_There is no question, therefore, but of soft-paste porcelain._ It is thus that Hellot sums up his report, written at the critical period when it was proposed to remove the Vincennes works to Sèvres, and place them under more immediate royal protection, and for this verdict we have every reason to be thankful. It is from this same memoir, _Recueil de tous les procédés de la Porcelaine de la Manufacture Royale de Vincennes_, that we obtain the most accurate details of the composition of the soft paste made at this time. It was a strictly private document, written expressly for the king by Hellot, who had recently been appointed to the direction of the Vincennes factory. This report was unearthed some time ago from among the archives at Sèvres. According to Hellot, writing in 1753, just as the Chinese combine the more fusible petuntse with the kaolin--‘a kind of talc which neither calcines nor vitrifies’--so with our frit, an artificial petuntse, we must mix, not an unctuous fusible clay, but some fine white infusible substance. Such a material is found in certain _marnes_ obtained from the gypsum quarries near Paris. The frit employed at Vincennes at this time--and the composition seems to have varied little up to the last days of soft paste in France--was essentially an alkaline silicate, containing also some lime and alumina, as will be seen from the following recipe:-- Fused nitre, 22 per cent. Sea salt, 7 ” Alicante soda, 3·7 ” Rock alum, 3·7 ” Montmartre gypsum, 3·7 ” Fontainebleau sand, 60 ” These ingredients, some of which are soluble in water, are fritted together--that is to say, imperfectly fused--in a part of the kiln specially reserved for them, great precautions being taken to regulate the heat. After reducing the frit to powder, the superfluous salts had to be thoroughly washed out by means of boiling water, before the substance was fit for mixing with the ‘body’ constituent of the paste. This body is prepared from the _grosse marne_ found at Argenteuil, by careful sifting and decantation. Six parts of the prepared frit are mixed with one part of the washed marl and with one part of a kind of chalk called _blanc d’Espagne_ (this last substance was afterwards dispensed with), and the whole thoroughly united by a grinding process which lasted for nine days. The resulting paste was made up into balls and allowed to ‘ferment’ for seven or eight months.