Vitruvius, in Book VII, Chapter III prescribes the application of three layers of sand mortar followed by three layers of plaster as the ideal support for the painted decoration. In my experience, working on Roman painted plaster from a number of sites in Southwark, just one layer of mortar and one of plaster was normally used, the Winchester Palace fresco was unusual in having two layers of mortar and two layers of plaster.
In Figure 2 the side view of a fragment of painted plaster from the edge of the lunette can be seen. Figure 3 illustrates the various layers of plaster and mortar in this fragment.
In Figure 2 it will be seen that the second and final layer of mortar is slightly darker than the first layer of mortar, and has a very faint reddish tinge. Within this second layer layer of mortar there are many small fragments of tile or brick, these were presumably added for the hydraulic qualities they would contribute, making the mortar more resistant to damp, a necessity for the mortar support situated in a bath suite.
The first layer of mortar, some 30 millimetres thick, contains a great deal of sand and gravel, but no tile or brick, it also contains many voids left by vegetable material, mostly straw or grass, this was added when the mortar was prepared in an attempt to prevent shrinkage and cracking of this first layer of mortar while it was setting.
The first plaster layer is white, the second and final layer is yellow due to the addition of yellow ochre to the lime plaster. The layout of Figure 4 is shown in Figure 5.
The tiny area outlined in red in figure 3, is shown in a highly magnified polished section in figure 6 with a key in figure 7. This area, some two millimetres high, is a section through the layers of paint, plaster and the very edge of the final layer of mortar.
Layers 1 and 2 are layers of paint, layer 1 is blue frit, or Puteoli blue on top of layer 2, a dark red ochre. The area shown in the polished section can be seen in figure 1 to the right hand side of the fragment where a line of pale blue paint borders the red ochre area and separating it from the yellow ochre.
Layer 3 is the yellow ochre plaster layer on top of layer 4 of white plaster Layer 5 shows the final layer of mortar.
A calcite crystal is seen at ‘a‘, one of a number in the section, is embedded in the final layer of yellow plaster, Professor Roger Ling has pointed out in his work ‘Roman Painting, 1991‘, page 204, that the presence of calcite crystals in this way may be deliberate rather than accidental in attempt to improve the brilliance of the burnished plasters lustre.