Technique and condition

Cross sections showed that the shadows of the robe had been built up on the opaque pale red ground using a two layer translucent glaze structure. The canvas is primed with a thin yellow ground which closely follows the topography of the canvas threads. The ground contains yellow, white, black and red pigment particles which are most likely to be yellow ochre, vermilion, bone black and lead white, although this was not confirmed using a dispersion of the pigments. The pale red ground contains large translucent particles, white, black, red and orange pigments. Under UV light in cross section some of the white translucent particles appear fluorescent and globular. This may be an example of a reaction of the medium with a pigment, creating a halo of crystalline material, and will be further investigated at the FOM Institute for Atomic and Molecular Physics, Amsterdam.

The first layer over the base is a semi-transparent glaze, consisting of a red lake pigment, vermilion, burnt sienna, white and yellow particles. The medium of this layer fluoresces with a pale green colour indicating that it contains a significant proportion of a natural resin. Cleaning revealed that the glazes exhibit a dull pink fluorescence when the painting itself is illuminated with UV light, the colour of the fluorescence clearly affected by the transparent lake particles in the layer above imparting a pink appearance. Dubois probably added resin to his paint to give increased transparency to the red glazes of the robe and his technique has been successful. Some of the artists in Rembrandt’s circle also used this technique in the seventeenth century, usually adding small amounts of pine resin varnish to their paint [1].

The top layer is a more transparent glaze of the red lake, with occasional vermilion and white particles. The medium appears quite fluorescent but with a pink colour, indicating possibly that the layer in cross section is illuminated by light coming from the resinous layer below, mimicking the effect seen on the paint surface. Varnish has soaked into the upper part of this glaze resulting in a green fluorescence towards the surface and it may have quite a porous texture. Thin horizontal lines of fluorescent material can also be seen inside this glaze, which are possibly interfaces between brushstrokes of the same paint and may be evidence that Du Bois applied a thin film of oil between painting sessions, so-called oiling out. The fluorescent channels contain dark material, possibly dirt trapped in between successive applications of glaze.

Dispersed samples of pigment from the upper glaze layer were examined using a polarising microscope. The major component of the glaze proved to be a red lake on an alum (aluminium hydrate or hydroxide) base. EDX analysis corroborated this observation, showing aluminium as a major element present in the sample. The lake particles showed a range of sizes, many of them quite large. The intensity of the colour varied from a pale pink to a bright crimson, the brightest particles being smaller in size. Some particles showed an uneven coloration across their surface. The dyestuff is likely to be of scale insect origin, kermes, cochineal or lac. The particles do not show the characteristic fluorescence of madder and a brazilwood lake would probably be more faded [2]. Occasional bright red birefringent particles were identified as vermilion and fairly large transparent isotropic particles were identified as ground glass. They do not show any colour at all and are thus unlikely to be very pale or faded smalt. Glass (as well as smalt) was recommended as a drying agent to be added to paint containing poor drying pigments such as red lake[1][3]. Tiny anisotropic chalk particles were identified and coccoliths were found, identified by their characteristic display of a stationary cross under crossed polars. The chalk was difficult to differentiate from small lake particles as it appears pink in meltmount due to its low refractive index. However well-separated particles were observed to ‘twinkle’. Dubois may have added chalk to his glaze to impart body to the paint – the surface of the glaze still retains grooves from his brush so it must have been of a pasty consistency when applied. Chalk was used by Rembrandt in this way as it did not affect the colour of the glaze and also gave extra translucency [1]. Some rounded transparent grains of a brown earth pigment were identified in the dispersion clinging to the base of the upper glaze. These probably originated in the lower glaze layer but were not separated enough from other material to allow proper analysis.

Certain passages have been painted using an interesting technique or unusual materials which warranted further investigation.

The Deep Red Shadows of the Robe.
Even with the naked eye, pale patches or inclusions of glittery material could be seen within the translucent glazes forming the deep red shadows and contours of the sitter’s robe. This could be more clearly seen when the grey scabby overpaint was removed from the grooved texture of the darkest shadows. Their appearance initially brought to mind scraps of gold or metal leaf. Under magnification the patches of material were revealed to be quite large and in worn areas of paint they broke the surface, appearing to be clumps of a pale yellow or white crystalline material. However in these areas they might have been affected by the reagent responsible for the damage. It was difficult to tell if the inclusions were incorporated into the transparent red glaze or lying along an interface on top of the opaque base colour. The surface was photographed under high magnification.

Glittery Patches

The cross-sections reveal that there are roughly horizontal bands of glittery material embedded above the interface between the upper and lower glaze. These bands are not flat and appear to be a conglomeration of pale particles within the glaze. It could be hypothesised that these are clumps of unblended pigment which sank towards the bottom of the glaze once applied. The reflectance of the clumps may have been increased if the gaps between the particles are filled with air. Orpiment was suspected to be the pigment forming the glittery patches but this pigment was not identified in any of the dispersions.

Yellow Lapel

The lapel of the sitters robe also appeared to be composed of a light yellow glittery pigment. This was examined and photographed under high magnification. This area was eroded, age cracks in the paint layer had been eaten away, presumably by ill chosen solvents used to clean the painting in the past. Samples of this paint were also taken (T07469/1 & 6). EDX analysis was carried out, identifying arsenic and sulphur as elemental components of the pigment and therefore confirming the identity as orpiment, arsenic trisulphide.


  1. Kirby and White, Rembrandt and his Circle: Seventeenth Century Dutch Paint Media Re-examined, National Gallery Technical Bulletin,
  2. Kirby and White, The Identification of Red Lake Pigment Dyestuffs and a Discussion of Their Use, National Gallery Technical Bulletin,
  3. Talley, M.K., Portrait Painting in England: Studies in the Technical Literature Before 1700, 1981, Chapter 6 the De Mayerne MS, pp.94-6.

Helen Brett
May 2001