WHEN working on Caravaggio's Cardsharps, which I published at the beginning of 1988,1 I had to question the assumption, more or less taken for granted during the present century, that Caravaggio himself could not have painted two closely related renderings of a single theme. That assumption must have originated in part from mod- ern prejudices concerning the nature of creativity, and is doubtless also partly to be put down to not unreasonable caution at a time when familiarity with Caravaggio's own hand was at a more rudimentary stage than at present, coupled with the daunting existence of a considerable number of early copies of his paintings. For this tacit habit of thought to be overcome it was necessary for firm documentary and technical evidence, together with convincing art-historical arguments, to be adduced in support of what, in strictly individual cases, the visual data might appear to indicate.

THE re-identification of the Lute-player painted by Caravaggio for Cardinal Francesco Maria del Monte, discussed in the preceding article by Denis Mahon, carries with it a number of implications for our understanding of the Lombard artist. Not the least of these is his reputed habit of painting directly from the posed model, as reported by his earliest biographers as well as by one of his most eminent patrons, the Marchese Vincenzo Giustiniani.' Giustiniani, of course, owned the related composition of the same subject, now in The Hermitage, Leningrad, and he certainly knew Del Monte's picture as well. The purpose of this essay is to examine the precise relationship between these two pictures – painted within a year or so of each other for two men who shared cultural and artistic interests – and to elucidate the creative process underlying their appearance.

IN 1983 the Pushkin Museum acquired a small oval portrait (Fig.20) which the present writer had been fortunate enough to discover in the home of Mikhail M. Ouspensky, a Moscow collector. Style and costume indicated that it was of the Dutch school of c.1600–10, while the fluency of execution and subtlety of technique were those of an outstanding artist. The image did not, however, conform to any of the established types of portrai- ture current at that time, whether in oil, miniature or engraving, although in the last oval portraits of this size were fairly common, and drawings for such prints – for example by Jacques de Gheyn II – also exist. The Moscow pictures, although individual enough as a portrait, does not present its sitter with the marked features of a Jacques de Gheyn portrait-drawing, nor with the bland reserve characteristic of those by Michiel van Mierevelt. Instead it shows a certain preciosity characteristic of some trends in late mannerist history painting, and in particular in the smaller works of Joachim Anthonisz. Wtewael. We hesitated over attribution to this artist, however, as no painting of this kind had, until recently, been accepted into his oeuvre.

IN 1897 the Staidtische Kunstsammlungen in Freiburg (now the Augustinermuseum), which are devoted almost entirely to the regional art of the Upper Rhine district, acquired from the estate of a local lawyer, Ludwig Riegel, a painting in oil on copper 'in Rottenhammer's manner' of St Michael vanquishing the rebel angels (Inv.no.12561; Fig.24). Classified as 'South German, mid-seventeenth century', it was relegated to the store- room, remaining unpublished and unknown to scholars, its artistic quality unappreciated. Recently it became possible to determine its true authorship in the simplest conceivable manner: by visual comparison with a painting recognised to be one of the major works of Giuseppe Cesari, known as the Cavaliere d'Arpino (1568-1640), which is among the treasures of the Glasgow Art Gallery.