The Yangzhou Conundrum
Hugh Moss and Stuart H.Sargent. August 2018

Over the past half-century, snuff bottles associated with Yangzhou have been the subject of a good deal of research, speculation, and re-assessment. The glass-overlay wares of the town famously see-sawed back and forth across the nineteenth century for several decades before we discovered evidence tying a group of them firmly to the last quarter of the century. It is all very exciting, but after going through so many rounds of re-assessments, we fear that of one of us may end up with a headstone that reads, “He was alive, but he changed his mind.”

We now have cause to re-date the other major glass products of the town, the enamelled-glass wares.

These consist of a broad group linked by distinctive enamels and style of painting. The broader group can then be divided into two: the first with simpler, usually floral designs, with thicker use of the enamels, often on coloured glass (caramel, turquoise-blue, black, and green for instance); the second, the classic wares, with thinner use of enamels, and more complex, painterly subjects usually, but not exclusively, on a distinctive range of white glass. The two can be shown to evolve one into the other, and we believe that the more sparsely decorated wares are the earlier ones. Apart from the stylistic evidence as one style evolves into the other, the whole group is firmly linked by the use of high-quality iron-red seal-script marks in an identical style, almost invariably either Qianlong nian zhi 乾隆年製 (‘Made during the Qianlong period’) or Guyue xuan 古月軒, although a few are unmarked and there is a tiny group with non-imperial designations.

The purpose of this article is to introduce evidence that the entire group, which we once thought were from the late eighteenth century are not.

Until we began research into the second part of our e-book on the Water, Pine and Stone Retreat collection, we favoured an attribution for these wares to the second half of the Qianlong reign.¹ We remain certain that they were made in Yangzhou but now believe they were made for antiquarian snuff takers and possibly for non-snuffing collectors as well, probably including the occasional foreigner, during the latter part of the nineteenth century. They reflect a growing preference among nineteenth-century Chinese snuff takers for old snuff bottles, which eventually gave rise to an antiquarian market catering to non-snuffers. This, in turn, encouraged the use of apocryphal marks. We must leave open the question of whether or not their makers expected all, or indeed any, of their Chinese clientele to be fooled by these designations, but we very much doubt that foreign collectors were advised that the marks were not to be taken seriously as an indication of date.

We have previously entertained some doubts about whether the group was as old as we thought. For one thing, the group is not represented in the William Bragge catalogue of books and “objects connected with the use of tobacco in all its forms,” published in 1880. The album of hand-painted watercolour illustrations produced for Bragge that is now in the Water, Pine, and Stone Retreat Collection does include porcelain bottles with apocryphal Yongzheng reign marks that can only have been made during the early or mid-nineteenth century, so we know that bottles with spurious marks already had a market by then.² But Yangzhou enamelled-glass bottles with such marks are absent. This suggests that they were not yet available to even the most dedicated English collector in the 1850s to 1880.³

In his 1880 catalogue Bragge lists 244 porcelain bottles from the collection and apparently none in enamelled glass. He does describe sixty bottles “in a great variety of rich colours, combinations of colours and kinds of glass,” but that does not sound like the enamelled-glass bottles that are our subject here. Did he disdain our Yangzhou bottles for some reason? Unlikely. He acquired and obviously appreciated a lot of painted-enamel porcelain bottles, and it is doubtful that he could have resisted the enamelled-glass equivalents had they been available; it is equally doubtful that, once he had acquired them, he would have let them disappear into a miscellaneous category of glass with “a great variety of rich colours,’ etc. No: if he had in fact owned any of the enamelled-glass bottles under discussion we would expect him to have illustrated at least some along with the many enamelled-porcelain bottles.

Bragge’s ignorance of our Yangzhou enamelled-glass bottles suggests, then, that the group had not yet reached the West when he was collecting.

There was an additional reason for our growing unease with the late-Qianlong attribution of the bottles: material evidence in the form of one bottle that is typical of the group but carries a mark that would not have been possible before the 1880s and may even date from the first decade of the next century (fig. 1). The enamelling on the bottle is indistinguishable in style and materials from the rest of the classic wares, but the mark tells us that it was made by or, more likely, for something called the Yangzhou Art Society (Yangzhou meishu she zhi 揚州美術社製).⁴

The term meishu 美術, “art,” is the key here. Meishu came into modern Chinese as one of several terms the Japanese employed to translate Western concepts that had no exact counterpart in East Asian discourse. These terms were written in Chinese characters (kanji) and for the most part they may be found in earlier Chinese texts, but they were relatively obscure and sufficiently free of associations and connotations that they could be borrowed to convey newly imported concepts. In the case of meishu, the term originally meant “fine (mei) techniques/skills/values (shu),” but the Japanese revived it (the characters are read bijutsu in Japanese) in the 1870s or 1880s as a translation of the German word kunst. Like many such neologisms in that era, it was quickly adopted in China as a term that would be equivalent to art/kunst/arte/искусство, and so forth. (Perhaps not surprisingly, it prompted considerable debate in China as to which cultural products were properly classified as “art” and which were not.)

The earliest mention of an “art society” we can find in China is in the summer of 1909, and it happens to be connected to Yangzhou. The Shanghai newspaper Shen bao 申報 announced the establishment in Yangzhou of a girl’s school called the Shougong meishu she 手工美術社 (‘Society of Handicraft Arts’).⁵ That school is probably not where our snuff bottle was made, as the skills being taught do not seem to have included enamelling or glassmaking, and it may not even be the same as the Yangzhou Art Society named on the bottle; but the appearance of a Society of Handicraft Arts in Yangzhou in 1909 tells us that the term meishu she, ‘arts society’, was used and understood in Yangzhou in that year. If the enamelled-glass Qianlong- and Guyue xuan- marked bottles were really from the late eighteenth and early nineteenth centuries, as we once thought, the maintenance of that tradition down to this Yangzhou Arts Society bottle in the 1900s, uninterrupted by the mid-century destruction of Yangzhou in the Taiping Rebellion, would have been nothing short of miraculous.

As a side note, a society with this name survived into the twentieth century and may well be the same institution. We have found a record of a Yangzhou bureaucrat heavily involved in the art and culture of Yangzhou, Xue Feng 薛鋒 (b. 1928), where he is identified as a ‘consultant to the Yangzhou Arts Society’ who contributed 400 RMB to an essay contest for young people in 1984.⁶

We long considered this snuff bottle a lone anomaly. Puzzling and worrisome as it may have been, its significance seemed outweighed by the number of questions that would have to be answered if we were to re-date the entire group to the late Qing.

One of these was the similarity of the reign marks in the group to the marks on imperial orders placed at Jingdezhen from the early 1740s until 1799, when the Qianlong emperor died (after having abdicated in 1795). All are neatly written in the same iron-red seal script, impeccably well drawn in most cases, and some had the reversed ‘s’ shaped element in the lower left quadrant of the character Qian 乾 reversed, which we have suggested may indicate production from the abdication years—or least late-Qianlong wares (or later fakes copying them, knowingly or not). The vast majority of the Yangzhou enamels had either this standard Qianlong, four-character mark in seal script or a seal-script Guyue xuan mark, also in iron red, exactly echoing imperial wares from the palace workshops, except insofar as they were in seal script (the enamelled wares painted at the palace bore regular-script equivalents, as a rule).

It was not only the nature of the marks but the fact that the vast majority of this group of Yangzhou enamelled-glass snuff bottles had them. There are very few snuff bottles of this type with either no marks or apparently non-imperial designations. That means that, if an obviously highly skilled enamelling workshop were making these snuff bottles a century later, they would have had to be devoting their talents almost exclusively to producing spuriously marked wares. That seemed unlikely. Taking the reign marks at face value seemed much more reasonable.

The rarity of unmarked and non-imperially marked enamelled-glass snuff bottles in the group not only argued for the group’s imperial nature and early date; it also stood in sharp contrast with the equivalent nineteenth-century glass-overlay production from Yangzhou. That production includes very few Qianlong-marked wares; the majority were made for specific patrons and were either unmarked or honestly marked. If enamelled-glass wares were made in the same time and place, we might reasonably have expected a far greater percentage of them to bear the names of local patrons (or to have no marks). That led us to assume that the two types of snuff bottle came from different times, and if the overlay bottles were indisputably late, the others were probably early.

Other indications of an earlier date were the state of natural wear across so much of the surviving output – they had obviously been made for use and were much used – and the fact that one was excavated from a tomb in Beijing in 2011 (fig. 2). In the photograph, the missing enamel decoration is not obvious, but in the hand the etching left in the glass by firing the enamels onto it is evident and identifies the bottle as obviously part of the classic output of this group. And snuff bottles from tombs around Beijing are typically imperial gifts that had been treasured by the deceased.

There seemed, in other words, more problems arising from assigning them to the late Qing than there were if we considered them imperial orders from the late-Qianlong. We took the majority view, but, as we have seen of late, democracy doesn’t always work out quite as one would hope.

We now have reason to believe that the enamelled-glass bottles cannot be from the earlier period, and strong evidence that they should be re-assigned to the latter part of the nineteenth century with at least some from the last two decades.

Recently we have taken notice of certain books from the 1880s designed to provide models for would-be painters to imitate. The books consist of series of illustrations in black and white of compositions taken from various well-known artists. Among those illustrations are several that are in their essential details identical to the coloured decoration on the enamelled-glass snuff bottles we are re-evaluating.

It is no accident that these books appeared at this time and had the impact that they did on snuff-bottle decoration. Lithography had been brought to Shanghai by the Jesuits in 1876; photolithography followed within a very few years. We do not have the expertise to tell whether the first versions of the pictures at issue were done by lithography or were copied by photolithography from drawings done on paper or silk with a brush. We know these books only through second-generation editions that we know were made by photolithography. The originals were unlikely to have been from traditional woodblock prints—not because such supple, delicate lines could not be achieved by woodblock carving, but because such high-quality production would have made the books unnecessarily expensive at a time when we know lithography was practised in Shanghai.

One of the publications we came upon is the Gujin mingren huapu 古今名人畫譜 (Paintings by famous artists ancient and modern).We shall refer to this book as the huapu. Again, we have not located a copy of the earliest edition, despite the fact that the book was once fairly popular (a copy was purchased by the writer Lu Xun in 1894, when he was only fourteen sui).⁷ ). It is generally said to have been published in 1888 and 1889 in Shanghai, although the publisher’s preface to the 1987 edition (see below) says it came out in several fascicles starting in 1888 and continuing for seven years.­ The 1888 date must be based on the date of the preface by Wang Tao 王韜 (1828 – 1897), which is reproduced in the 1987 edition, but we do not know the basis for saying it took seven years to complete the publication. The earliest edition we have been able to examine is a photolithographic edition from 1926, also published in Shanghai, held by the University of California, Berkeley. It is based on an edition held by the Jiaoying shuwu 蕉影書屋, presumably a bookstore.⁸ Then there is the 1987 reduced-format edition, from which we have taken relevant illustrations for this article.⁹ The 1926 and 1987 editions have almost the same number of pictures—360 and 258, respectively—but they are not exactly the same selection, each edition having pictures that the other lacks. Also, the 1926 edition is remarkably disorganized, especially in the first two fascicles (there are six fascicles, with traditional thread bindings); the 1987 edition (a single volume) is organized thematically.

The differences between these two editions of ostensibly the same book alerts us to the fact that books like this were not treated like sacred texts that had to be preserved in their pristine condition at all costs. The purpose of the 1987 book was evidently also to serve as a tool for the aspiring artist, and if rearranging and changing the selection of pictures improved it, so be it. For us, the important thing is whether a given set of pictures provides the model or models for a given snuff bottle and when that model was available to the enameller.

A second source is the Gujin mingren huagao 古今名人畫稿 (Draft paintings by famous artists modern and ancient), which we shall refer to as the huagao. It seems to have come out in three fascicles, beginning in the late 1800s and ending in the early 1890s. There is a 1984 Beijing reprint that has all three collections. Each collection has its own preface; the first, dated 1888, is also attached to the 1987 version of the huapu, as mentioned above (the 1926 edition has no prefaces!). The second preface is dated 1889, and the third, 1890. A 1973 edition from Taiwan with only the first two is ­a facsimile of an 1891 edition from the Hongbao zhai 鴻寶齋 in Shanghai. An edition of the first collection alone (with 49 pictures) in the Bayerische Staatsbibliothek is dated 1896.¹⁰ In comparison with the huapu, this book has about one hundred fewer pictures, 238. and it arranges the pictures in its own version of a theme-based sequence. (For example, the first collection has ten pictures of human figures, thirty of ladies, and twenty of landscapes.) Each collection has its own preface; the first, dated 1888, is also attached to the 1987 version of the huapu, as mentioned above (the 1926 edition of the huapu has no prefaces!). The second preface is dated 1889, and the third, 1890.

Despite the difficulty of using prefaces to date books when those prefaces may have been taken from one book and added to another, it seems reasonable to say that enamelled-glass snuff bottles whose main decoration is evidently copied from the pictures in these books were not painted before 1888. We shall have more to say about this in a moment, but first let us look at some actual examples.

The most represented artist in these books, by far, is the Yangzhou painter Wang Su 王素 (Xiaomei 小某, 1794 – 1877). whose name is also associated with late-Qing Yangzhou glass carvings, as we have seen.¹¹ His pictures are signed Xiaomei Su, Xiaomei Wang Su, or Xiaomei, and have seals. Some of these Wang Su paintings—more precisely, the black-and-white versions we see in these books—are clearly the model for the decorations on Yangzhou enamelled wares.

For example, fig. 3 reproduces a page from the huapu of a painting by Wang Su that includes a facsimile of the signature Xiaomei Wang Su.¹² Fig. 4 shows the same picture transferred to an enamelled-glass snuff bottle. The only significant difference is in the treatment of the rockery at the lower left (side by side comparison ).

Fig. 5 from the huapu, signed Xiaomei,similarly matches fig. 6, the unusual cylindrical bottle decorated with goats in exactly the same pose.¹³ Wang Su’s setting for the animals has been changed, and his interesting crossing of the horns on the two darker goats has been rejected in favour of a more conventional depiction. But the positions of the animals and their relative lightness and darkness are the same (side by side comparison).

Fig. 7 is signed Xiaomei Su. It matches the design of boys on wading water buffaloes on one side of one of the more famous examples from the collection, one with an apparently non-imperial mark (fig. 8): Yangzhen qingshou 養眞慶夀 (‘Nurture one’s true nature, celebrate a long life’) (side by side comparison). On the other side is a distinctive scene of an equestrian archer shooting a goose in flight (fig. 9) which matches fig. 10 (side by side comparison).

This mounted-archer theme was a popular one, and there is a glass-overlay bottle (dated 1867 and probably from Yangzhou) that gave it quite a different treatment.¹⁴ The archer picture on this enamelled-glass bottle is based on a 1755 painting by Yangzhou artist Hua Yan 華喦 (1682-1756).¹⁵ When we described this bottle for the 28 May 2010 auction of the Bloch Collection (lot 98), we said that the enameller/artist had added background mountains to give a sense of space and height, whereas Hua Yan had relied on the vertical format of the hanging scroll to create the same effect. We now know that the addition of mountains was the work of Wang Su—probably in a painting auctioned by Christie’s, New York, 2 June 1988 (lot 91); although we have not seen that painting, its copies in the huapu and huagao have the mountains that Hua Yan’s version does not have (see fig. 10). There are also two paintings by Ni Tian 倪田 from 1889 and 1890 that have the background mountains and seem close to the Wang Su version as we know it through the lithographic copy.¹⁶

Let us return now to the dating question. Of course, the unquestioned identification of the primary source of some of the designs as Wang Su proves beyond a doubt that they, at least, cannot have been made prior to 1799, as we had formerly proposed – Wang Su was a child of only six sui in 1799. But does that really mean the bottles could not have been painted any time before 1888? Could not the enameller have copied directly from the original paintings before the huapu and huagao were published?

The first answer is that it is unlikely that an artist working for an enamelling workshop would have the same access to collections as one working for a publisher. Perhaps it would happen in isolated cases, but we are dealing with multiple instances here. In fact, the publishers may have been art dealers who copied inventory as it passed through their hands. That is easier to imagine than lithographers (let alone enamellers) going around knocking on doors and asking if they might have a look at the family art collection.

The second answer is that if the two artists, the enameller and the lithographer, looked at the same painting independently, their copies would surely be more different from each other; they would not match each other in so many details, time after time.

Now, we cannot rule out the possibility that Yangzhou enamelled-glass workshops were aware of what the Shanghai publishers were up to, especially since a high proportion of the artists whose paintings were being copied were associated with Yangzhou. If we had a compelling reason to date these snuff bottles to before the publication of the huapu and huagao, we could certainly point to the length of time it must have taken to transform hundreds of paintings into black-and-white lithographs and propose that the glass enamelling workshop had pre-publication access. But that seems unnecessarily speculative, and in any case it would only push those bottles copied from the huapu and huagao back to the mid- to early 1880s.

To return to our enamelled scene of the boys on buffaloes, it is quite possibly related to glass-overlay versions from Yangzhou we can now also date to the mid- to late century (fig. 11 and another bottle once in the Bloch collection).¹⁷ Despite the difference in the medium, the similarity in the concept suggests that the overlay version is also derived from the huapu or huagao, and thus dates from no earlier than the 1880s. This is helpful corroboration of our understanding that the glass-overlay art flourished in Yangzhou about that time.

Figs. 12 and 13 are also a match (side by side comparison), the bottle being an extremely rare miniature, ovoid form for the school, recently from the Lonsdale collection. This time the original is taken from a painting by Fei Danxu 費丹旭 (Xiaolou 曉樓, 1801 – 1850).¹⁸ Fei is not known to have lived, worked, or even visited Yangzhou. His career took him to Shanghai, Suzhou, and especially Hangzhou. In his case it seems most likely that Yangzhou enamellers were less likely to have seen the original painting by Fei than to have seen the copy in one of the Shanghai publications.

We expect to find other matching designs as we find time to check the huapu against the much larger group of bottles in our research files, expanding the research beyond the group gathered in the Water, Pine, and Stone Retreat collection. We already know that the huapu or the huagao provided the Ye family of inside-painted artists active from the early 1890s into the twentieth century with several designs. Some are copied directly and very obviously from the illustrations; others were modified or rethought completely as combinations of figures from the illustrations.¹⁹ Ye Zhongsan 葉仲三 was particularly good at this. It may never be possible to know which book they were using, but it must have been a fairly early edition; the most recent example we have seen in our current research on the bottles in the Water, Pine, and Stone Retreat Collection is dated 1919.

More work is needed to make complete sense of all the late-Qing Yangzhou wares and their stylistic sequence, a task we shall continue to undertake in Parts II and III of the publication on the Water, Pine and Stone Retreat collection. It is evident, however, that as part of a thriving and vibrant snuff-taking culture in Yangzhou, both the glass overlays and the enamels flourished in the second half of the nineteenth century, creating a group of distinctive, high quality wares, primarily for snuff-taking patrons.