This Virgin, whose cheek touches that of the Christ Child, is based on the Byzantine Elousa type, or Virgin of Tenderness. The best-known example of the type is the Italo-Byzantine Madonna of Cambrai, bequeathed to the Cambrai Cathedral in 1426 as an original work by St Luke, on which many early Netherlandish paintings were based.7 Interestingly there is an extract from a Marian hymn inscribed along the edges of the Virgin’s cloak.8 This is one of the three known versions of this Virgin and Child, and was published as a work by Bernard van Orley by Friedländer in 1909.9 A second version is in Pommersfelden (Schönborn Castle), while the whereabouts of a third, which was with the Hague art dealer K.W. Bachstitz in 1932, are unknown.10 Friedländer described the painting in connection with another Virgin and Child in Ottawa, and dated both of them c. 1520.11 The work in Ottawa also provided the clue for the attribution to Van Orley because it was the left half of a diptych with the portrait of Margaret of Austria on the other half, whose court painter Van Orley was.12 Friedländer considered that the Rijksmuseum version was the equal of the one in the Hague gallery,13 but Baldass felt that the latter was a later version derived from the one in the Rijksmuseum, and suspected that the Pommersfelden panel was the original.14 Going by the available photographs, the Amsterdam version is the best of the three. In the other two there is no text on the trim of the cloak. The one formerly in the art gallery has an aureole around the Virgin’s head, and a veil has been added over the Child’s body, the originality of which is doubtful. The delicacy of the Virgin’s face in the Amsterdam panel argues for the attribution to Bernard van Orley, and is very comparable to those in other paintings which are generally accepted as being by him.15 This is underscored by the affinity with the Ottawa Virgin and Child. The faces have the same slightly elongated shape, closed eyes, small mouth with full lips, long nose and light blush on the cheeks. The hair is also meticulously rendered. This type is easily distinguishable from similar Virgins from the workshops of Quinten Massijs and Joos van Cleve. The main factors arguing against the attribution to Bernard van Orley include the decidedly feeble depiction of the Virgin’s hands. Their rather unconvincing anatomy, the exaggeratedly long fingers and nails are atypical for Van Orley, who tended to paint chubby, expressive hands. The underdrawing, too, provides insufficient indications for a more precise attribution. Too little is known about the division of labour in Van Orley’s workshop to speculate about whether or not the master would have touched up a face in an assistant’s product. It seems unlikely, though, that more than one hand was responsible for a work produced in a relatively small series. (L. Hendrikman/J.P. Filedt Kok)