The Marakovic Collection
Hugh Moss and Stuart Sargent


Part I: The Imperial Phase
The Yongzheng Era, 1723 – 1736

With the advent of the relatively short Yongzheng era in 1723, we step onto firmer ground. Reliable reign-marked wares remain rare and are confined to certain art forms, but we have a treasure trove of information in the archives of the palace workshops, the Zaoban huoji chu 造辦活計處. (Many readers will recognise the workshops by the shorter name Zaobanchu 造辦處, the name adopted in 1759 when the administrative structure was formalised.) Abbreviated, enigmatic, confusing, these archives are a compilation of short-hand administrative documents keeping track of a wide range of products made at the court for the purposes of the Imperial Household Department – everything from saddles and candles to high art. Detailed records apparently began only with the first year of the Yongzheng era, although a Yongzheng entry notes one Kangxi event: thirteen glass snuff bottles were dropped off for temporary storage the next-to-the-last day of the Kangxi era (which by the Western calendar was 3 February 1723).

We can glean other glimpses of Kangxi production, as well. In the early days of the Yongzheng reign, we see numerous snuff bottles in a variety of materials being brought to the miscellaneous workshop, the inlay workshop, and the enamelling workshop to serve as models for further production; these snuff bottles must have been leftovers from the just-ended Kangxi era.

It is the enamelling workshops of the new reign that provide us with our most plentiful and reliable landmarks. The records are full of orders for both glass and metal snuff bottles decorated with painted enamels; and we can match some of these up with extant examples, which is of course the Holy Grail of archival research.

Extant examples of Yongzheng painted enamels on metal are almost as rare as those surviving from the Kangxi reign, but there are two in the Marakovic collection. One of these (fig. I.22) is of great interest from the standpoint of using the archival records. It is almost certainly related to a Kangxi-marked bottle that recently left the Bloch collection (Bonhams, Hong Kong, 28 November 2011, lot 166), a famous enamelled metal bottle with prunus on a ruby-pink ground (fully discussed in Hugh Moss, Victor Graham, and Kao Bo Tsang, A Treasury of Chinese Snuff Bottles, vi: Arts of the Fire,no. 1066). The Bloch bottle fits the description of a snuff bottle mentioned in the third year of the Yongzheng records: In mid 1725 (YZ3/5/19), an ‘enamelled red-ground white prunus blossom snuff bottle’ was delivered to the enamelling workshop with instructions from the emperor to ‘make a picture based on this snuff bottle, and then fabricate a bottle just like the way it is drawn’. If the model were from the Yongzheng period, we may assume that the instruction would simply have been to make some more of the type made on such and such a date; the need to take a drawing and copy it suggests that this is the first of this series to be made in the Yongzheng and that the model was from the previous era. The bottle was not returned to the palace for nearly four years (KZ7/4/2), so the enamellers may have been leery of depending on a drawing alone in their efforts to replicate it.

It can be difficult to keep track of quantities of certain types made at the palace workshops. In the case of white prunus designs on a red or ruby ground there are a dozen or so entries for such bottles, but some are repeats, and on one occasion the order is for ten to be made with black grounds and red grounds without specifying how many of each; when the order is noted as complete, only two pairs are delivered, and we don’t know what colour ground they had. There is also one case where two bottles were submitted and criticised, after which an order was received to cover them and re-paint them with larger blossoms, so presumably the original design was painted over. It seems, however, that about a dozen on a red ground were produced and delivered by the enamelling workshop from 1727 to the end of the era. Very few have survived, but the Marakovic example, bearing a Yongzheng mark, is one of them.

(In 1987 Yang Boda 楊伯達 caused some confusion by publishing a bottle with a similar design and stating that it came from Guangzhou, but no such bottle is mentioned in the voluminous tribute records he was drawing on for his book, and he later indicated in private correspondence that his statement was based on his own sense of the development of the art, not on those records. Based on positive evidence from the archives for the palace workshops in Beijing, we remain confident that prunus designs on pink or red grounds were northern, not from Guangzhou.)

The real prize of the Yongzheng bottles in the Marakovic collection is shown in figure I.23. It is of a unique shape and design for a Yongzheng enamel, but what really crowns it, literally as well as figuratively, is the highly imperial stopper, made originally for this very bottle. The emperor’s most elaborate hat for ceremonial occasions was of similar shape, and the pearl finial was reserved for the use of only the loftiest members of the imperial family. The elaborate gem-encrusted stopper with its pearl finial is an undoubted reference to the emperor’s hat, and we can safely assume that this bottle was made for either his use or that of his empress, if she also took snuff (and the indications are that women took snuff as readily as men, perhaps from the outset).

The slightly protruding lip, extending beyond the plane of the enamel on either side of the thin metal body of the bottle, gives greater strength to this vulnerable part of the bottle and protects the enamel at the weak point where it meets the exposed metalwork. This bottle and the previous one have what we may call a ‘naturally wide’ mouth; the lip is approximately equal to the thickness of the materials forming the lip, which in the case of metal and enamelled-metal bottles makes for a narrow lip and a wide mouth. (Later enamelled-metal bottles often have a flattened ring or ‘washer’ welded on atop the neck, creating a wide lip and reduced mouth.) The naturally wide mouth may be a distinguishing characteristic only of enamelled-metal snuff bottles from the early period of palace production, but it may also prove a useful trait in helping us to identify other early types.

The conch shell and fylfot seen in figure I.23 are readily recognised as Buddhist designs, the conch representing the power of the Buddha’s teachings and the precious swastika being a sign on the Buddha’s breast representing the Buddha-mind in everyone. From pre-conquest times, the Manchu rulers had patronized Tibetan and Mongolian Buddhism, always attempting to co-opt as much religious influence as possible; by the Qianlong reign, the idea that the Manchu ruler was a universal Buddhist king invested with moral authority over all cultures was being openly and aggressively promoted. Not surprisingly, the range of Buddhist works of art from the reigns of the Yongzheng and Qianlong emperors made at and for court use is immense.

There are a few references to painted enamel on glass in the Yongzheng records. Two entries from 1728 sound like one of them could be the famous bamboo-shaped bottle decorated with a bamboo and spiders design in the National Palace Museum in Taiwan, which bears a Yongzheng mark. The archival records refer to the subject matter of an enamelled-glass snuff bottle as jiejie shuangxi 節節雙喜; jiejie means both ‘every joint of the bamboo’ and ‘time after time’; shuangxi means ‘doubled happiness’, and this is often symbolized by a pair of spiders, because seeing a spider has long been taken as an auspicious omen in China. The painted decoration of the NPM bottle includes a pair of spiders, and it is generally accepted that it is one of the very bottles referred to in the archives. Thus, we have both archival and material evidence for the existence of enamelled glass bottles in the Yongzheng period.

There is some evidence, admittedly slim, that one enamelled glass bottle in the Marakovic collection might date from the latter part of the Yongzheng reign (fig. I.24). In mid 1736, the first year of the Qianlong reign (QL1/5/17), the records note that an imperial order was received at the enamel workshop to fire enamelled glass for the emperor’s inspection. Three days later, two transparent blue glass snuff bottles with painted enamels were presented to the emperor. A response was received noting that there were not many flowers on the snuff bottle and that next time more should be painted, and that all should have a reign mark. Here, right at the beginning of the Qianlong reign, we have a valuable indication of the existence of two sparsely enamelled blue glass snuff bottles, neither with a reign mark.

Enamelled blue glass is extremely rare prior to the Guyue xuan 古月軒 wares of the late Qianlong, when one or two older plain blue glass bottles were enamelled. We are aware of only one surviving enamelled blue glass bottle that is obviously early enough to relate to the record, and it fits the description given in the records. The bottle (, no. 00677) is decorated with fruiting and flowering pomegranate branches.

Now, that bottle is similar in palette and style to the Marakovic example in figure I.24, although the enamels look a little more degraded from the passage of time on the blue-ground example. The Marakovic bottle is also relatively sparsely decorated, suggesting a Yongzheng trend that was reflexively repeated in the early years of the Qianlong era until the emperor expressed his preference for denser floral designs, which then become a standard for the rest of the reign.

Another reason for proposing a Yongzheng date for the Marakovic bottle is that it is unmarked. From the beginning of Qianlong and for the rest of the reign, enamelled glass and metal wares made at the palace workshops were predominantly marked, either with a four-character Qianlong mark or, after 1767, sometimes with an alternative Guyue xuan mark, which we explore later. The record of 1736 is very clear that in responding to the two blue glass enamelled bottles, the emperor’s wish was that in future ‘all should have reign marks’, which implies that marks were not routine previously.

In the Kangxi period we have no indication of production elsewhere for the court, although we may assume it began not long after snuff bottles became popular. The moment snuff and a passion for fancy containers to hold it began to flourish at court, the fine workmanship and exotica of outside workshops would quickly have been sought beyond the palace walls. The first solid evidence of it, however, is from the Yongzheng period, when a series of reign-marked Guangzhou enamel bottles of imperial designs were sent to the court, either as tribute or in fulfilment of imperial orders. Of all surviving marked enamels from the Yongzheng era (there are not many), more than half are southern. Figure I.25 shows a spectacular example, one that is typical of the finest Yongzheng southern production for the emperor.

The art of painting with enamels on metal was introduced simultaneously to Guangzhou and the palace workshops. Enamellers from Europe who brought the art and the materials with them to introduce to the court first set foot in the Celestial Kingdom at Guangzhou, where they would present their credentials and be housed until summoned north. This process could take months, sometimes years; the frustrated experts spent their time showing their skills to local mandarins in order to impress them and hasten their journey northward so they could fulfil their underlying mission of proselytizing the Chinese at the highest level. The art of painted enamels on metal was already well established at Guangzhou a decade before the end of the Kangxi period, and by the Yongzheng it was under spectacular control – often better control than they could manage on a regular basis at the court. The enamels were clear and perfectly fired, surfaces of great integrity were achieved, and fine painting was mastered, albeit of local style even when reproducing northern designs. The Marakovic example in figure I.25 is exemplary. It also features the wide mouth and narrow lip of its northern counterparts, characteristics that remained typical in Guangzhou well into the Qianlong reign. The inside of the foot rim is filled with enamel almost to the level of the rim; in effect, we have a protruding foot with a slightly concave centre. The concavity is just enough for the bottle to stand on the foot rim and for the reign mark to be protected from wear. This compressed, elongated-oval form is also a typical and popular southern enamel shape from the Yongzheng reign through to the end of the Qianlong. Absent the reign mark, we would still recognize this as an imperial product by its yellow ground, which was reserved for imperial use, and by the formalized floral design with its generalized auspicious motifs of longevity and good fortune.

Porcelain snuff bottles do not appear to have been enamelled at court during the Yongzheng reign. There are copious records of blank porcelain arriving from the porcelain centre at Jingdezhen and being stored against future use and records of the enamelling of bowls, dishes, and so forth, but there is not a single mention of a snuff bottle. It seems likely that, with enamels on glass and metal being produced at the court, it hadn’t yet occurred to anyone to bother exploring porcelain as an option, at least as part of the standard output. Moreover, the translucence and resonance of porcelain, so prized for bowls and dishes, were wasted on a small vessel packed with snuff. Only one Yongzheng reign-marked enamelled porcelain snuff bottle exists – and it is southern, though it copies a Beijing shape and decorative style. One or two unmarked monochrome porcelain snuff bottles might date from the Yongzheng, but they are equally likely to belong to the Qianlong era. Regular porcelain snuff-bottle production does not seem to have begun on a noticeable scale until the early Qianlong period, when Tang Ying 唐英 (1682 – 1756), supervisor of the government kilns at Jingdezhen, was instructed to make ‘no more than’ fifty snuff bottles a year of enamelled porcelain – a modest quota.

With the imperial glassworks well established after a quarter of a century of production, and with demand expanding rapidly, not only for fancy bottles but also for convenient, easily produced containers for gifts of snuff, we may expect large numbers of imperial glass bottles from the Yongzheng reign. Hundreds appear in the records, and many undoubtedly survive, but identifying them remains problematic. Three ruby-red glass examples of typical imperial octagonal faceted forms exist with credible Yongzheng marks; none is significantly crizzled or typical of very early transparent glass production, but that is not enough to cast doubt on the marks, since such problems were largely solved for the majority of Yongzheng glass.

Many surviving unmarked glass bottles are of a style that could date from the Yongzheng reign, and it is reasonable to expect that at least some of them do. There are many faceted ruby-red glass examples of various shapes and faceting designs that do not carry reign marks. Many other popular colours are represented also, but red seems to have been among the most popular colours at court; it had novelty value and exotic foreign appeal, since the ability to make ruby-red glass had been developed in Europe in the late seventeenth century and had been introduced by the Jesuits only in the previous reign. The process of making this colour from colloidal gold was first hinted at by Florentine chemist Antonio Neri (1576 – 1614) in his 1612 book L’Arte Vetraria, but it was not until a little before 1679 that Johann Kunckel (1630? – 1703) perfected it. A hot new colour in Europe, it would have created a sensation when introduced to the imperial glassworks in or shortly after 1696. Snuff bottles made from red glass would soon have become one of the most sought-after gifts, since the material would have been so readily recognized as imperial even before taking the form into account. Snuff bottles made with some of the other colours were probably more dependent on having a standard imperial form to convey the prestige they were supposed to carry.

Figure I.26 shows a bottle whose form is similar to two of the ruby-red glass bottles with Yongzheng reign marks mentioned above. It could date from any time between the Yongzheng and the early decades of the nineteenth century, and its narrow mouth may be one indication of a Qianlong or later date, but it serves to illustrate a type that would have been produced in the Yongzheng era.

In addition to ruby red, all the known Kangxi colours must have been used by Yongzheng glassmakers, almost certainly augmented by one or two more, and there must have been experiments in combining them in decorative schemes.

One of the most important colours was, of course, imperial yellow; it would have been produced at the glassworks throughout the dynasty. The term imperial yellow as used in the snuff-bottle world actually covers a range from a pale lemon yellow to a rich, egg-yolk yellow. Figure I.27 shows one example. This bottle has a wider mouth than the bottle in figure I.26 – not in absolute terms, but relative to the lip, which is the significant measurement. With the bottle in figure I.26, there seems to be a deliberate attempt at a narrower mouth, whereas the one in figure I.27 represents simply a practical size.

We may be certain that by the Yongzheng era, and almost certainly during the Kangxi, a good many variations on faceted forms would have been explored, going far beyond the octagonal profile with its various panel shapes. This is one of those variations. It suggests a bottle contained in a protective pouch of some sort, but the pouch is composed of eight facets. This artistic concept is known from the Qianlong, but it is worth stressing yet again that just because an idea, form, or type existed in the Qianlong era it does not follow that it was invented then. Common sense, and indeed the records, suggest otherwise. For example, Chinese snuff bottles set with European watch faces are usually dated to the late Qianlong era, which is probably generally correct, but the Yongzheng records also include snuff bottles that appear to be of the same type.

Figure I.28 shows a colour we know was made at the glassworks from the Kangxi era onwards, and the bottle exhibits another variation on faceting. Once the art of faceting was introduced, it would quickly have occurred to the novelty-hungry court with its rapidly growing demand for snuff bottles to start faceting any existing shapes they admired. This is a faceted meiping or ‘prunus-blossom vase’.

The bottle in figure I.29 has a much lower centre of gravity. It is also of an imperial colour known from the outset, although its un-crizzled state suggests it is unlikely to be earlier than the Yongzheng.

Figure I.30 shows a bottle that offers an intriguing innovation: each facet is concave. Of the standard sapphire-blue colour popular at court, it has a simple protruding flat foot without a foot rim. We tend to see the less complex detailing characteristic of later work as a sign of a diminished sense of commitment, but it could also be a function of demand and purpose. Individual bottles ordered for the emperor and his family or for distribution to high officials and important individuals would have been made to a very high standard – the highest possible. On the other hand, orders for large numbers of relatively plain glass bottles to be used as fancy containers for gifts of snuff on occasions such as banquets for the emperor’s birthday or for prizes to large numbers of bannermen on regular hunting trips north of the Great Wall would not have required the same levels of care and attention to detail. (Banners were the military units that formed the foundation of the Qing state; there were Manchu, Mongol, Chinese-martial and other banners.) Many early bottles from tombs have only a simple protruding flat foot, yet they are very carefully made. This design probably facilitated the making of relatively large numbers of bottles; it does not equate to a decline in artistic standards resulting from lack of funds and interest.

Standard imperial faceted bottles, particularly those of an octagonal profile, span imperial snuff-bottle production from the late seventeenth century to the end of the dynasty, probably. Many of them have valuable, credible reign marks to guide us as to their evolution. This group was an early focus of the Marakovic collection, so they are well represented here. Another major selection is in the Barron collection, and one or two other collectors formed important groups of them when they were coming so bountifully onto the market recently with the archeologically unsupervised excavation of thousands of private tombs around Beijing – a flood of bottles that slowed to a trickle with surprising abruptness after all the Beijing Olympic venues were completed. Discounting the fakes that gushed forth to mingle with the newly excavated bottles, there is no doubt that a vast number of such bottles were buried with their owners as emblems of favoured social status. They were produced in far larger numbers than we had imagined, and in a much wider range of colours, many of which would have been in production by the Yongzheng era.

The faceted bottle in figure I.31 is of a pale emerald green; similar bottles are found from the eighteenth century in a much darker emerald green and variations between. Others are in a more lime-green colour. This bottle has the formal and surface integrity of an early example and a naturally wide mouth, as opposed to a self-consciously wide one. It could date from either the Yongzheng or the Qianlong era.

Among early examples we also find glass that is colourless (with or without crizzling and glass disease); we find white; various intensities of sapphire blue from pale to very deep, almost purplish blue; aquamarine blue, often an intense, deep colour resembling more closely the highly valued blue tourmaline used at the time; translucent cobalt blue; turquoise blue and turquoise green; ruby red, again from dark to almost pale pink; transparent brown; the yellowish brown known in the records, we believe, as ‘wine yellow’; emerald green; a range of lime greens; transparent olive green; translucent, sometimes opaque green with or without streaking, which apparently imitates (or at least resembles) jasper; purple; pink; a rare opaque or semi-translucent blood-red colour; lilac; and imperial yellow. We also find realgar glass, either of the predominantly orange-yellow range, or with a bright scarlet surface colour. Then there are the mixtures, where one colour bleeds into another – say, a pale purple or grape colour blending into colourless glass, the variation emphasized by the faceting.

Finally, there are the rarities with surface decoration of fragments of other glass, such as ruby pink splashes as raised panels on colourless glass, or, as in figure I.32, with fragments of aventurine glass—this is one of only two known examples, the other belonging until recently to the Bloch collection (Bonhams Hong Kong, 25 May 2011, lot 127).

Aventurine glass was invented in Venice and became an obsession with the early Qing glassmakers. At first it was imported in blocks to Guangzhou and thence to the palace workshops. Apparently it could not be melted without losing its character, so it was either used as fragments in blown wares or worked from solid blocks; it was not blown. When an entire bottle is of aventurine glass, it is always carved like a hardstone. This remains true even after the Chinese discovered the art of making aventurine glass successfully at the palace workshops in the 1740s.

Both the Bloch bottle, with faceted panels on either main side, and the Marakovic bottle in figure I.32, with rounded convex panels on either main side, were initially blown into moulds that included the panels, the facets on the sides, and the gutters around the perimeter of the panels. The fragments are not only stretched in places, suggesting the blowing process, they are also evenly spread over those different contours and up the neck; that would be impossible if a globular bottle were blown and then ground down to form the desired shape, for the grinding would remove the flecks from the surface of the glass in critical places.

A three-century-old portrait of a refined lady gives us unexpected assistance in identifying at least one bottle that can be no later than the Yongzheng reign but probably predates it. The portrait comes from a set of portraits of young women that was commissioned by the Yongzheng emperor before his accession to serve as a surrounding screen for his favourite apartments at the Yuanming yuan. In 1732 they were ordered remounted as 12 hanging scrolls to preserve them. They have been described as being portraits of the Yongzheng emperor’s twelve consorts, but James Cahill demonstrates that they are more likely to be discreetly seductive scenes in a rather racy, slightly decadent Yangzhou 揚州 style (James Cahill, Pictures for Use and Pleasure: Vernacular Painting in High Qing China, University of California Press, 2010, pp. 48 – 49). Quite apart from the appeal of their main subject, the extraordinarily detailed paintings include a mouth-wateringly exotic range of ancient and contemporary works of art in cabinets and on tables. Whether or not these are palace ladies, they are surrounded by objects similar to items with which the prince who commissioned the paintings would be at home. Beside one of the women is a ruby-red snuff bottle that can only have been made at the imperial glassworks (fig. I.33). It is of elongated form with a quatrefoil profile and raised rounded-rectangular panels on each main side. Cahill convincingly suggests that the paintings date from between 1709 and 1723, based on seals that appear on paintings within the paintings and match actual impressions of seals belonging to the future emperor. We do not know of any bottle that is exactly identical to the one in the painting, but there are many similar flattened, elongated forms in ruby-red glass with flat panels that might be of the same period. One example is the Marakovic bottle in figure I.34.

These raised panels are worth a short digression. Snuff is prone to forming lumps with the slightest introduction of dampness. Much of it travelled for months over vast distances, often from Brazil viaEurope, and then was stored in large glass containers, sometimes for years. When lumps formed, a convenient method of dealing with them prior to snuffing was to put a portion of snuff on a flat or concave surface and crush them out, using perhaps the spoon or the stopper; a cabochon of stone would be the perfect tool for the job; even a finial would serve adequately. Eventually snuffing paraphernalia came to include small dishes for preparing the snuff, with the general idea, it is claimed, that each snuff taker had a private dish and, when snuff was passed around, used it to crush an individual portion. Sometimes dishes and bottles came in matching sets. The earliest of these separate dishes that can be dated with any certainty is no earlier than the mid Qianlong era. The vast majority that are dateable by style or material are mid to late Qing. Some exist that might be earlier, in well patinated ivory or plain stone, but not a single one of those suggests that it must predate the mid Qing. Common sense has it that the separate dish evolved at this time. For a series of dishes in the Marakovic collection, see figures IV.1-9.

A large number of bottles, many of them early, have a plain panel on each main side that is flat, convex, or concave, sometimes surrounded by a raised rim. We have long held the view that these were intended as integral snuff dishes. An entirely different theory has been proposed recently: that the plain circular panels on these snuff bottles were not functional at all, in the normal sense; they were symbolic mirrors to ward off evil spirits. Despite the importance of mirrors in Chinese culture and the general auspicious symbolism of the circle, we remain unconvinced by this argument, for two main reasons. The first is that if one felt the need for such a talisman, one could attach a token mirror to the belt or wear one as a pendant (as was in fact done) —a symbolic mirror on a snuff bottle hidden away in a pouch would seem to present little threat to malevolent forces. Secondly, a large number of early bottles have these panels, but their appearance in late Qing production, after the separate snuff dish seems to have become popular, is greatly reduced. More separate dishes, fewer integral dishes: that suggests that the former were taking over the function of the latter, which further suggests that they both held snuff for one purpose or another. The mirror theory is a stimulating one, and there is no question but that the round panels are potentially symbolic (as well as aesthetically pleasing), but we have not seen enough evidence to dissuade us from continuing to view these panels as integral dishes.

The panels on the bottle in figure I.34 are, like many others, not circular, of course, further weakening the mirror theory. A typical flat-panelled example in imperial yellow is shown in figure I.35; it does have round main-side panels.

The relatively small size of this particular example prompts us to note that, although the standard-sized snuff bottle represented by some of the imperial enamels on metal of the Kangxi period (whether painted or inlaid) continued throughout the dynasty, smaller sizes seem to have been popular alongside them during the early decades of the eighteenth century. There were probably two reasons for this. The ever-growing production of bottles as containers for gifts during the eighteenth century as snuff took hold in wider and wider circles prompted relatively smaller sizes – to limit the amount of precious snuff distributed. The seriously small bottles we now call ‘miniatures’ that evolved by the Qianlong reign might have been a natural progression of this trend. A gift of snuff was all it took to grant imperial benediction; there was no need to dole it out by the shovel-full. Many standard-size bottles continued to be produced, however. The elite were not short of fine snuff and might prefer larger bottles for their daily needs. The emperor was given ample imported snuff from missionaries keen to gain his favour, so would have had no need to husband his supply.

Another way of interpreting the development of small sizes is that they were designed as a novelty or as containers for samples of rare or precious snuffs. Just flashing a miniature bottle might have been a subtle way of declaring that one was a connoisseur qualified to judge fine grades of snuff.

Figure I.36 also has the flat panel. It is in typically imperial ruby-red glass; it also has the naturally wide mouth of some early bottles. But one more feature identifies it as a likely early-eighteenth-century bottle (its condition and lack of crizzling suggest that it is unlikely to be from the very first phase of production). The design on the narrow sides is a formalized version of bamboo sections, an idea that probably occurred to palace glassworkers at some time during this period. Similar bottles have been excavated from tombs, some in dismal condition, which suggests they have been in the ground for a long time. Some snuff bottles of this form have engraved designs that appear to be early and under missionary influence in their style, which would add support to an early date for the design.

The bottle in figure I.7 was a possible Kangxi-era turquoise-blue snuff bottle; we now propose another turquoise-blue glass bottle, in figure I.37, as a possible Yongzheng product. In neither case are we brimming with confidence about the date, but we do want to go on record as recognizing that the colour would have been a staple at the time. The records note turquoise glass: on one occasion in 1732, towards the end of the reign, the emperor was so dissatisfied with a turquoise glass bottle because it was darker than the original material that he ordered that it be melted down and recycled. The bottle shown in figure I.37 is as fine a colour as one could hope for in terms of the gem material. Again, it is too plain to date accurately with any degree of confidence, but nothing in the colour, simple form, elegance, size, or size of mouth would rule out a Yongzheng date.

Before moving on to glass involving less guesswork, let us introduce an example of colourless glass in the Marakovic collection that is very rare (fig. I.38). It is unaffected by crizzling, so presumably is not from the earliest years of the imperial glassworks. The engraving, done with a spinning wheel, is of distinctly European style, placing the work in a small group of early engraved-glass bottles that suggest direct European influence, possibly even decoration by a European hand, although this particular one is less indicative of that than some where it is blatantly obvious. A bottle in the Bloch collection (A Treasury of Chinese Snuff Bottles, v: Glass, no. 864) also had a European touch, and we suggested a dating range of 1690 – 1720, but that now seems to be at least a couple of decades too early. A very similar specimen has come to light bearing a Qianlong mark; even if we interpreted that to mean that the bottle was made in the very first year of the era, 1736, that is still later than we would have once guessed these bottles to be. Dating early snuff bottles is an exercise that teaches humility.

If plain glass was produced in large quantities from shortly after the establishment of the glassworks in 1696, we can be certain that carved glass was as well, including overlay glass, as suggested by the recent discovery of a credibly marked Kangxi red glass overlay (Water, Pine and Stone Retreat Collection). This seems borne out by three instances in the Yongzheng records where overlay glass, some of which might have been from the Kangxi era, was brought into the workshop for one purpose or another. But there is only one record of overlay-glass articles actually being produced: a pair of overlay blue chi-feline (chi-tiger) snuff bottles (tao lan chihu biyanhu 套藍螭虎鼻煙壺), completed in the fourth year of the era (YZ4/2/10). Unfortunately, we cannot be sure whether the overlay was cut back as a cameo or was simply sutao 素套, a well-known type of early glass that has distinctively different interior and exterior colours but is not carved as a cameo. We are currently unable to recognize any extant examples of Yongzheng cameo overlay glass.

Without marks, we can do no more than propose possible carved-glass Yongzheng snuff bottles, such as the one in figure I.39. The size of the mouth relative to the narrow lip is interesting: it seems self-consciously wide rather than naturally wide as we see on enamels on metal from early imperial production. It may reflect an early attempt to match in glass the proportions we see in enamel-on-metal bottles, but we are not sure what to make of this; it may be a later product with the wide mouth related to the known mid-Qing fashion.

It must be remembered that different materials create different options. A very wide mouth and narrow lip on a delicate amber bottle wouldn’t last long, for example; the smaller the mouth and the wider the lip, the stronger the upper neck. A wide mouth on a metal bottle, on the other hand, would be less trouble to produce and perfectly practical. In the glass-working process, the important factor to keep in mind is that the initial form was blown into a mould, always leaving an uneven inner neck surface. It was necessary to work the interior of the neck to make it cylindrical to receive a stopper. The junction between the drilled cylindrical interior of the neck and the interior surface of the blown body of the bottle is commonly observable on a large number of glass snuff bottles of most types. A number of factors could influence how extensively the neck had to be bored out. An awkward blow leaving an irregular interior to the neck could have required the removal of quite a bit of material to produce a properly round inner neck. The width of a mouth, therefore, can be a matter of fashion in some cases and a consequence of remedial carving in other cases. We cannot always guess which explanation is best.

Another possible factor relating to mouth size is the requirements of the snuff. This is something we know very little about in the post-snuff world, but it is worth remembering that snuff-bottle style had to adjust to some extent when fashion in snuff itself changed. Snuff was the starting point; container characteristics such as material, opacity, mouth size, interior finish, and so forth could and did vary for aesthetic and economic reasons, but they all had to take into account what was best for the storage of snuff and the preservation or enhancement of its flavour. And snuff was not a homogenous commodity: there was imported snuff from Brazil (most prized), Spain, France, and Scotland (made with Chesapeake tobacco); there was domestic snuff from Shandong, Sichuan, Guangdong, Zhejiang, and Jiangsu. Each of these had different qualities and prices, and although it is likely that we shall never be able to match snuff types with bottle types with any precision, it is very likely that elite snuffers had a very good idea of what kind of mouth, interior, material, or decoration was appropriate for what kind of snuff.

With this particular bottle (fig. I.39) one more possible reason for a wide mouth occurs to us. After being blown in the imperial glassworks, the bottle might have been sent to have its neck drilled in the adjacent grinding and polishing shop, where simple carving, smoothing, and perhaps faceting was done. But for the carving of the design, it would have been sent around to the lapidary workshop where jade and other stones were carved. It is not out of the question that, having carved the surface so well, the lapidary may have decided to widen the neck – an easy enough process that could have been achieved in very little time, although we don’t know why the lapidary would have taken this extra step.

This design of kui 夔 dragons making up characters was a Qianlong standard in both plain and overlay glass and in other materials such as nephrite and Duanstone (the shale used for inkstones). As with so many types and designs, however, what was popular in the Qianlong era did not necessarily start in the Qianlong era. In 1731 (YZ9/4/19), the Yongzheng records refer to a large snuff bottle in ‘sky-after-rain’ blue glass decorated with two kui 夔 dragons being brought in to serve as a model, so we know that both carving on glass and kui were standard at the time.

The bottle in figure I.39 is not of the standard form that was to evolve for this design: it is smaller, more of a compressed sphere with a short neck, and has no foot. This could be evidence for a Yongzheng date. The standard eighteenth-century version has a foot, and the curve of the body tapers gently into the cylindrical neck.

We do not yet know when overlay glass of the cameo variety began. There is an intriguing case of two sets of twelve tea cups in ‘clear sky after rain’. One of these was submitted to court in 1709, according to Yongzheng archival records for late 1725 (YZ3/11/3). It was carved by a Cheng Xianggui 程向貴, from Guangdong, and it is said to be kehua 刻花, which is often supposed to denote overly carving. The other set, made sometime before its maker, another Guangdong man named Zhou Jun 周俊, returned with Cheng to Guangdong in May 1715, was not submitted to court until late in 1724 (YZ2/10/28). It was ‘plain’ overlay, sutao 素套, which surely designates uncarved overlay. Zhou Jun was an apprentice craftsman in the household of the man who presented both sets, Ding Zaobao 丁皂寶. Ding might have been a eunuch in the upper echelons of the palace bureaucracy; he was commissioned by the Yongzheng emperor to plant pines at Mount Tai in 1731 to replace trees destroyed by a slide. We know nothing else about him, which is a pity. In answering the 1725 query, the jade workshop was eventually able to give the dates when both sets of cups were offered to the court and by whom, but only Ding Zaobao knew where the two men from Guangdong had gone. The need for all this detective work indicates that the uncarved cups, at least, were made outside the official glassworks, evidently in a glass-carving workshop Ding Zaobao was running on the side. It makes sense that the uncarved overlay set was kept nearly a decade by Ding before he offered it to the court, considering that it was the work of an apprentice in Ding’s own household.

The story of the overlay tea cups proves that the technique of layering colours was known well before the end of the Kangxi era. What it does not tell us is whether kehua was cameo overlay or simply engraved designs on an overlay colour. We think it was the latter, for very few cups with contrasting cameo designs are known, and none of them predate the mid eighteenth century. In contrast, there are a number of cups and other wares from the Yongzheng and early Qianlong reigns that are decorated with surface engraving.

That said, if one cameo overlay bottle survives from the Kangxi era, we may expect the existence of glass-overlay snuff bottles dating from the Yongzheng even if we can’t identify them. A possible candidate is the spectacular Bloch realgar-glass carved bottle with a cameo design of chi 螭 dragons (A Treasury of Chinese Snuff Bottles, v: Glass,no. 972; Bonhams Hong Kong, 28 May 2010, lot 78). It might even be earlier than Yongzheng. The bottle in figure I.40, with its simple overlay, integral snuff dishes, and relatively wide mouth also might be as early as the Yongzheng era, possibly before, as it shares certain characteristics with the recently-discovered Kangxi-marked example. Alas, there is no way to prove that it is not Qianlong, and although it is probably imperial, there is no way to prove that, either.  Snuff bottles represented in the Yongzheng records include a relatively large number of embellished types – mixed media involving more than one material. Among them are snuff bottles of ‘foreign lacquer’ (probably inspired by Japanese makie-e 蒔繪 gold lacquer painted designs).

There may be references to bamboo veneer in the Yongzheng records, although they are ambiguous at best. If bamboo-veneer snuff bottles were produced among our earlier snuff bottles, then there are a few possible candidates for so early a date, including the bottle in figure I.41. It is one of the rarer forms and is softer, slightly less emphatic than some that we may assume to have been made in larger quantities during the Qianlong reign. Nevertheless, an early Qianlong date is as likely for this bottle as a Yongzheng one.

Whenever it began, the art of bamboo veneer was highly developed in the first thirty years of the Qianlong era. We deduce this from a bamboo-veneer ruyi sceptre that is inscribed with a Qianlong imperial poem written specifically for it in 1756 (Gerard Tsang and Hugh Moss, Arts from the Scholar’s Studio, no. 121; also inSotheby’s Hong Kong, 8 April 2010, lot 1728).

Amber is frequently mentioned in the Yongzheng records in relation to snuff bottles, sometimes as an embellishing material, sometimes as a ground. Some of our extant amber bottles must date from the earlier part of the eighteenth century, but without a single credibly reign-marked amber work of any kind from either the Kangxi or Yongzheng eras, we have no reference point for dating them. The bottle in figure I.42 is one of two known amber snuff bottles of this unusual basket-weave shape and style, both of which are well worn and convincingly old. The Marakovic example may be from the earlier stages of the production of basket-weave forms, given its careful attention to detail and unusual three-tiered weave, which was not the standard for the mid Qing; after many years of repetition, designs tend to simplify and become more perfunctory in execution, but that has not happened yet here. This amber snuff bottle may have been carved in the Yongzheng era.

As to the hard materials, they are many that were definitely produced at the palace workshops and elsewhere for the court, and we need only guess at one or two here to represent the many. The bottles in figures I.43 and I.44 are two versions of a rare agate design that was standard in glass; it was certainly produced in the Qianlong reign and probably before, as well. Some examples are smaller than others, more rounded and compact; those may be the earlier examples. The design is a formalised version of the mallow, kui, which has many varieties. The one evoked here is probably huang Shukui 黃蜀葵, Abelmoschus manihot, of the Malvaceae family. It is known by various names in both Chinese and English, where it is called hibiscus manihot or aibika. The leaves, roots, and flowers can be used in medicine, and the plant was praised for its beauty in poetry at least as early as the Song dynasty. But the fact that the flower ‘inclines its heart’ toward the sun and follows its course across the sky has long made it a symbol of loyalty. It and its serrated palmate leaves appear in other art forms as well, but mallow-form glass bottles, many of which have scalloped profiles to match the shape of the petals, were a standard type for Qing production, popular from the Qianlong period onwards, although rarely bearing a reign mark.

Early in the Yongzheng records we find one blue glass snuff bottle embellished with coral mallow flowers and beeswax amber animal masks—it was one of the bottles Prince Yi 怡親王 (Yunxiang 允祥, 1686 – 1730) brought in just a few days after the beginning of the era (on YZ1/1/9), making it a likely Kangxi product. (It was refurbished within six days.) There is one mention of a ‘mallow-shaped snuff bottle’ of unspecified material being brought to the inlay workshop in the spring of 1726 (YZ4/2/17). The shape (shi 式, in the language of the archival entry) was probably like that of the bottle in figure I.44, where the hexagonal profile of the bottle follows the outline of the flower. The small size and rounded detailing of the Marakovic bottle might indicate an early example in the series; of course, we cannot identify it with the bottle brought to the workshop in 1726, but we know that the palace workshops certainly worked in agate and that some of the few credible Yongzheng-marked hardstone vessels other than snuff bottles are in agate.

Any number of perfectly plain, well-made bottles made from agate and other hardstones may date from the Yongzheng period, but the lack of decoration frustrates even educated guesswork as to their proper dates. The bottle in figure I.45, of a style popular at court, could have a dating range of a century and a half, at least. The carnelian snuff bottle in figure I.46 is of an unusual form, but the mouth is of a type commensurate with an early date. The Yongzheng records relate that a ‘red agate’ peach-form snuff bottle was brought to the jade workshop early in the second year of the reign (YZ2/1/28) to have its material recycled. That indicates that material like this carnelian was used in the Kangxi-Yongzheng period, but that supposition does not narrow down the dating range of the Marakovic specimen.

If the Yongzheng emperor was copying turquoise in glass and complaining in 1732 when he spotted a poor match of colour (see above), then genuine turquoise pieces must have been worked at or for the court. One of the only known stone bottles with a Yongzheng mark is a small turquoise bottle with mask-and-ring handles (Robert Hall, Chinese Snuff Bottles XV, The SB Collection [London, 2011], no. 60). As with any engraved mark, the regular-script reign mark could have been added later, but the bottle itself is convincingly early and, with a wide mouth and small size, what we would expect of a Yongzheng example in any case, regardless of the credibility of the mark. The patina and smoothing of the surface detail from handling are also commensurate with considerable age. The Marakovic example (fig. I.47) is more generic. It is of a shape popular during the Qianlong era in porcelain, but we do not know whether stone inspired porcelain or vice versa. Nor can we tell whether this is another Yongzheng turquoise or a Qianlong bottle; all we can say is that it, too, is extremely well patinated and early.

Jade, which was so highly valued in Chinese culture, would have been used for snuff bottles from the outset. Many of the types we associate with the Qianlong era were surely present earlier than that. The well-known palace carving style of the Qianlong is established as such only because a Qianlong reign mark appears on a fairly wide range of them; this does not mean bottles in the same style were not carved earlier and left unmarked. It is true that nephrite was in relatively short supply before the middle of the eighteenth century; tensions with Zunghar Mongol empire were too high for stable trade relations between the jade-producing regions and the Qing empire. The Yongzheng records for late 1732 note the emperor’s efforts, through the grand minister of the Imperial Household Department, Haiwang 海望 (d. 1755), to locate and acquire good jade material, calling on the supervisor of the Suzhou manufactory to help in the search. It should be no surprise, then, that nephrite snuff-bottle production was rarely mentioned in the Yongzheng archives. Despite its name, the jade workshop in the Yongzheng era worked mainly on agate, glass, and other materials where lapidary skills were needed.

Nevertheless, there are candidate nephrite bottles that can be assigned to the Yongzheng era, some perhaps to the Kangxi. We propose three Marakovic examples to represent what was probably a reasonable number of Yongzheng nephrite bottles.

One is the bottle in figure I.48. Double gourds were a standard palace form in many arts from the Kangxi through to the end of the dynasty. Among them is a small group of white nephrite compressed waisted-gourd snuff bottles with matching original screw-threaded stoppers, represented here by the bottle in figure I.48.

We know from the archival records that the concept of the screw-threaded top was familiar at the Yongzheng workshops. Screw-threaded stoppers were added to four aventurine snuff bottles in November 1726 to convert them into oil bottles; six years later, in the tenth year of the era, a set of eight-sided glass ping (bottles) for the storage of snuff was made with bronze screw-threaded stoppers.

The double-gourd form we see in figure I.48 is also attested in the Yongzheng records. On 5 April 1725 (YZ3/2/23), Prince Yi brought in an agate double-gourd form snuff bottle for refurbishing; note that if the bottle was more than three years old, it would have been made in the Kangxi era. In the seventh year of the era (1729), an amber bottle in this form, inlaid with coral, was completed.

Yellow nephrite was a favourite of the Kangxi, Yongzheng, and Qianlong emperors, coming closest among the colours of their beloved jade to the imperial colour. We can be certain of extensive Qianlong production in the material, and we can also assume earlier production. The bottle in figure I.49 is probably from the Qianlong era, but we couldn’t rule out an earlier date. It has a non-specific mark on the foot stating simply that it is zhenwan 珍玩, ‘for precious amusement’, a common enough mark on snuff bottles; there is no reason why the addition of such a mark should not have begun prior to the Qianlong.

Finally, among the early nephrite bottles, figure I.50 shows another palace type, with its off-white nephrite, low-relief archaistic carving and understated elegance combined with impeccable formal integrity. It too may predate the Qianlong reign by a few years.

Before leaving the Yongzheng reign, let us briefly list some of the types found in the first few years of the records, even if there are no obvious equivalents in the Marakovic collection (or in some cases, anywhere else, either). It will suffice to give some idea of the range produced at the time. Many of the materials are used in combination, as embellishments or inlays. There are: glass, including aventurine glass; agate, including white agate; ivory; coconut shell; amber, including yellow amber and waxy amber; hawksbill turtle (tortoiseshell); painted enamels on metal and glass; coral; black clam-shell (possibly a skin of some sort, as we aren’t sure how to interpret this); black crystal; mother-of-pearl; gourd; and lapis-lazuli.