At the time that some mural painting of the thirteenth century was discovered in the Sainte Chapelle at Paris an analysis was made of the colours by MM. Dumas and Persoz. The result showed that the whites were probably a preparation of lead, perhaps flake white; the drapery blues were phosphate of iron, probably native, and the other blues were ultramarine; the bright red with which the aureole round the angel's head was painted was vermilion, the browns and yellows were ochres; the greens were made with these ochres mixed with phosphate of iron, and the roses and violets were probably obtained from shell fish by a simple mechanical preparation.*

* The question of the Tyrian purple was investigated in 1859 by Lacaze-Duthiers. He found that when the matter which becomes purple is taken from the murex it is white or slightly yellow, sometimes slightly grey. (It is a little band on the lower face of the mantle between the intestine and the gills.) Submitted to the action of the son it becomes first citron yellow, then greenish yellow, then green, and then tarns to violet which becomes darker and darker. At the appearance of the violet a penetrating smell of essence of garlic is given off. By varying the quantity of the material and the length of exposure to the sun, drawings may be made of vigorous tone and delicate broken tints. Lacaze-Duthiers made photographs with it which Cros and Henry say were quite successful. The matter is soluble in water and alcohol before being influenced by the light, but when it has become violet it is perfectly insoluble. "If you wish to have a fine colour," says Pliny, " for 50 lbs. of wool mix 200 lbs. of buccin to 111 lbs. of purple, in this way this superb amethystine colour is obtained." Purple also meant a red colour. "During my youth," says Cornelius Nepos, as reported by Pliny, " violet purple was the fashion, and was sold at 100 denari the lb. (about 237 francs the kilo) soon after the red purple of Tarentum, and afterwards the dibapha of Tjre, of which the lb. cost more than 1000 denari." Pliny distinguishes between purple and " conchylienne " colour, " for each the matter is the same only for the latter 'buccin' is not used, also they pour into the decoction of purple urine and water in equal parts, and then add half as much purple again. It is by means of an incomplete saturation that this delicate colour so much praised is obtained, so much the brighter as the wool has taken 1 ess dye".

Paillot de Montabert says that he found the use, of egg yolk only gave much force and unctuousness to the colours; that the mixture of the white with the yolk gave tenacity and greater transparency; while the white used alone gave still stronger and more lively colours, but less soft and more friable. Resins may be added to the yolk, rendering it still more tenacious, harder, and therefore less permeable, and the volatile oil of wax can also be added to this gluten to delay its drying. He says " the process is very rich and splendid, and as powerful as oil." Yolk of egg can be bleached by exposure to light when thinned with spirits of wine. White of egg may be kept in a dry state to be liquefied as required.

He also gives very full directions for painting in distemper. The palette should be of tin, with a piece of leather round the hole through which the thumb goes, to prevent it from galling the flesh. The other end should have a raised edge and several hollows to hold the colours, which should be ground in water only. The size must be kept liquid by being put over some warm coals. "If it gets solid on the palette it may be melted by warming. Take care that the colours do not get dry at the same time, and thin the size if necessary." At the end of each day the palette should be washed and dried to preserve it from rust. When working on a large scale the tints must be mixed in pots, and the palette becomes much too large to handle; it then stands on a box by the painter. The most important point is the keeping the surf ape of the cloth, wood, or wall on which the painting is being executed damp. Walls must be perfectly dry in their structure or the colours will fade in drying, and they will crack. If the wall is prepared with lime, lime must be used in the painting, which then becomes fresco secco. If it is intended to use chalk whites the wall must be prepared with chalk or plaster white in very thin coats. The plaster must be as good and even as possible; when thoroughly dry give it one or two coats of very hot size, stronger than that which is to be used with the colours. If it is rather uneven, mix a little whiting or chalk with the size to make it more even, or even plaster which has been exposed to the air and well ground. When this is dry the surface is to be scraped as uniformly as possible. The design is then drawn on the surface with charcoal quite lightly, and when the arrangement is settled the outline is gone over with a small brush and very pale colour. The colours should be tried on plaster squares, or on panels prepared in the same way as the wall, or on a warm tile, chalk, or umber, all of which absorb the moisture immediately and show what the colours will be like when dry at once. Tints which look the same when wet are often quite different when dry. Keep the pots of colour warm, and stir them up each time you put the brush in them, as the colours are apt to sink to the bottom. Use the colours strong and bright, remembering that they lose half their strength and more in drying. So that one need never fear heaviness, but rather weakness and too great greyness. Remember, too, that burnt earths change less than other colours, and that lake and black change scarcely at all. The brushes should be very soft— those made of goats' hair are good for uniting the colours.



By Domenico Ghirlandajo, in the Church of Ognissanti, Florence.