Csornay Boldizsár - Dobos Zsuzsa - Varga Ágota - Zakariás János szerk.: A Szépművészeti Múzeum közleményei 97. (Budapest, 2002)

FORRAI, KORNÉLIA HOÓS, MARIANNA-SOMOS, ÉVA: The Research and Restoration of a Painting on Panel, Transferred to Canvas, by Domenico Puligo. Madonna and Child with Saint Sebastian and Saint Roch

period were often prepared with red-brown paint, applied with brush; sketches of this kind could not be shown by infrared reflectograms either.) We have found that very little was left of the original ground in this picture, too. The cotton muslin cloth

19 that had been lined directly onto the back of the painting, was presumably adhered to it with a starch-glue paste.

20 Once the muslin was attached, the picture, presumably still wet, was put into the press.

21 After it had dried up, a thicker linen cloth was attached to the muslin

22 to reinforce it, but at that time no relining adhesive was applied at the edges. It is quite possible that the painting was once again taken under pressure. Another linen canvas had been fixed to the prepared stretcher,

23 and the two canvases were adhered to each other. As at that stage the painting had already been attached to the stretcher, there is reason to believe that no pressing was performed then. This action was followed by tacking the middle canvas to the stretcher and soaking off the protective coating from the paint surface. Not until all these processes had been fulfilled did they fix the muslin to the stretcher.

24 The glue was not perfectly removed from certain areas of the paint surface, which may have caused whitened and opaque stains to appear scattered on the surface. To remove this side effect of wet treatment, the surface got a new "varnish" layer.

25 Retouching was applied only in areas of losses, stoppings and the immediate proximity of these. This may have been, according to our view, the final step within the process of transferring the painting to a new support. Nevertheless, at a later date, the picture was subjected to a major restoration and, presumably, a number of subsequent regenerating varnishing (figs. 33-37). The crack that can be observed in the three cross-sections (figs. 33-35) promotes our understanding of the sequence of layers and events. The substance at the bottom of the sample (being yellowish and gelatinous in the standard photograph, red in the tinted one, and light, fluorescing in yellow, under UV light) had, through the already existing crack, flowed along to the protective coating. We assume it to be the starchglue paste. The other adhesive, which is seen in the standard photograph to have been slightly penetrated into the crack from above, is darker and more bluish under UV light Thin-yarned, loosely woven, therefore light and well-falling, women's dress material of cotton, silk or wool. It was named after the Turkish town of Mossul. 20 The adhesive was inspected with wet chemical analysis and cross-section tinting process (iodine test, protein test). The results were compared with descriptions, and enabled us to deduce the probable adhesive. 21 Presumably that was the moment when the weave of the canvas was impressed into the paint film. 22 Zoltán Szálai helped us in examining the textiles. 23 The contemporary sources do not provide any information as to whether the canvases had been initiated. 24 Each of the three canvases were fixed to the stretcher with a separate row of tacks, and each were stretched for the first time. The marks of handworking are recognizable in several places of the stretcher. 25 We did not investigate the material of this coating, but it is known that the practice of mixing shellac or glue with boiled starch was a usual one even in the middle of the 20th century. The surface of a painting treated with such a mixture became flat and uncracked. and the opaque varnish became transparent again.