IN recent months there has been increasing awareness of our studies of the mural painting of Guidoriccio at the siege of Montemassi, long regarded as a textbook example of Gothic art of the trecento (Fig.4, upper mural). Our purpose in what follows is to publish some substantial evidence we have not published before, evidence to support our view that the Guidoriccio is not a documented work of 1330, not by Simone Martini, and not of the trecento. We are presenting this material in relation to evidence we have published previously, in order to provide a record of the present state of our research and also to point out the lacunae, which we hope will eventually be filled.

THIS contribution to the problem of the Guidoriccio fresco in Siena had modest origins. Its aim was to explain the nature of the issues, to sort out those arguments which appeared to the writer to be central from the many which seemed to him to be peripheral and occasionally irrelevant and to indicate where crucial evidence was still lacking. The occasion of its composition was the immediate aftermath of the Simone Martini convegno held in March 1985. This was in almost every respect an excellent meeting. For a number of reasons, however, the Guidoriccio question was not effectively discussed.

ONE of the great revelations of the cleaning of the Judgment of Solomon (Fig. 18) was the rediscovery of three or more stages of a sixteenth-century masterpiece, hitherto obscured below successive restorations. The decision to clean this painting was prompted by the need to consolidate the paint surface, which was flaking in many areas. Without the removal of the thick varnish layer, consolidation of the paint would have only a limited success. In addition, the varnish showed a marked discolouration, distorting the colour balances, while the extensive restoration disguised the painting as a whole. For these reasons it was, therefore, decided that cleaning would be both necessary and beneficial.

AS part of the project to redecorate and re-hang the main gallery of seventeenth-century Italian painting in the National Gallery, London, it was decided to clean and restore the only two works in the collection by Giovanni Battista Salvi, called Sassoferrato: The Virgin and Child embracing and The Virgin in prayer. Neither painting had received any form of treatment (other than minor blister treatment in the case of The Virgin and Child embracing) since the nineteenth century. Both had become extremely disfigured by old varnish layers, which were not only very discoloured, but also had deteriorated to such an extent that the once transparent films had become cloudy and opaque. While the discolouration of a deteriorated natural resin varnish is most obvious over areas of light-coloured paint, the loss of transparency is more marked over the darker colours. It is also possible that varnish over dark areas of paint is inclined to deteriorate faster than that over lighter areas, because dark colours absorb more heat, therefore accelerating the oxidation and break-down of the resins. This may account for the extreme opacity of the varnish in the shadows of the drapery folds in The Virgin and Child embracing, particularly in the green curtain and blue robe, where little could be seen of the strongly three-dimensional modelling suggested in the X-radiograph made as a preliminary to the cleaning (Fig.45).