The manufacture of true kaolinic porcelain was begun in 1769, but the soft paste continued to be made for another thirty years, side by side with the new ware. It was not till the year 1804 that it was finally abandoned by Brongniart, the new director. He found the soft-paste ware unsuitable for the big pieces now ordered by the Imperial Government. The paste was difficult to work, the preparation was expensive, and the dust formed both from the paste and from the lead glaze was injurious to the health of the workmen. One or two attempts have since been made at Sèvres to revive the old ware, but they have fallen through in every case. Brongniart, in 1804, to provide funds for the impoverished works and to pay the arrears of wages to the workmen, threw on the market the large stock of plain white soft paste that had accumulated in the magazine. Now at that time there were in Paris many skilled porcelain painters, some of them ex-employés at Sèvres, and others, men who made a living by painting on the plain ware sent from Limoges and other factories. These ‘chambrelans’ (they painted at home, _en chambre_, and corresponded to our English ‘chamberers’) were now employed by the dealers who had eagerly bought up the ware that Brongniart had parted with.[193] They painted and gilt this white ware in imitation of the Sèvres porcelain of the best period so successfully that the services they turned out have found their way into royal collections. This ware, in fact, forms a group by itself, quite apart from the later imitations of the _pâte tendre_, which, in every degree of merit and demerit, are now found in the china-shops of Europe and America. M. Garnier points out three signs by which this pseudo-Sèvres may be recognised: 1. The green prepared from the newly introduced chromium is of a warm yellowish tint, and displays none of the submetallic tints so often to be seen in enamels coloured by copper, as in the _famille verte_ of China. 2. The gold on this bastard ware, burnished with an agate polisher, differs in quality of surface from the old gilding worked _au clou_. 3. The date-marks and painters’ monograms were copied at hazard from the old pieces--at that time no list of these marks had been made public--so that, for example, the monogram of a gilder may be found on a piece decorated in colours only. CHAPTER XVIII THE HARD-PASTE PORCELAIN OF SÈVRES AND PARIS