The colour of ultramarine is brought out by successive heatings. Green portions, more or less in quantity, are often formed in the crucibles, especially on the first ignition. On repeated heating they pass into a blue tint. Artificial ultramarines are said to be seldom entirely freed from all traces of the green modification, and are therefore less beautiful than the natural varieties, having a shade of green or grey. This defect, however, is certainly not discernible in Guimet's products, which sometimes incline so much to purple as to require neutralizing with a little Prussian blue. Depth for depth, the artificial are darker and less azure than the natural varieties, but the superiority of the latter consists not so much in their greater purity of hue, although this is considerable, as in their far greater transparency. The finest French ultramarine is never so transparent as the native; it is brilliant, it is powerful, it is permanent, it is nearly--but only nearly--transparent. Possessing in a subdued degree the characteristics and qualities of the genuine, it works, washes, and dries well; and is useful either in figures, draperies, or landscape. Rivalling in depth, although not equalling in colour, the pure azure of native ultramarine, it answers to the same acid tests, but is sometimes distinguished therefrom by the effervescence which ensues on the addition of an acid. Not a bubble escapes in such case from the natural blue; unless, indeed, as occasionally happens, it retain a portion of alkali, with which it may have been combined in the preparation, but from which it should have been freed. Darkened as a rule by fire, factitious ultramarine becomes dingy blue, and at last white, when strongly ignited for a long time; and is, like the true variety, decolourised by ignition in an atmosphere of hydrogen gas. At a high temperature, this effect is even produced by silica, whence the unfitness of ultramarine for painting on glass or porcelain; and simply by a prolonged red heat the blue is rendered white. Being unaffected by alkalis, it is eligible in mural decoration, and is particularly adapted to siliceous painting, on account of the silica and alumina which it contains, two substances with which a soluble silicate readily unites. If artificial ultramarine be mixed with a soluble silicate, for example silicate of potash, and be laid on a properly prepared ground, it will become so firmly fixed, says Mr. Barff, that no amount of washing nor the slow action of moisture will remove it, or affect its brilliancy. Judging from the behaviour of ultramarine, therefore, if the colours employed in siliceous painting contain silica and alumina, they should adhere as firmly to the surface on which they are placed; and this is really the case. It is possible to produce a mixed solution of aluminate and silicate of potash which will remain liquid for twenty-four hours. If, while in the liquid state,