Since 1931 Jan Vermeyen's teacher has been thought to have been Jan Gossaert rather than Jan van Scorel. Yet the Holy Family in Haarlem (Fig. I), which may claim to be his earliest known work, clearly shows the influence of both these artists. The main obstacle to the acceptance of a close relationship between Vermeyen and Scorel was a problem of chronology: Scorel returned from Rome to settle in Utrecht in I524, while Vermeyen was supposed already to have entered the service of Margaret of Austria as a court painter in I525. However, the documentary evidence available allows of a somewhat different interpretation. It is likely that Vermeyen did not get his official appointment until May I529 and that he moved from Utrecht to Malines during the course of I527, possibly at the same time as Jan van Scorel left for Haarlem and because of the same unrest in Utrecht. A petition handed in by Vermeyen after the death of Margaret of Austria on 27 November I530 mentions a number of works that must have been executed before that date. These include not only two portraits of Erard de la Marck, Bishop of Liege (cf.Fig. 2), but also one of a certain Jean Denis, which is stated in the inventory of the princess's estate to have been done by candlelight. In view of the description of this lost work, the Marriage at Cana recently acquired by the Rijksmuseum (Figs. 4 and 5) may well be a relatively early work. A date in the early I530's would be in keeping with the costume of the lute player (a portrait?) shown there, while the style of painting is close to that of the portrait dated I53 I of Felipe de Guevara (see Note 6) . Also very similar in style, particularly in the use of chiaroscuro, is a Head of a Woman, a fragment in the Kunsthistorisches Museum in Vienna (Fig. 6) which is attributed to Jan van Scorel, but differs from his work precisely in the greater contribution the chiaroscuro makes to the suggestion of plasticity. A Holy Family by the Fire in Vienna (see Note I9), though also dealing with the effect of artificial lighting, would seem to represent a somewhat later phase in that the modelling of the figures lends them a relief-like character. There has been a tendency to date this phase around I545, i.e. in the proximity of Vermeyen's etchings on which the date, if there is one, is invariably I545 or I546. Some of the etchings do bear a close resemblance to the Holy Family by the Fire, but it should be borne in mind that they may well reproduce much earlier works, as is demonstrably the case in quite afew instances: a portrait of before I530 (Note 3), a type of woman already found in the early Holy Family in Haarlem (Note II) and motifs that Vermeyen must have recorded during his visits to Tunis in I534-5 and Spain in I539 (Note 25) . Since the etchings are not, therefore, reliable documents for dating other works in a similar style, a painted copy of a lost work from the Tunisian campaign takes on an added significance. This is Rubens' copy of the portrait of Mulay Ahmed, now in the Museum of Fine Arts in Boston (Fig. g) . Although Rubens' interpretation of Vermeyen's original no doubt contributes to the powerful qualities of his copy, something of the firm plasticity and colourful appearance of the sitter must be due to the protoope and may be taken as typical of Vermeyen's style of the mid-I530's. Thus like Maerten van Heemskerck, though superior to him as a painter, Vermeyen seems to have broken away from Jan van Scorel's influence in the course of the I530's.