There is, however, a class of porcelain which does not fall well into any of the above divisions, but which is historically of great importance. The blue, purple, and yellow glazes of this ware were _painted_ on the biscuit after a preliminary baking of the paste, and then fired, not in the hottest part of the furnace, but in what we may call the _demi grand feu_. The glaze of this ware contains lead, and this fact and the method of the decoration may be held to give it a position bridging over the interval between our first two classes and the third--that of enamelled porcelain. This ware, _painted on the biscuit_, dates, however, from an earlier time than the latter class, and must not be confused with it. As I have pointed out, these types did not entirely replace one another, for the earlier forms continued to be made by the side of the later. One of our principal difficulties in discussing the early wares of China is to reconcile and co-ordinate the various types described in old Chinese books with the few specimens surviving at the present day. Of these scanty examples we can point to scarcely any in public collections; the rare pieces that have been brought from China are in the hands of private collectors in England, France, and America. In the Chinese authorities we find as early as the tenth century references to porcelain which was ‘blue as the sky, brilliant as a mirror, thin as paper, and as sonorous as a piece of jade’; an emperor who reigned just before the accession of the Sung dynasty (960 A.D.) demanded that the porcelain made for him should be ‘of the azure tint of the sky after rain, as it appears in the interval between the clouds.’ Compare with these descriptions the thick paste, barely translucent, the heavy irregular glaze, greyish white to celadon or pale blue, of the few specimens of undoubted antiquity that have survived to our day. How can we reconcile the tradition with the material evidence? Two explanations have been given of the discrepancy. According to one theory, all the more delicate and fragile pieces have disappeared ‘under the hands of time’ (or shall we say more definitely under those of endless generations of housemaids?), only the heavy, solid specimens surviving. The other theory is simpler: it is that the writers of the books are apt to fall into exaggeration when speaking of any matter that has the sanction of age--that, not to mince matters, they are as a class great liars; and this is a point of view that commends itself to those who have any acquaintance with Chinese literature.[28]