Technique and condition
The painting is on a mahogany panel made of a single, unbevelled piece of wood with the grain running vertically. There is a layer of mushroom brown paint all over the back with a Winsor and Newton label over it. The support is in excellent condition.
It is primed with a thick, white ground, almost certainly applied by the colourman. It is not known whether Millais used the recorded pre-Raphaelite technique of applying a second ground before starting to paint and working wet-into-wet. Many cross-sections would be needed to establish this but the edges of the painting are in such good condition that we could not justify taking samples of this nature. The ground is very securely attached to the support. There are long, very fine, horizontal cracks in the ground and paint over most of the composition; they are spaced between 5 and 30mm apart, their edges are flat and there is no sign of flaking.
Infra-red photography reveals confident linear drawing in graphite or fine crayon in Mariana's torso and arms. Sketchier lines indicate the back of her skirt, the window embrasure, the position of the tablecloth in front of her and the stool behind her, which originally was conceived at a more foreshortened angle to the viewer. None of these lines is easily visible with the unaided eye.
The painting is outstandingly well preserved. Apart from the horizontal cracks, which are visible only in strong reflected light, and some very fine, barely distinguishable drying cracks in the dark curtain near the top right corner, the paint surface is virtually unblemished by age. A short line of darkened paint on the top edge at the right corner may have been applied to cover abrasion from the frame but it is possible that the artist did it.
The technique of painting is varied. In most areas - the dark background, green foliage, flesh tones, shadows of the blue dress and the white cloth in shadow - the paint is semi-opaque and applied thinly in little hatches and dabs with fine brushes, so as to exploit the reflective properties of the white ground. In other parts - the stained glass windows, red seat and the leather wall hanging - the paint is thick and glutinous-looking and was applied in blobs of various sizes. Either it was fully thixotropic or it was applied while the painting was placed horizontally face up because it has not run at all. Analysis with GCMS ( K.J.van der Berg, FOM Institute, Amsterdam, 1999) of red paint from the stained glass window at the left edge shows that the medium is a fusion of copal resin and oil. On the other hand a sample of green from the leather wall hanging (top edge) revealed heat-bodied linseed oil with no additives. It looks as if Millais used two different blue pigments for the dress; the shadowed parts look like Prussian blue while the highlit area resembles ultramarine. The highlit blue looks much built up.
The varnish is a natural resin applied fairly thinly. It may be the original, as there are no signs that the painting has been cleaned. As it had sunk a little in the dress and parts of the dark background, a fresh coat of dammar resin in white spirit and Shellsol A was brushed on top after the painting was acquired by the Tate and had been surface cleaned with Stoddard Solvent. When dry, the surface was polished with a soft brush to reduce the glossiness a little.
Rica Jones, (with additional information from Dr Joyce Townsend and Dr Klaas Jan van der Berg)