In fresco, though a ground is prepared for applying the paint layer, the two are hardly distinguishable in the completed work. In fresco secco the lime water used as a medium is of the same composition as the ground; while in buon fresco the pigment is incorporated in the outer layer of the ground itself. Also, fresco when painted did not have a protective layer put over its surface, though later generations have sometimes tried to supply one.
In contrast, in the painting processes still to be described, those of tempera and oil, the paint layer is quite distinct from the ground and is often of different composition, and a protective layer is almost invariably applied. Thus, we are now concerned with the most elaborate form a painting can take.
It has already been mentioned that this term is sometimes applied to various kinds of size painting. Historically, there is nothing against this. Cennini uses the word ‘tempera’ practically as the equivalent of ‘medium’, and writes of a size and of an oil tempera. In practice, however, and as the result of modern usage, it is better to confine the term to that form of painting in which the medium is egg.
In medieval Europe this was the medium most extensively used for painting on wood panels, until in the course of the fifteenth century it was displaced by various types of oil and varnish media. Very occasionally, egg tempera was used on linen or canvas; but in the Middle Ages, paintings with textiles as supports were generally for temporary purposes, such as banners and decorations at religious and other festivals, and as noted above, the medium in such cases was likely to be some form of size paint. Especially was egg tempera used in Italy; and in that medium, the great altar-pieces and smaller panel paintings of the fourteenth and earlier fifteenth century were carried out.
It is not surprising, therefore, that the standard account of how to use tempera in its most elaborate form is that given by Cennino Cennini (see the previous chapter on Fresco). It is, indeed, largely due to him that, after long disuse, there came a revival of tempera painting, which continues today. In the earlier nineteenth century, reflecting a revived interest in early Italian painting, Cennini’s treatise was rediscovered and transcriptions and translations into various languages were made, thus putting at modern painters’ disposal an exact and authoritative account of early Italian methods.
Invariably, in tempera painting, the wood panel has to be prepared to receive the paint. Even if carefully planed and sandpapered, it is apt to be uneven in texture, due to knots and inequalities in grain, and so would absorb the paint unevenly; while resins in the wood would certainly in time soak into and discolour the paint. To the panel, therefore, a ground of gesso (a mixture of size and plaster of Paris) was applied.
First, however, Cennini recommends that the panel be given a coat of size, saying: ‘And do you know what the first size, with water, accomplishes? Not being so strong, it is just as if you were fasting, and ate a handful of sweetmeats, and drank a glass of good wine, which is an inducement for you to eat your dinner. So it is with this size; it is a means of giving the wood a taste for receiving the coats of size and gesso.’
The next step was to fasten to the panel with size thin linen or canvas, in strips if the panel were large. This had a double purpose; it provided a finely toothed surface to which the gesso could adhere, and it provided further protection against resins from the wood discolouring the gesso or the paint. The gesso was applied in several coats.
The earlier ones were what Cennini calls gesso grosso (thick gesso), which consisted of plaster of Paris mixed with size made from parchment clippings. Plaster of Paris is calcium sulphate prepared by roasting gypsum or alabaster. When mixed with a little water or size, it crystallizes into a hard mass, familiar to everyone in the form of plaster casts. Every layer of gesso had its surface carefully smoothed, but the gesso grosso always had a slightly granular surface, to which the next coat of gesso could adhere firmly.
The gesso ground was finished with a layer or layers of gesso sottile (thin gesso), as Cennini calls it. This is prepared by mixing plaster of Paris with a great deal of water, which prevents it setting, and reduces it to a very fine powder, which in Cennini’s words, ‘will come out as soft as silk’, and is then mixed with size. Finally, the gesso sottile was scraped as smooth and flat as possible with a knife.
For large paintings, several pieces of wood had to be joined together, and this was done with a glue made by mixing cheese with quicklime, a homely blend that time has proved to be both effective and permanent. In the Middle Ages, also, it was usual to make panel and frame together, the frame in the case of larger altar-pieces being as a rule elaborately carved. That is why, to his description of preparing the panel, Cennini adds instructions for covering the frame also with gesso grosso and gesso sottile to prepare it for being gilded.
Even in Cennini’s time the elaborate procedure he details was sometimes simplified, as it often is today. He himself, in the case of small paintings, says that it is enough to size the panel, and put on several coats of gesso sottile; and modern restorers, when they have had to transfer a tempera painting from an old and decayed panel to a new support, generally find that the canvas or linen under the gesso, and the layers of gesso grosso, are only used in the case of large panels composed of several pieces.
The gesso, too, was not necessarily made from plaster of Paris. In Cennini’s time, according to his own account, wood ashes mixed with size might be substituted for the gesso grosso. Again, chalk (whiting) can be used throughout instead of the plaster, as commonly happens today; while today, also, canvas is sometimes used as a support instead of wood. But no substitute gives quite so firm and smooth a surface as plaster of Paris, nor does canvas give so firm a foundation to the gesso as does wood.
For painting, powdered pigments are mixed, at the time of painting, with egg yolk alone, or with both yolk and white. These are composed of the same substances, though in very different proportions. In the white, the binding element is albumen and there is very little fat, and by itself it forms a brittle film, soluble in water. The yolk, however, contains a considerable proportion of oil as well as albumen; and these together dry into a tough film, hardly soluble in water. Anyone who has tried to wash an egg spoon some time after use will know this. A mixture of white and yolk naturally increases the proportion of albumen in the medium, and so makes it somewhat more brittle, and more liable to be affected by water.
Both white and yolk contain large proportions of water; but it is usual to add water as a diluent, to make the paint easier to work. For the same reason, a little vinegar used to be added to the yolk, since this made it somewhat less greasy. Today, a little acetic acid is sometimes used for the purpose. In both white and yolk is a fatty substance called lecithin, which also contributes to easy working, by helping to emulsify (i.e. make into a mixture substances otherwise repellant to mixing) the oil and the non-fatty parts of the medium.
Tempera dries quite quickly, even when considerably diluted with water. Mixed with a pigment its transparency is between transparent water-colour and oil, and cannot be relied upon to conceal anything underneath it. Consequently, modifications and corrections are difficult to make, once the paint has dried. This has several consequences in practice.
Almost invariably, before painting begins, a detailed drawing of the subject is made on the gesso. Cennini, for example, advises that the drawing be first made in charcoal, practically effaced, and then gone over with ink and a brush, the remains of the charcoal then being brushed away. Preparatory to this drawing would, of course, be a number of other drawings and studies, similar to those made in preparing a cartoon for fresco.
On the basis of this drawing, the paint is built up in a series of layers, much in the same way as in transparent water-colour. Frequently, the system of light and shade was indicated by a wash of terre verte (green earth colour), over which the other colours were laid, light at first, to be strengthened and enriched by subsequent layers, if required.
This green underpainting is often to be seen in fourteenth-century paintings, especially in the shadows on the flesh. A detail of a fourteenth-century Sienese painting, both the underlying drawing and the terre verte lay-in of the shadows can be seen. For the underlying drawing see also the head of an angel from a painting by Cosimo Tura. Over this green underpainting, the forms would then be fully modelled, and details added by smaller touches of paint. These touches were not fused but laid side by side, like hatching or stippling in water-colour, since the paint having to be thick enough to allow precision of touch and density of substance, dried quickly, and so did not allow one touch to melt into another easily.
When the painting was finished, and thoroughly hard and dry, varnish was applied both as a protection, and to improve appearance. Cennini recommends waiting as long as possible before varnishing and adds, ‘if you varnish after the colours and their temperas have run their course, they then become very fresh and beautiful, and remain in pristine state forever’. Cennini does not give a recipe for his varnish; but it was apparently not free-running, since he speaks of its being applied with the hand or a sponge, which suggests a resin dissolved in a drying oil, of the kind described later in the section on oil painting. In modern practice a coating of wax is often used to protect the surface of tempera paintings.
Various methods have been used to retard the rapid drying of tempera, and so give it greater malleability. Free use of water does something, but makes tempera somewhat like transparent watercolour in use, and is not practicable in the final stages of a painting. Some modern. recipes therefore recommend the addition to the egg of a considerable quantity of oil.
This certainly retards drying, and probably toughens the paint film (though this is hardly necessary); but the oil is liable to turn yellow, which may explain why some modern tempera paintings have darkened and discoloured. In addition to difficulties of manipulation, tempera has similar, though less marked, limitations to those of fresco. The albumen in both the white and yolk of egg contains sulphur, which in contact with pigments containing lead or copper, such as white lead or verdigris, may cause formation of black sulphide of lead.
The range of tone in tempera, though greater than that in fresco, is less than that in oil, and marked variations in the thickness of different parts of the paint layer are not possible, so that the painter has fewer means of securing dramatic emphasis than in oil. But in luminosity and subtle delicacy of colour, and in beauty of texture, tempera is one of the most remarkable processes available to the painter. Moreover, its results are very durable. The medium encloses and protects the pigments well, and is not itself liable to disintegrate or to change in colour and transparency.
It is a much disputed question whether tempera paintings were intended to be seen as they now appear in many museums, the paint covered only with a thin layer of wax or colourless varnish with a dull surface, which does not affect the quality of the colours and the high key. It has been argued that the varnish of which Cennini speaks was intended both to increase the richness and depth of the colours, and to give the whole painting a mellow tone.
Certainly, in paintings where the underlying modelling is in the dull green of terre verte, a varnish toned slightly yellow gives what is, to some modern eyes, a more harmonious and unified effect. Cennini says nothing about toning his varnish, though there is reason to believe that in any case it was dark and yellowish. The reasons he gives for its use are that it gives the colours freshness and beauty and preserves them.
If this means, as it may mean, that the varnish gives the colours the appearance they had when first applied and still wet, then the case for the early painters using what was in effect a toned varnish is strengthened, since the varnish would have just this effect. However, we are unlikely ever to know how the medieval painter intended his tempera painting to look; since in the rare cases where a painting retains some of its early varnish, and this has not been covered by varnish of a later date, the early varnish is so discoloured and fragmentary as to give no clue to what it originally did for the appearance of the painting.
Hits: 89