Of the recipes given by the early treatises, I will quote a few, for in reality they are all the literature we have upon the subject. Eraclius, who wrote in the twelfth century, gives accurate directions: "Take ochre and distemper it with water, and let it dry. In the meanwhile make glue with vellum, and whip some white of egg. Then mix the glue and the white of egg, and grind the ochre, which by this time is well dried, upon a marble slab; and lay it on the parchment with a paint brush;... then apply the gold, and let it remain so, without pressing it with the stone. When it is dry, burnish it well with a tooth. This," continues Eraclius naively, "is what I have learned by experiment, and have frequently proved, and you may safely believe me that I shall have told you the truth." This assurance of good faith suggests that possibly it was a habit of illuminators to be chary of information, guarding their own discoveries carefully, and only giving out partial directions to others of their craft. In the Bolognese Manuscript, one is directed to make a simple size from incense, white gum, and sugar candy, distempering it with wine; and in another place, to use the white of egg, whipped with the milk of the fig tree and powdered gum Arabic. Armenian Bole is a favourite ingredient. Gum and rose water are also prescribed, and again, gesso, white of egg, and honey. All of these recipes sound convincing, but if one tries them to-day, one has the doubtful pleasure of seeing the carefully laid gold leaf slide off as soon as the whole mixture is quite dry. Especially improbable is the recipe given in the Brussels Manuscript: "You lay on gold with well gummed water alone, and this is very good for gilding on parchment. You may also use fresh white of egg or fig juice alone in the same manner." Theophilus does not devote much time or space to the art of illuminating, for, as he is a builder of everything from church organs to chalices, glass windows, and even to frescoed walls, we must not expect too much information on minor details. He does not seem to direct the use of gold leaf at all, but of finely ground gold, which shall be applied with its size in the form of a paste, to be burnished later. He says (after directing that the gold dust shall be placed in a shell): "Take pure minium and add to it a third part of cinnibar, grinding it upon a stone with water. Which, being carefully ground, beat up the clear white of an egg, in summer with water, in winter without water," and this is to be used as a slightly raised bed for the gold. "Then," he continues, "place a little pot of glue on the fire, and when it is liquefied, pour it into the shell of gold and wash it with it." This is to be painted on to the gesso ground just mentioned, and when quite dry, burnished with an agate. This recipe is more like the modern Florentine method of gilding in illumination.