Pieter Brueghel the Younger based his paintings of The Four Seasons on the series conceived by his father as designs for engravings that had been commissioned by his publisher Hieronymus Cock (1518–1570) (fig. 1). Cock clearly recognised the commercial potential of such a set, drawing on the long, rich visual history of representations of the months or seasons, which originated in medieval books of hours. Bruegel the Elder made drawings for Spring, signed and dated 1565,² and Summer, signed and dated 1568,³ but following his death in 1569, the series was completed by another leading artist of the time, Hans Bol (1534–1593), who was evidently heavily dependent on Bruegel’s example and working very much in the spirit of his contemporary. The inn to the right in this scene, for example, is based on a real tavern (in Hoboken, outside Antwerp), which appears in the Elder’s drawing of 1559,⁴ and in subsequent engravings. Unfortunately neither of Bol’s preparatory drawings for the series appears to have survived, however. Cock’s prints were published in 1570 and very quickly became widely known.

While Brueghel the Younger certainly made complete painted sets of The Four Seasons (possibly only two of which remain intact),⁵ it is clear that the enterprising artist was also willing to execute individual versions or pairs of the seasons according to demand. The present painting, for example, was one of a pair of pendants, the other representing Spring, also with Galerie Robert Finck in Brussels in 1960.⁶

The Winter Landscape exists in only ten to twelve autograph versions, all of which are signed, one securely dated 1621, and two others possibly dated 1613 and 1622.⁷ The form of the signature here – ‘Breughel’ as opposed to ‘Brueghel’ – indicates a date after 1616, when the artist changed the spelling of his name.⁸ These autograph versions of the composition are all executed on panels of similar size and proportion, strongly suggesting that the design was transferred by a tracing available in the Brueghel workshop, though variations in details such as the actions and number of figures on the ice vary between them.

Nor has the artist slavishly followed Cock’s engraving. As well as updating some of the figures’ dress, Brueghel has adapted the composition, with several changes to the staffage, the position of the trees, framing the scene with trunks on either side, and the sparser background, such as the bare, snow-covered bank to the left, rather than the vineyard visible in Cock’s engraving. These amendments reflect the artist’s understanding of the painted format, which is served better by this spacious arrangement than the more crowded appearance of the graphic scene.

1 Both in the Kunsthistorisches Museum, Vienna, inv. nos 1838 and 625, respectively; see S. Haag et al., Bruegel, exh. cat., Vienna 2018, pp. 214 ff., cat. no. 75, reproduced in colour pp. 224–25; and pp. 212–13, cat. no. 71, reproduced in colour pp. 210–11.

2 Vienna, Albertina; see Ertz 1988/2000, vol. II, p. 537, reproduced p. 538, fig. 412.

3 Hamburg, Kunsthalle, Kupferstichkabinett; see Ertz 1988/2000, vol. II, p. 537, reproduced p. 538, fig. 413.

4 London, Courtauld Institute; see H. Mielke, Pieter Bruegel: die Zeichnungen, Belgium 1996, pp. 55–56, cat. no. 44, reproduced p. 167.

5 The set sold London, Christie’s, 7 July 2016, lot 6 (£6,466,500), and that in the National Museum of Art of Romania, Bucharest; other formerly complete series are believed to have been since dispersed. 

6 See Ertz 1988/2000, pp. 590–91, cat. no. E610, reproduced.

7 See Ertz 1988/2000, pp. 601–02, cat. nos E665–E677. The uncertainty over the number of accepted works is due to the likelihood of the duplication of entries. Another autograph Winter Landscape, not published in Ertz, was sold London, Sotheby’s, 8 July 2015, lot 6 (£1,085,000).

8 Dendrochronological analysis carried out by Ian Tyers of Dendrochronological Consultancy Limited, has found that the latest dated heartwood ring of the three boards making up the panel is 1549. With a minimum expected number of eight sapwood rings likely to be missing, it is possible to assume a terminus post quem for the felling of the tree, of circa 1557. None of the boards in this panel are of a typical width for Baltic timber, however, meaning it is neither possible to know by how much they were trimmed, nor to provide an accurate likely usage date.