The tenor of these criticisms will not be endorsed by the modern collector of K’ang Hsi porcelain. _Famille verte_ porcelain is enthusiastically sought, and even indifferent specimens command a high price, while the really choice examples can only be purchased by the wealthy. As to the inequalities on the surface, the second of the three reasons hazarded by d’Entrecolles is nearest the truth. The enamels used by the Chinese porcelain painter contain a remarkably small percentage of colouring oxide, and one of the characteristics of _famille verte_ colours is their transparency. To obtain full tones and the contrast between light and shade (even to the limited extent to which the Chinese use this convention) it was necessary to pile up the layers of colour at the risk of unduly thickening the enamel. But the connoisseur of to-day finds nothing amiss in these jewel-like incrustations of colour, so long as the enamels are pure and bright, and have not scaled off or suffered too severely from the wear to which their prominent surface is exposed. It seems[314] that when the porcelain was destined to receive on-glaze enamels (without any underglaze blue) a special glazing mixture was used in which only one part of the softening element[315] was combined with thirteen of the ordinary glazing fluid. This glaze was very white and strong, and too opaque to do justice to an underglaze blue. There is a reference in the first letter of Père d’Entrecolles to a white colour which was used on the “porcelain painted in various colours.” It was fluxed with lead like the other enamel colours, and it was also used mixed with the latter to modify their tint. In fact there can be little doubt that it was arsenical white, an opaque white familiar on the Yung Chêng and Ch’ien Lung porcelains, and prominent in the _famille rose_ palette, but not usually suspected of such an early appearance as 1712, the date of the letter in question. The designs of the _famille verte_ porcelain, like those on the blue and white, are first traced in outline and then filled in with washes of colour. The outlines are in a dry dull pigment of red or brown black tint, inconspicuous in itself, but acquiring prominence when covered with transparent enamel. M. Grandidier tried to formulate certain rules for these outlines which, if reliable, would simplify greatly the task of dating the porcelains. On Ming ware, he said, the outlines were blue; on K’ang Hsi wares the face and body outlines were red, those of the vestments and other objects black. Unfortunately the first of these generalisations is wholly wrong, and the second pointless, because only partly right.