This bust-length portrait shows a man in three-quarter profile wearing a black bonnet with earflaps, and a fur-trimmed black and dark blue gown over a white shirt with a gathered collar and a red jacket. The panel is signed and dated very prominently on a trompe l’oeil piece of paper attached to the background with two pins. The painting has been regarded as a self-portrait by Jacob Cornelisz since it was bought in 1887.4 This was based on the man’s twisted pose, the result of the artist looking sideways in a mirror in order to paint his portrait, combined with Jacob Cornelisz’s monogram. The identification was further confirmed by the double portrait that Jacob’s son Dirck (c. 1497-c. 1567) painted of his parents, which is now in Toledo (fig. a). That painting, which can be dated around 1550, shows exactly the same man as in the Amsterdam panel, but this time holding brushes and a maulstick as he works on a portrait of his wife. X-radiographs and infrared reflectograms of the double portrait revealed that there was a second portrait of Jacob Cornelisz beneath that of his wife.5 Initially, then, the artist was shown painting his self-portrait. A tracing made on transparent paper (fig. b) showed that the Amsterdam portrait is exactly the same size and shape as that of Jacob Cornelisz in Toledo, and the same is true, although reversed left for right, of the second portrait of Jacob Cornelisz in that painting, which was ultimately replaced by that of his wife but is visible in an X-radiograph (fig. c).6 This means that there was a model for this portrait in the workshop, probably a pricked cartoon, which could be used whenever needed, either as it was or in mirror image. The mechanical underdrawing of the Amsterdam painting, which consists solely of contour lines, confirms the suspicion that the Amsterdam portrait was based on that cartoon (fig. d). Although the signature on the Amsterdam painting and the relationship to the one in Toledo make it clear that we are dealing here with a product from the workshop of Jacob Cornelisz van Oostsanen, the attribution of the former to the master himself is problematical, as is its dating. Van Eeghen suggested in 1986 that Jacob Cornelisz may have died in 1532, ruling out the possibility that he painted the self-portrait.7 In addition, this painting differs technically from others by Jacob Cornelisz. It has a hard, smooth surface, whereas most of his paintings have thick layers of paint applied in a draughtsman-like manner with a small, stiff brush and thick, sturdy paint. Even the paint layers of his last dated works of 1524 and 1526 (SK-A-1349 and SK-A-668) which are themselves departures from his earlier paintings as regards technique and iconography, are not as thinly brushed as this self-portrait. Another discrepancy is the abundant use of lead white in the light passages in the face and neck, as can be seen in X-radiographs. This is not found in the X-radiographs of other paintings by Jacob Cornelisz in the Rijksmuseum. The strong chiaroscuro, with almost black shaded passages, is also unusual, as is the so-called turbid medium effect in the shaved area of the face, which was created by applying a semi-transparent layer over relatively dark paint. The combination of the above-mentioned factors makes it clear that this is not an autograph self-portrait by Jacob Cornelisz van Oostsanen, nor can it be attributed to the painter of the double portrait in Toledo, Dirck Jacobsz, because the style is different. The latter work is executed more softly in light pastel colours, and the paint surface is thicker and, in the portrait of the woman in particular, the figure is modelled with quite broad, lively brushstrokes. The most likely explanation is that the Rijksmuseum portrait is a repetition of an older self-portrait that is now lost.8 That painting would have been one of the first Netherlandish self-portraits. The Amsterdam copy might be the work of Jacob’s grandson, Cornelis Anthonisz (?-1553), whose Copper Coin Banquet of 1533,9 with the portraits of 17 members of the Amsterdam crossbowmen’s civic guard, may portray a different kind of head but is not that far removed from the Rijksmuseum portrait in its technique. (Daantje Meuwissen)