16 Another bottle from the Bloch Collection, also in sapphire-blue glass is not marked, but can be dated to the first phase of Imperial glass production ( fig. 3 ). It is of very similar form, but in place of engraved decoration are four raised beads defining panels and, on the two main sides, acting as integral snuff dishes as well. Here the crizzling seems entirely due to the chemical imbalance of the ingredients, the usual problem inherited from Europe at the time, rather than from burial. The surface is crizzled but has none Fig. 4. Colorless glass bottle with extensive crizzling, early phase production. of the characteristic iridescence or surface weathering typical of buried glass. Both seem to suggest that a standard form for early Palace snuff bottles, although many others existed, was a compressed sphere, with panels that could also act as snuff dishes if required. On the Kangxi-marked example, although the narrow-side panels are not raised, they are suggested by the use of oval designs. Other bottles exist that can be reasonably dated to the earlier phase of glassmaking from the inception of the Imperial glassworks into the Yongzheng period, even if we cannot always completely rule out manufacture at a slightly later time. There is a colorless glass bottle of the type known, somewhat confusingly, as “ crystal ” which has extensive crizzling ( fig. 4 ). This sort of lead glass excited Chinese attention in the early Qing period because of its clarity and capacity for holding hot liquids. It was much emulated in China and with the need for crystal- clear lenses at the court was certainly produced there in the early years. Another colorless glass bottle which is even more likely to date from the Kangxi period has extensive interior and exterior crizzling typical of Kangxi transparent glass and is unquestionably the result of the so-called “ glass disease ” arising from an unbalanced recipe ( fig. 5 ). Faceting was another glass- decorating technique introduced by Stumpf and his contemporaries in the dying years of the seventeenth century. It became a popular court decorative technique, and not only for glass, but must have begun in the Kangxi period, and a bottle such as figure 6 might be so early. Another obvious candidate is a colorless glass double-gourd shaped bottle which is not only diamond-point engraved but seems to suggest the hand of a European engraver ( fig. 7 ). It would be unusual for a Chinese artist to use two bands of what is essentially a border pattern to fill the main decorative area of the upper bulb. The floral band of the lower bulb is also distinctly un-Chinese in style. In place of the elegant, flowing, continuous formalized floral scroll the Chinese had been producing for a thousand years, there appears a disjointed design where the elements of the floral scroll keep getting lost beyond the frame. This appears to be by a European hand, possibly even that of Stumpf who is recorded as having mastered the arts of producing the glass, manipulating it on the blow- iron and carving it. What of the other colors Wang Shizhen informs us were being made prior to 1702? Red, purple, yellow, white, black, green and brown are all specified. We know that ruby glass was produced as a staple at court, made to a recipe introduced from Europe, and this may be the red—although the description “ red like fire ” raises the intriguing alternative that he might have been referring to realgar glass, which often looks remarkably like fire. However, we will go with the usual interpretation and assume he refers to ruby-red glass. It seems likely that this color, the result of a tiny amount of colloidal gold in the mixture, was a fairly closely guarded court secret for some time after its introduction. It was, apparently, not produced at Boshan, for instance, until the early nineteenth century, despite the regular comings and goings of Fig. 3. Sapphire-blue glass bottle carved with a circular panel on each side, 1696–1730. Bloch Collection.