2,000 sous reyals of Valencia, was purchased before 2nd May 1444.54 This allows for the possibility that the painting was acquired directly from the posthumous workshop, close to the date of its closure, although it is equally possible that Gregori simply persuaded its owner, then resident in Bruges, to part with it. A second painting of St George, also lost, was likewise acquired in Bruges some years later, this time for Leonello d’Este, who, in 1446, paid the lawyer and cleric Antonio Domaschi in Bruges the sum of 100 gold ducats for the painting.55 Whether or not the Este George was modelled on Alfonso’s St George and the dragon, its acquisition cannot be used to argue that Van Eyck’s workshop remained open in 1446, as the document does not name the painter. What of the hypothesis that one or both of Van Eyck’s surviving family members – Margaret and Lambert – ran a posthumous workshop in Bruges? The new evidence regarding The Wild Sea changes the biographical evidence for Margaret. Weale’s misidentification led him to believe that Margaret was ‘still living’ in 1456, but the last record of her now dates from 1446, when her name appears in the records of the Bruges lottery. Margaret had not remarried, as she is referred to as Jan van Eyck’s widow.56 She is listed as receiving 2 pound groats, which in all likelihood represents the income from an annuity issued by the city.57 The fact that her name is not listed in a register of all the annuities held by the city, which starts in the year 1450/51 therefore indicates that Margaret had probably died before that year. The new date for her death clearly undermines the theory that Margaret ran a continuing Van Eyck workshop in the city until around 1450. In addition, there is no evidence to suggest that she was a painter, something which seems to have been possible for women at this period only if they were trained within the family.58 Furthermore, the memorial list of the Bruges guild of imagemakers and saddlers, which begins in 1450, does not record a single widow.59 Although the regulations did permit a widow to take over her deceased husband’s business, we can infer that this option was rarely if ever taken up in practice, unless as a short-term measure – and perhaps in such cases, the widows were not listed. Indeed, the very fact that some three years after Jan’s death Margaret moved away from the house and workshop on the Gouden-Handstraat favours a different hypothesis: that Margaret had no need for this large house with its working spaces and substantial outbuilding, and moved into a smaller house. Finally, it cannot be ruled out that Margaret left Bruges entirely. The final surviving document concerning Van Eyck’s family – the financial gift from Philip the Good to Jan and Margaret’s daughter Lievine, dated February 1449 O.S. – does not refer to Margaret by name, but it does indicate that at the period of her mother’s death Lievine entered the monastery of St Agnes in Maaseik, in the Mosan region.60 Recent research has confirmed that Margaret’s natal family, like her husband’s, originated from the diocese of Liège, making it possible that she returned there after his death.61 Taken transmits a text from a written document, see Jones 2006, op. cit. (note 19), pp.73–81, at p.77. 54 For the documents on Dalmau, see Archivo del Reino de Valencia, Mestre Racional, 59: Batlia general de Valencia. Comptes d’administraciò, Valencia, 1444, fols.273v–274r, 2nd May 1444, and fol.283v, 22nd August 1444, see also J. Sanchis y Sivera: Pintores Medievales en Valencia, Valencia 1930, pp.114–15; and R. Weiss: ‘Jan van Eyck and the Italians’, Italian Studies 11 (1956), p.15. 55 We are grateful to Geoff Nuttall for sharing his interpretation of this document with us and Paula Nuttall for her insights into the reception of the St George paintings. For Leonello’s mandate, see Archivio di Stato, Modena, Camera Ducale, Computisteria, Mandati, vol.7 (1445–1446), c.317v. 56 SAB, OA, 273, fol.3 identifies her as ‘de wed(uw)e Jans va(n) Eyck’. On the lottery, see L. Gilliodts-Van Severen: ‘La loterie à Bruges’, La Flandre 1 (1867– 68), pp.5–26, 80–92 and 160–95; II, 1868–69, pp.408–73; III, 1869–70, pp.5– 110, for the widow of Van Eyck, see p.9. 57 On Margaret and the Bruges lottery, see further Dumolyn, Geirnaert and Speecke, op. cit. (note 7). The interpretation of this document is uncertain: either the city paid Margaret together, all these points make it possible that Margaret played a role in the winding down of Van Eyck’s Bruges workshop in the early 1440s, but unlikely that she continued to manage a posthumous Van Eyck workshop in Bruges until c.1450. This leaves us with the figure of Lambert van Eyck. Whether Lambert worked alongside his brother Jan in the 1430s and early 1440s requires further study. The evidence that he was a painter combined with the commercial success of Jan’s business certainly make it a reasonable hypothesis in principle that he took over his brother’s workshop; however, no evidence has yet emerged to show that he re-established the workshop in Bruges after the sale of Jan van Eyck’s house and workshop in 1443–44. After 1442 we lose all trace of Lambert, and it remains possible that he subsequently left Bruges. It is remarkable that Steppe failed to identify any paintings that could have been made in a Lambert van Eyck workshop at this date, speculating only that he could have produced unspecified copies of Jan van Eyck’s lost Holy face. Clearly, the hypothesis of a Lambert van Eyck workshop is not needed to account for the surviving panel paintings in Van Eyck’s style that may have been made in Bruges or its vicinity in the second half of the 1440s. Such works could equally be divided (hypothetically) among workshops established by former journeymen or apprentices, on one hand, and those of painters in Bruges who had not been formally trained by Van Eyck but who had assimilated his style and techniques, on the other. The arguments presented here potentially reduce the roles played by Margaret and Lambert in the posthumous workshop, but they also refocus our attention on Jan’s journeymen. Is it possible after all that a trusted journeyman took over the workshop’s management after Jan’s death? More particularly, the hypothesis that the workshop closed in spring 1444 reopens the question whether Petrus Christus was active in the posthumous workshop before 1444, when he paid to become a burgess in Bruges, in order to establish himself as a master in his own right. As mentioned earlier, scholars in recent decades rejected this possibility on the grounds that Christus would already have become a burgess by default if he had lived in Bruges for more than one year and one day; it now appears unlikely, however, that this particular rule was operating in Bruges at that period – and that makes it possible that up to the moment that he became a burgess in July of 1444, Christus had been living and working in the city.62 Other solutions cannot be excluded: for example, Christus could have come to Bruges in July 1444 to take advantage of a temporary reduction in the rates for becoming a burgess ordered by Philip the Good on 24th January 1440 (1441 N.S.) and set to expire four years later.63 What is needed now, therefore, is a complete scholarly reappraisal of the nature of Christus’s relationship to Van Eyck, involving both thorough reassessment of the laws and regulations that shaped the activity of painting in Bruges, and fresh study of the documentary, physical and visual evidence. the income from an annuity out of the lottery proceeds, or alternatively the city ‘paid’ her the annuity in the form of a stake in the lottery. It is uncertain why Weale settled on the year 1456 when he stated that Margaret was ‘still living’ in that year, as the accounts of the Obediëntie of the Chapter of St Donatian’s end slightly later, in 1459– 60: this may have been an error on his part. For the records of the town lottery, see SAB, OA, 273, fols.3 and 14v. 58 See Reynolds, op. cit. (note 8), p.22; and M. Droz-Emmert: Catharina van Hemessen: Malerin der Renaissance, Basel 2004, pp.37–38. 59 For the memorial list, which is a copy made c.1490, see SAB, OA, 314: Beeldenmakers and https:// erfgoedbrugge.be/memorielijstvan-de-brugse-schilders/, accessed 5th January 2022. 60 Weale, op. cit. (note 2), pp.xlix–l, no.36. 61 H. Callewier: ‘A new document on Jan van Eyck: his request for a confessional letter in the Vatican Archives’, Simiolus 43 (2021), pp.16–25, p.19. 62 New research on this problem by Jan Dumolyn will be published in a forthcoming publication. 63 For this explanation, see Martens, op. cit. (note 23), p.15. the burlington magazine | 164 | february 2022 129