This portrait of Peter Paul Rubens (1577-1640; for a biography of whom see e.g. SK-A-346) was acquired with a portrait of the jurist and historian Jan Caspar Gevaerts (1593-1666; SK-A-2319) for the prestigious Van Winter collection as the work of Anthony van Dyck. Later both were sold as such to the Dutch government in 1907, and catalogued by the museum as autograph until 1926, and in the case of SK-A-2319 until 1934, when the present painting was omitted from the catalogue. Evers in 1944 attributed the present work to Van Dyck,5 but Burchard rejected the attribution nine years later,6 and recently Vey has followed other authorities by describing it as a copy.7 The work is to be associated with the grisaille oil sketches executed by Van Dyck in preparation for his Iconography, a series of etched and engraved portraits of famous personages, mostly contemporary. The earliest reference to such a set of prints occurs in March 1632.8 The prints were preceded by, at the least, ad vivum drawings (or drawn copies of likenesses) and oil-sketches in grisaille, the latter on a scale similar to the eventual print. The purpose of the oil sketches in the process is not certain – on the rare occasion when Van Dyck etched a portrait himself none seems to have been executed – but they must have been intended to act as a guide for the engraver. The great majority of the extant grisailles – most of which are thought to be autograph – is in the collection of the duke of Buccleuch and Queensberry at Boughton House, Northamptonshire. This group has a provenance that goes back to the artist Sir Peter Lely (1618-1680), but contains some replicas, or near replicas, and is uneven in quality.9 The museum picture closely connects with (but is larger than) the design for the portrait of Rubens at Boughton, which has been claimed to be the working grisaille10 executed in the process of creating the print.11 But differences between these two and some clumsy passages – especially in the ruff – suggest that the Buccleuch sketch may itself be a copy and not a prototype from which the print was made. It is generally agreed that the museum picture is a copy of the Buccleuch sketch; the 1976 museum catalogue suggested it was made in Van Dyck’s studio, and Luijten has more recently claimed that the original grisailles were kept by Van Dyck and that he ‘had his pupils copy them’.12 Such a circumstance would be a repetition of what was recollected as having happened with the Apostle series13 during Van Dyck’s first Antwerp period (c. 1613-1620) and would explain the large number of extant copies of the grisaille sketches. Vey has implicitly rejected this theory by referring to ‘a number of able painters [in Antwerp and Brussels] who could translate a Van Dyck drawing or painting into a grisaille … It goes pretty much without saying that their grisailles would have subsequently found their way into Van Dyck’s workshop stock.’14 The museum picture is not a grisaille, but rather is executed in shades of brown, and although Vey has described all the Boughton sketches as grisailles, Luijten has stated that some sketches – perhaps not those that were connected with the process of making the prints – were ‘composed in brown or greenish hues’.15 This work was executed with no underdrawing in brunaille. The weak handling of the Rijksmuseum painting suggests quite a distance from Van Dyck or his studio, if the latter did indeed make such copies. It is quite possible that this brunaille was painted after, but not long after, Van Dyck’s death, but precisely when is impossible to estimate. The hand seems not the same as that which executed the Gevaerts portrait (SK-A-2319). Klein has estimated that the oak support (from the German/Netherlandish region) would have been ready for use ‘possibly from 1630’, while ‘more plausible’ would be a date from 1640.16 The prototype from which Paul Pontius (1603-1658) made his engraving for the Iconography would have been executed during Van Dyck’s ‘second’ Antwerp period (c. 1627-1632/1634), around 1630. Van Dyck must have known Rubens’s appearance well from early in his career and we know that they had encountered one another17 in the early summer of 1628, so it is curious that (as has been remarked) Van Dyck endowed his sitter with a thinning head of hair, for Rubens was already bald by the middle of the second decade.18 Is it possible that Van Dyck was unaware of this, because Rubens may have remained covered (i.e. did not remove his hat) when they met? Vey elaborated Liedtke’s19 proposal by suggesting that the sitter’s ‘forward leaning pose’ was inspired by Rubens’s own Self-Portrait in the Rubenshuis.20 Gregory Martin, 2022