Читать книгу Silk, Slaves, and Stupas - Susan Whitfield - Страница 12

TWO

A Hellenistic Glass Bowl

A PHOTOGRAPH DOES NOT DO JUSTICE to the object shown in plate 1b.¹ Although its outside is roughened by weathering, the inside still shows the original rich deep blue of the glass. The dating, its shape, and the groove under the rim suggest it was a late Hellenistic piece, probably made in the Levant. We have to imagine it fresh from the workshop, over two thousand years ago, smooth and unblemished. But it was not admired for long. A century or less after its casting, it was interred, with two similar bowls, in a tomb at Hengzhigang in South China. Here it remained until the tomb was excavated in 1954. Most of its short life was probably spent in the bowels of one or more ships, making their way across the Indian Ocean and the South China Sea with a cargo of trade. Was it once one of many—hundreds, possibly thousands—as seen in ships dispatched in the opposite direction from China with their loads of ceramics? Was it made for trade? And why was it placed in the tomb? Before addressing these questions, let us start with how and where it was made.

The origin of this and the other two bowls is far from certain. In the original Chinese archaeological report they were simply designated as coming from outside China, and such glass of this early period was typically labeled “Roman” by Chinese archaeologists. This is the designation that came with the vessel when it was displayed in Brussels in 2009.² Roman was simply a

For places mentioned in this chapter see Map 3 in the color maps insert.

catch-all for anything Western. On the basis of its date, its shape, the groove under its rim, and the weathering, Julian Henderson has suggested that it is late Hellenistic, made in the Levant and similar to many other pieces excavated there. This is plausible, as there is other evidence that late Hellenistic or early Roman bowls were being imported into China.³ However, there was also a glass industry in China at this time, and Brigitte Borell, who has studied it, suggests that the bowl might rather be a product of an industry local to Southwest China and Vietnam.⁴ Working with objects from the past often involves accepting uncertainty, and while new analytical techniques such as isotopic analysis may promise to answer some of our questions they are unlikely to resolve them all.

THE ORIGINS OF GLASS

Glass is produced from materials occurring naturally that need only heat to transform them, such as silica, found in sand, and alkaline plant ash, the latter used as a flux to reduce the melting temperature of the silica. Volcanic eruptions, nuclear explosions, asteroids, and even mundane events such as burning haystacks can produce glass if the raw materials are present.⁵ Before humans started producing glass they worked obsidian, volcanic glass naturally created when lava flow—rich in silica—cools rapidly. The presence of iron and aluminum makes the resulting brittle and shiny substance dark. It is found worked by humans from 700,000 BC, fractured like flint to create sharp blades and arrowheads. It continues to be used today by some surgeons for their scalpels, producing a sharper and smoother cutting edge than steel.⁶ Naturally occurring glass was also used for decorations: it has been suggested that the carved scarab found in the breastplate of King Tutankhamun (d. 1323 BC) might be from a piece of glass formed by a meteorite crashing into the Egyptian desert.⁷

Most manmade glass is made from silica and a flux, but the process means these are not arranged in any regular way like the lattices of solid materials. The bridges between the atoms of the silica and oxides are broken, and other atoms, such as sodium and calcium, are distributed fairly randomly. It is therefore termed an amorphous material—not like a usual crystalline solid or like a liquid (figure 4). Plastic is another amorphous material.

FIGURE 4. Structure of crystalline solid, liquid, and amorphous materials. After Corning Museum of Glass, “What Is Glass?” December 2, 2011, www.cmog.org/article/what-is-glass.

Glassmaking was the last to be developed of three main nonorganic human technologies (the other two being pottery and metalworking). Before glass the same materials and technologies were developed for the production of vitreous materials such as faience and Egyptian blue. Faience is a hard, highly colored glazed material often reflecting the light and is made from the same materials as glass—silica and plant ash—although fired at a temperature about two hundred degrees Centigrade lower than that required to produce glass. It also is produced on one firing, whereas glass requires two: one to produce the raw glass and then a second to work it into beads and other objects. Faience is found in Mesopotamia from around 4000 BC, in Egypt in the fourth millennium, and at an Early Harappan site in South Asia around 2700 BC.⁸

Egyptian blue, considered to be the earliest synthetic pigment, was made by repeated high firings from silica, copper alloy filings or crushed ore, calcium oxide, and a fluxing material. It was probably developed to imitate lapis lazuli, a stone imported to Egypt over three thousand miles from the mines in present-day eastern Afghanistan.⁹ It appears from around 2500 BC, and later examples show that it was also being made in Ugarit in present-day Syria. It continued in use until around AD 400.¹⁰

The first known manmade glass—in the form of beads—is found in northern Mesopotamia and dates to about 2500 BC.¹¹ The raw glass—known as soda-lime glass—was made from silica and plant ash fused together in crucibles in small quantities. Silica by itself has a melting temperature of over 1700ºC, but when it is mixed with an alkali, such as soda, this can be reduced to about 1000ºC. The alkali is called a flux. The resulting glass slag was then fired again and fashioned into beads by various means, such as winding the molten glass around a clay-coated wire. Glass beads came to be produced in various colors, the glass colored and made opaque with the addition of raw materials. Some of the resulting beads resembled semiprecious stones such as turquoise and lapis lazuli.¹² Raw glass from Eridu in present-day Iraq, dating from around 2300 BC, is made from silica and plant ash, combined with a cobalt-rich material to produce this blue coloring.¹³

Tutankhamun’s death mask contains lapis lazuli along with locally made deep-blue glass. Presumably the substitutions were made because of the cost and rarity of lapis. By this time glass technology had spread into Egypt, possibly taken there by Mesopotamian craftsmen. Beads of the same composition as those on Tutankhamun’s mask are found in northern France and Scandinavia, evidence of early trading in glass.¹⁴ Some scholars argue that glass technology was also developed in the Indus valley and that some beads previously classified as stone are in fact weathered glass. The evidence is uncertain, but given the production of siliceous faience there by around 2700 BC, the development of glass would not be unexpected.¹⁵ Firm evidence for glass production across the Indian subcontinent is datable to around 1450–1200 BC, and there was contact between the Harappan cultures in the Indus valley and those in Mesopotamia by land and sea.¹⁶

The first glass vessels, rather than beads, were also found in northern Mesopotamia and date from around 1500 BC, probably produced under the Hurrian kingdom of the Mitanni.¹⁷ They were core formed and made from silica and plant ash.¹⁸ Making vessels in this way required the fusion of large volumes of glass inside furnaces that could reach approximately 1150 to 1200ºC. The molten glass was used to coat a core of dung and clay that had been shaped around a rod. Trailed-on glass provided surface decoration. Colored glass started to be produced systematically, with the addition of antimony leading to opaque white, yellow, and turquoise, and the addition of cobalt to produce the deep blue. These required further technological innovations: the glass had to be heat treated, so that crystals of the antimony or other substance would form, and then had to be cooled slowly to develop the opacifiers. This resulted in a stronger glass. Sources and supply of the raw materials were also required, not all available locally. This period also sees the appearance of mosaic glass vessels produced in molds and decoration using marbling, further technological innovations.

As well as the export of beads and, later on, finished vessels, unworked glass—raw furnace glass or ingots—was exported as neighboring cultures developed glass technology and were able to work glass into local products. The discovery of a fourteenth-century BC shipwreck off the coast of southwestern Turkey near Kas gives us a glimpse into the network of trade and exchange of this pre–Silk Road time across western Eurasia and Africa.¹⁹ The wreck contained, among many other items, amber from the Baltic, ostrich eggs, ebony and ivory from Africa, an Egyptian gold scarab inscribed with the name of Nefertiti, Cypriot pottery, Canaanite jewelry, pomegranates, cumin, and almonds (and a stowaway house mouse). Copper ore for the 345 ingots (each weighing twenty-three kilograms) came from Cyprus. It was probably a merchant ship of the Syro-Palestinian peoples who lived on the eastern coast of the Mediterranean, and was on its regular circular voyage from Syro-Palestine northwest to Cyprus, thence to the Aegean, and occasionally as far west as Sardinia, then back home via North Africa and Egypt.²⁰ The ship’s cargo included about 175 translucent cobalt-blue and turquoise glass ingots, about fifteen centimeters in diameter. These are known from textual sources to be called mekku-stone, and analysis has shown them to be identical to Egyptian and Mycenaean glass.²¹ Around 9,500 glass beads and 75,000 faience beads, along with others in carnelian and other semiprecious stones, were also found. Many of these would have been items for trade, but others were probably the personal possession of crew members.²² Scholars have also suggested that the one ton of tin found in the Mediterranean shipwreck might also have come from mines in Central Asia.²³

Glass technology developed in different cultures across West Asia and Europe in the following millennium. Glass started being worked in Greece around the thirteenth to twelfth centuries BC; it was fused in northern Italy around the eleventh to tenth centuries BC; and thereafter the technology moved across Europe.²⁴ From about 800 BC there was a significant development in the composition of glass made in western Eurasia with the use of the sodium-rich minerals natron or trona largely replacing plant ash as the alkali. Natron is found on the edge of the desert northwest of Cairo at a place called Wadi el Natrun (Natron valley). Natron was a purer and denser source of the alkali required for glassmaking and, unlike plant ash, did not need prior preparation. It had been used for embalming bodies in Egypt as early as about 2000 BC, but it was relatively rare. However, it became the main source of the alkali in glass in the Levant and Europe for one and a half millennia. The Roman scholar Pliny the Elder (ca. AD 23–79) recorded one account of the start of glassmaking near the Belus River: “A ship of [Phoenician] natron merchants came to shore and when the men were scattered all along the beach preparing their meal, since there were no stones to support their kettles, they put pieces of natron from their ship under them. When these had caught on fire and the sand of the shore mixed with them, there flowed transparent streams of a new substance, and this was the origin of glass.”²⁵

The Greek historian Strabo (64/63 BC–ca. AD 24) also wrote of glassmaking in the region, identifying another source of sand further north along the Mediterranean coast near Sidon. Excavations over the past two decades in Beirut have given evidence of both glassmaking and glass working, and thousands of glass vessels have been uncovered.²⁶ These date from the late Hellenistic period onwards and include the type seen in the bowl under discussion, namely a blue monochrome hemispherical bowl with a decoration of a single groove below the rim. They were made by casting, that is by either pouring the molten glass into or over a mold or heating the raw glass inside the mold.²⁷ Although plain bowls are found, ones with a grooved decoration are most common, the groove probably being cut by a wheel and not formed by the mold.²⁸ Large assemblages have been found of such bowls in this region at every type of settlement.²⁹ Ruth Jackson-Tal suggests that this shows they “were available to most classes of society,” were stored by being stacked upside down, and were pre-dominantly used for drinking.³⁰ Glass vessels became the choice for tableware, preferred over gold and silver as they did not give off an odor.³¹ They were also cheaper, a mark of the transition that took place around this time from the production of “small numbers of luxury core-formed cosmetic vessels to the mass production of simpler cast drinking vessels.”³² As Henderson observes, since silica, in the form of sand, was readily available, economies were dependent on the supply of fuel for the furnaces and the supply of natron, still the principal alkali flux, but also on the development of sagging. This method of casting bowls by placing a flat disk over a mold meant that the mold could be reused and was a quick and cheap method of production.³³

The glass for our bowl could have been produced in one of the tank furnaces uncovered in Beirut or in one of the other glass production centers in the Levant or North Africa. It was not necessarily the case that the vessel was produced in the same center as the glass, but given the ubiquity of this form of vessel in the Levant at this time we can assume it was probably made there along with hundreds or thousands of other similar items. Although many questions about this piece may never be answered, the composition and source of the material for its production could be confirmed using scientific analysis.

SCIENTIFIC ANALYSIS

The chemical composition of ancient glass is complex, with many different compounds found in many different combinations. A variety of oxides are formed from the primary components, such as the sand, quartz, and various alkali fluxes, as well as in the colorant and opacifying materials. In addition, other oxides associated with the impurities are found with the various minerals used for coloring. Up to the 1960s, various analytical techniques were in use to identify these components, but from this time more efficient techniques became available. These include X-ray fluorescence (XRF), scanning-electron microscopy (SEM), particle-induced X-ray emission (PIXE), and various methods of spectroscopy. Each technique provides slightly different kinds of analytical information, and often several techniques are used to cross-check or add to results.³⁴ In isolation, the chemical analysis of a piece of glass will not necessarily add greatly to our understanding. Data need to be collated from a variety of sources in order to build a picture of glass production in different regions at different times. Material found at production sites is a key part of this, as is a knowledge of trade patterns.

Analysis over the past half century has resulted in a much greater understanding of glass: for example, identifying that natron starts to be used from around 800 BC and that glass produced in central China from around 500 BC—discussed below—uses barium oxide and lead with potassium as the flux and is not of the soda-lime type found across western Eurasia and North Africa. These methods would, for example, confirm whether or not the glass for this bowl was made in the Hellenistic or Chinese worlds. Qualitative XRF analysis has been carried out on one of the bowls discovered in the tomb. Such analysis indicates all the elements present in the glass, without providing information on the quantities. This analysis showed the presence of potash, lead and barium, along with silica and various alkalis.³⁵

Two decades ago Henderson noted that “glasses made at two different areas using the same technological tradition but with slightly different raw materials will probably contain recognizably different trace elements and will possibly have different stable isotope signatures.”³⁶ Since then, techniques of isotopic analysis have been developed that have the potential to enhance further our understanding of glass and other material.³⁷ The relative levels of strontium and neodymium in Hellenistic glass have distinguished two different sand sources used in making such glass.³⁸ As Henderson points out, there might be a compositional and isotopic distinction between the sands found on the beach at the Belus River and those at Sidon, indicating the two sources and providing a means of distinguishing glass melted in the two areas.³⁹ An analysis of these sands awaits. Trace element analysis has already provided evidence for more than one production area for Middle Hellenistic glass, one of them possibly in Italy.⁴⁰ This is just the start of a potentially very rich methodology for mapping and understanding in much greater detail the sources of glassmaking, glass production, and glass trade.

In the case of our bowl, however, we are still reliant on its style, form, and dating, along with the basic information provided by the XRF analysis. The former points strongly to this being a late Hellenistic bowl produced in the Levant. The latter introduces an element of doubt into this hypothesis. Without further information the results are inconclusive but certainly leave open the possibility, as Borell points out, that this bowl was produced in China.

GLASS PRODUCTION IN ASIA

In South Asia, as noted above, faience is found in Indus valley cultures, and there are signs of glass production there in the early second millennium BC.⁴¹ The evidence remains fragmentary through the first millennium. Glass with a high aluminum content is found at Rupar in the Punjab from the beginning of the first millennium BC and a high-potassium (potash) glass from Hastinapura in Uttar Pradesh around the middle of the first millennium BC.⁴² From about 400 BC potash glass is also found in Southeast Asia, though the production sites of the Indian and Southeast Asian glass are still uncertain for this period. But by around 200 BC, production sites on the southeastern coast of India were producing glass beads. These, now known as Indo-Pacific beads, were traded to Southeast and East Asia as well as to Africa.⁴³ “The Indo-Pacific bead industry produced one of the, if not the, most widespread and ubiquitous trade item of all time.”⁴⁴ This glass is distinguished by its high aluminum and low lime content.

Potash glass continued to be produced in Southeast Asia, including at sites around the Tonkin Gulf in what are now Northeast Vietnam and Southwest China. A detailed analysis of the potash glass from South and Southeast Asia of this period distinguished several subgroups, suggesting different production sites, possibly including one in Southwest China.⁴⁵

Elsewhere in Asia, glass beads are found in tombs in the Tarim basin, in present-day Chinese Central Asia, dating from around 1000 BC, and in central China from around the fifth century BC.⁴⁶ Scientific analysis shows that most are soda-lime glass, which suggests that they were imports from West Asia, but the crudeness of some of the beads and the presence of lead and magnesium has led some scholars to suggest the possibility of the start of local glass production in Central Asia at this time.⁴⁷ Glass from other Tarim sites dating from the fifth century BC onwards shows more sophistication and is of various compositions, including soda-lime. But other examples show significant levels of barium oxide and lead with potassium as the principal alkali. This glass seems to have been imported from China, as we see evidence of glass technology with these materials in central China from this time. No other place is known to have produced glass of this composition.

Both the relatively late appearance of glass in China and the use of barium oxide are features worthy of comment. By the end of the second millennium BC, potters in China were using furnaces to fire pots with ash glazes at temperatures of about one thousand degrees Centigrade and at higher temperatures within the early first millennium. Glazes are vitreous substances, and it has been suggested that the technology for glazing ceramics in Mesopotamia evolved from the glass industry.⁴⁸ In China, the use of glazes seems, however, to have preceded glass.⁴⁹ The ceramic technology in the Chinese region continued to develop, and by the end of the first millennium semivitreous glazed stonewares were produced that were both hard and able to hold hot liquids. This led to the production of porcelain, which, being vitreous and semiopaque, even more closely replicated some of the qualities of jade. The materials used for stonewares in southern China contained virtually no clay but instead consisted of fine mica or hydromicas. These were 6 to 10 percent potassium oxide, which, when fired, melted some of the silica in the clay body into a stiff glass and toughened the resulting stoneware.⁵⁰ So the technology was available for making glass.

Possibly as early as 1000 BC the Chinese produced a synthetic pigment, called Han blue or purple, that was used for about a thousand years. It was similar in composition to Egyptian blue but was synthesized at a higher temperature. Elisabeth West FitzHugh and Lynda Zycherman have tentatively suggested that the discovery of Han blue—made from barium copper silicate—may have been a serendipitous accident of glassmaking. However, there is currently no firm evidence for glassmaking in China by 1000 BC.⁵¹ The earliest finds are potash-lime glass beads made in central China from about 800 BC. They are replaced from the fifth century along the central Yangzi valley by lead-barium and potash glass.⁵² Han purple is seen used to decorate glass beads found in burials from the second half of the first millennium.⁵³

Lead-barium glass remained the predominant glass in central China up to the early first millennium AD and is found throughout China, north and northwest into the steppe and Central Asia, and south to the sea. Brill and others have suggested that the turbidity that barium produced resulted in a glass resembling jade and that this glass was produced as a jade substitute.⁵⁴ The lead made the glass more brilliant and reduced its melting temperature.⁵⁵ As Gan Fuxi observes, it was natural for the Chinese to use lead as a flux, as they had long experience of using it in bronze working. In addition, both lead and barium ores are found in large quantities along the Yangzi River valley.⁵⁶ The use of saltpeter or potash as an alternative flux to make potash-lime glass is also not surprising given a long history of use in China.⁵⁷ Glass vessels were mold-made, again utilizing a technique well honed from bronze technology.

The evidence therefore points to glass production starting in China around 800 BC but adapting local techniques. Does this suggest that Chinese artisans experimented using familiar technology to emulate this foreign product? But why would they wish to produce it? As discussed above, we see glass being produced for aesthetic and economic reasons. There is no reason to think that glass would have been cheaper to produce than the high-fired stoneware in China, although there might be an economic argument for producing it as a substitute for jade. The argument that technologies develop from aesthetics is an interesting one, made in a structuralist context by Cyril Smith, and might lead us to conclude that the human desire that was met by glass in Mesopotamia, Europe, and Egypt, for example, was largely satisfied in China by high-fired pottery.⁵⁸

THE AESTHETICS OF GLASS

Why was glass produced? What set it apart from other available materials? Perhaps, as Smith suggests, it was “the desire for decorative objects” that led to the discovery of the materials, processes, and structures of faience and then glass technology.⁵⁹ The invention of Egyptian blue was driven by an aesthetic desire for the deep blue pigment but possibly also by economic need. The aesthetic appetite had been whetted by the introduction of lapis but could never be satisfied given that the logistics and cost of importing the material from thousands of miles away ensured it would always be relatively scarce and expensive. Blue glass could substitute for the rare stone. Later in Southeast Asia translucent prismatic-cut glass beads appear resembling the beryl crystals found in South India.⁶⁰ Beryl is often used in Buddhist relic chambers, representing one of the seven treasures of Buddhism, but glass is also found, possibly as a crystal substitute (see chapter 4).

In the early cultures of what is now China, jade became the most valued stone, representing imperial and religious power (see chapter 1).⁶¹ It was worked from Neolithic times into copies of weapons and tools, but also into forms that clearly had a ritual meaning. It is probable that much jade was imported two thousand miles from Khotan.⁶² This, and the skill and time required to work it, probably made it as valuable to the Chinese emperors as lapis was to the Egyptian pharaohs.

It is therefore not surprising that some of the early glass objects discovered in China, such as glass bi, seem to be emulating jade.⁶³ Given the value but also the expense of jade, it might be expected that other materials that could simulate it would be sought. But while glass has some of the same qualities that were valued in jade, such as a certain translucency and hardness (glass is 5.5 and jade is 6 on the Moh scale of hardness), it is more brittle, and, perhaps more importantly for an aesthetic argument, has a different feel. Glass, as an amorphous substance, is warm to the touch, whereas jade, as a crystalline material, is initially cold but then slowly warms in the hand. The jade collectors of Khotan waded barefoot through the river, since they were said to be able to identify jade from its feel on their feet.⁶⁴

As Hsueh-man Shen has observed, early Chinese cultures did not know where to fit glass in their taxonomy. This divided “stuff ” into the elements of metal, wood, water, fire, and earth. The elements of pottery were apparent: it was made from earth and transformed by fire. But the elements of glass were unclear to the Chinese for many centuries. Comparisons were made with pottery but also with metal, precious stones (particularly jade), and even water.⁶⁵ This ambiguity is reflected in the terminology, with words adopted from outside. The term liuli 琉璃 came in during the Han period and was used to refer to glazes as well as to opaque glass and gemstones. Both Chinese characters making up liuli contain an element or radical for jade (玉). It possibly derives from the Sanskrit word vaiḍūrya, referring to blue and green stones, including lapis lazuli, as does the Chinese term boli 玻璃, which became the primary term for translucent glass.⁶⁶ Its possible Sanskrit origin—sphaṭika—also referred to a crystal or quartz.⁶⁷ Both terms almost certainly came in with Buddhism. We do not know what terms were used to refer to glass products prior to this.

Cecilia Braghin argues that glass technology remained a “marginal tradition in China” and that production “appears to have been encouraged by contacts with imported glass artefacts.”⁶⁸ Was this also because the tradition, developed early in the cultures of China, for jade and ceramics covered the aesthetic range offered by glass, so that as a material it was seen to offer nothing new? Jade and ceramics were relatively plentiful, and the technology had been honed over thousands of years by the time glass was introduced into China. In Braghin’s scenario, the original impetus for glassmaking in China came from glass beads from West Asia. It was then revived by the arrival of Hellenistic vessels, as found in central China and, as in this piece, in a tomb on the southern coast. This is supported by the two references to glass in the histories of the Han period (206 BC–AD 220), which mention the emperor importing it from Central Asia and from the kingdoms in what is now southern China (see below). The third- to fourth-century alchemical text Baopuzi Neibian (Book of the master who embraces simplicity, inner chapters) also notes the process of glassmaking in the South while acknowledging that it was not invented in China and was also imported. Ascribing glassmaking to foreign craftsmen is found in the Chinese histories. For example, the Beishi (History of the Northern Dynasties; 386–581) notes the visit of a merchant from Central Asia in the mid-fifth century.⁶⁹ The Suishu (History of the Sui dynasty; 581–618) contains a reference to a man called He Chou (540–620), described as a descendant of a Sogdian family who were specialists in technology of the region. He is credited with reviving glass technology in China at this time.⁷⁰ Its appearance in tombs of the elite shows that, even if glassmaking never became central to the culture, glass objects were nevertheless valued.⁷¹

TOMBS OF THE SOUTH

The tomb where our bowl was found is in present-day Guangzhou in the far south of China and is dated to the Former Han (206 BC–AD 9). This area has an interesting history. Qin Shihuangdi (r. 246–210 BC), the “first emperor” of China, was originally ruler of one of the so-called Warring States who managed to conquer the other kingdoms and set himself up as the emperor of a united China in 221 BC. He expanded Qin rule southwards, into what was home to very different cultures from the agricultural heartland or the steppe to the northwest, collectively known by the Chinese as the Yue peoples. The region to the south of the Nanling Mountains was called the Southern Yue—or Nan Yue (Nanyue) in Chinese.⁷² The military commander of this region for the Qin was named Zhao Tuo, and at the end of the Qin dynasty (221–206 BC) he declared Nan Yue a separate kingdom (in Vietnam—Yue nan—it is called the Triệu dynasty). The kingdom included a belt of land along the sea extending into present-day Vietnam, as well as modern China’s Guangxi and Guangdong provinces. The capital was in Panyu, modern-day Guangzhou, where royal tombs have been excavated.⁷³ Zhao Tuo, his grandson Zhao Mo, and their three successors largely retained the kingdom’s autonomy until 111 BC, when an army sent by the Chinese Han dynasty managed to impose their rule. Although there were a number of uprisings—most famously that of the Trưng sisters in AD 40—these were unsuccessful, and the region remained under Han Chinese authority.

The brief published archaeological report on the 1954 excavation of this tomb dates it to the later part of the Former Han period, suggesting that it probably dates from after the fall of Nan Yue, but a legacy would have remained especially given the diverse population of the region. When the tomb of Zhao Mo, the second ruler of Nan Yue (d. 122 BC), was excavated nearby, the tomb objects showed influences from the cultures of the steppe, West Asia, Vietnam, and central China. There was a spectacular collection of jade, and, most notably, the corpse was dressed in a jade suit. This was a tradition developed in China and reserved for emperors and their families; it was believed to confer immortality. Zhao Mo’s jade suit was composed of 2,291 plaques, many of them reworked from other objects. But the tomb also contained glass beads and twenty-two blue glass plaques with gilt bronze frames, measuring ten by five centimeters. Five pairs had been wrapped and placed face to face in a bamboo container for burial. Two were discovered in the burial chamber of the wives, but six were in the main burial chamber. The glass is lead-barium as made in central China, but the plaques, in their shape, size, and position around the middle of the corpse, seem to be belt plaques, as commonly found on the steppe. Chinese gentlemen wore long robes, while the horsemen of the steppe wore short robes—more suitable for riding—and hung their dagger and other accouterments from a belt. Zhao Mo’s family came from northern China where the steppe met the settled, and Lukas Nickel suggests that he might have had mixed ancestry, which “made him appreciate a drinking horn, robes in nomadic fashion and a box with foreign-inspired decoration,” all objects buried with him.⁷⁴ This would explain the belt plaques, although the use of glass remains unusual. Glass beads, necklaces, and a glass bi were also found in the burial.

Over two thousand other tombs from the same period have been excavated, some near Hepu and Guixian, also part of the Nan Yue kingdom and nearer the Vietnamese border. While many contained glass beads, only eleven contained glass vessels—one or two ribbed glass bowls (figure 5).⁷⁵ These, along with another bluish-green bowl found in a tomb in central China, were originally identified as Roman.⁷⁶ The latter is still believed to be early Roman—showing that some glass vessels did indeed make their way to China at this time.⁷⁷

FIGURE 5. Form of glass bowls found in Hepu and Guixian, South China. After Borell (2011: fig. 3.1).

Borell concludes that these were tombs of those who held a status just below the top rank. The mold-made shape of the bowls is distinctive, unlike anything Roman or Hellenistic, and analysis has shown them to be potash glass with a low magnesium level, which suggests a mineral source for the potash, possibly saltpeter. Borell has argued convincingly that these were products of a local industry and that such glass was exported from this region by the maritime routes: similar pieces have been found in Arikamedu in southeastern India. She places the dating of the start of this industry to the middle or late Former Han, that is, after Nan Yue. The use of glass imported from central China in the Nan Yue royal tomb suggests that the material was valued and might have prompted the start of this local industry. But the industry could also have been influenced by glassware coming from central China, from elsewhere in South or Southeast Asia, or from further afield.

The preliminary scientific analysis of one of the bowls found with the one under consideration here from the Hengzhigang burial has shown some presence of potash, and given this Borell suggests it might also be a product of a local workshop. However, the presence of some potash is not in itself inconsistent with the bowls being Hellenistic, and without further evidence the shape, color, weathering, and design still point toward a Hellenistic origin. If we accept this—allowing some element of uncertainty—then the question remains how these bowls came to Guangzhou.

Analysis of the many beads found in the other burials shows that they are of four different types, including the local potash and the Chinese lead-barium glass. Borell concludes that this is evidence for “a complex network of interregional exchange.”⁷⁸ A large part of this exchange was almost certainly by sea.

MARITIME TRADE ROUTES

Maritime trade routes connected Africa and Eurasia from earliest times, with trade relations, for example, between the Harappan civilization of the Indus valley and Mesopotamia in the third millennium BC.⁷⁹ By the second half of the first millennium, spices from India were reaching Greece via the ports of southern Arabia.⁸⁰ Within a few centuries, sailors were making use of the monsoon winds: the Greek historian Strabo (64/63 BC–ca. AD 24) tells of a man called Eudoxus of Cyzicus who made the return journey twice, setting sail from the Egyptian Red Sea in the time of Ptolemy VIII Euergetes II (r. 145–116 BC).⁸¹ This use of monsoon winds opened up easier access to ports in southern India and, from there, to Southeast Asia.

At Arikamedu near Pondicherry in southeastern India, excavations have shown the port developing from the mid-third century BC with a distinctive type of pottery produced in the Hellenistic Mediterranean, rouletted ware, being imitated in Arikamedu from the mid-second century.⁸² The presence of unfinished agate pendants in a Southeast style were, as Bérénice Bellina points out, probably also imitations intended for export to Southeast Asia, where Indian imports are common.⁸³ The site also contained numerous locally produced glass and stone beads that were also widely exported.⁸⁴ Imported glass vessels discovered here, although not in great numbers, include soda-lime Hellenistic and South China potash glass.⁸⁵

The southwestern coast of China and the coast of northern Vietnam bordered the Tonkin Gulf, protected from the ocean by the island of Hainan. Petroglyphs found along the coast suggest the Yue people were accustomed to traveling by sea. In 1975 the excavation at Zhongshansilu to the east revealed a shipyard dating from the third century BC. It is estimated that this could have been used to build ships twenty-nine meters long and three to six meters wide that could carry a cargo of twenty-five to thirty tons.⁸⁶ The Chinese official history of the Han dynasty describes a route from near Hue, in present-day Vietnam, to what is now the island of Sri Lanka via the Malay Peninsula, several stops in Myanmar, and Chennai on the east coast of India.⁸⁷ The Chinese were said to have carried silk and gold to trade for pearls, precious stones, and other items—including one named biliuli in Chinese, which could mean glass.⁸⁸ A poem from a later period shows the continuing importance of this region for trade, including slaves (see chapter 10): “Argosies laden with slaves from the sea—rings in their ears; / Elephants laden with girls of the Man—bodies bound with bunting.”⁸⁹

A design on a vessel in Zhao Mo’s tomb is a depiction of four ships, clearly showing a rudder that would have been essential for such navigation (figure 6).⁹⁰ Clay and wooden models of both river and seafaring ships have been found in other Han-period tombs.

FIGURE 6. A line engraving on a bucket from the tomb of the king of Nan Yue (ca. 122 BC), Guangzhou, showing a wooden boat. After Erickson, Yi, and Nylan (2010: 166).

The transport by sea of this glass bowl and its two companions at this time from a Red Sea or Gulf port via India and possibly several other landfalls to South China in this period is therefore certainly possible. We can tentatively date the production of the glass bowl to the late second century or early first century BC and its burial in the tomb to the first century BC. We do not know whether it was exported directly after production, possibly taking a year or more to reach its final destination. Given that it was found with two other similar bowls, we might hypothesize that they were part of a much larger glass cargo. How did they come into the possession of the buried man, or his family? What was their significance? Were they even used before being buried with him, or were they acquired solely for burial?

These, and many more questions, will probably always remain without definite answers. However, because the elements used to make glass have a distinctive geographical signature based on their isotopes, refined methods of isotopic analysis combined with trace element analysis developed in the past few decades have enabled tests that will confirm the geographical source of the materials and therefore suggest the probable production area. Such isotopic analysis of this piece would provide support—or otherwise—for our hypothesis: that it was Hellenistic and made in the Levant. Discoveries of other pieces, in tombs, in shipwrecks, or elsewhere, might also add support or suggest other stories we can tell about this piece. This story is not yet finished.

If this is confirmed as a Hellenistic piece, then it was one of many such items made as an everyday drinking cup and not as a luxury item. But far away in South China it was rarer and almost certainly considered as both exotic and a luxury: it might never have been used for its original purpose. A possibly analogous situation is seen in the Chinese ceramics exported by sea to Europe from the fourteenth century. These were usually everyday items or pieces made for the export market and were far from being considered luxuries in China. But once they reached Europe they acquired a new and much higher status.⁹¹ The status of the glass vessel—as a foreign luxury—might account for its inclusion in the burial, like the silverware ewer and Sasanian glass bowl that were buried with Li Xian (see chapter 5). In his study on the trade of glass beads in Asia, Peter Francis has argued that “the farther from the manufacture one goes, the more a product is apt to be treated as a luxury item.”⁹² Most burials contained glass only in bead form—not whole vessels—and this tells us something about the status of the buried man: the excavation team concluded he was a member of the elite.

The tomb was one of three excavated from the same small tumulus at Hengzhigang in the northeast of the city of Guangzhou. They were dated to the middle or late Former Han period by their structure, size, and burial goods.⁹³ The broken fragments of the three bowls were discovered in tomb no. 1. This had a burial chamber six meters below the original surface level of the earth and oriented south southeast. The surviving shaft was completely filled with fine sand. There were two gutters on the floor of the chamber used for the sleepers that supported the wooden burial chamber. Only a few blackened traces remained of the coffin.

Most of the goods had been placed in the north and west of the chamber. Among them were seventy-one ceramics, a bronze tripod, three glass bowls, and one glass bi. One of the bowls was placed in the north of the chamber; the other two, along with the bi and a bronze belt hook, were placed next to the coffin. The only other glass found was a green belt hook in tomb no. 2.⁹⁴

The glass bowl had changed over the two thousand years of burial. It did not decay and disappear in the way of organic material, but its surface and composition were affected by the long period underground, making it look very different from when it was first produced. This so-called weathering is typically seen on buried pieces and is more common on potash glass than on soda-lime. The glass bi found in the same tomb did not survive excavation but disintegrated into dust, perhaps suggesting that the material was severely weathered potash glass. But the type of weathering seen on the bowl is consistent with other soda-lime pieces made in the Levant.

Moisture is the primary cause of deterioration, as it causes the alkali ions to be slowly leached out and replaced with hydrogen ions from the water. This usually occurs within a few years of the burial, and then a new cycle starts, so that the different layers can often be seen. They range in thickness from one to about twenty-five micrometers. Sometimes they protect the glass from further weathering by providing a sort of laminated layer that slows down access of moisture to the glass underneath.⁹⁵ The presence of moisture is not surprising in a burial in this region and is supported by the fact that no surviving organic materials were found in the tomb and that both the corpse and the wood of the coffin and the tomb chamber had completely decayed. The excavators note the custom of this time of using sea sand to pack the wooden chamber—evidenced by the sand found in the shaft—and suggest that the moisture from the sand caused the wood to decay.

The glass bowl, which started life as a utilitarian object and became an exotic luxury, is now an aesthetic and historical artifact. It continues to be considered an item of value in the Chinese context. This is shown by the fact that, instead of remaining in a local or provincial collection or museum, it was sent to become part of the collections of the National Museum of China in Beijing (formerly the History Museum). As demonstrating one of the links across Eurasia in premodern times it has also acquired significance in a “Silk Road” context and has appeared in several exhibitions and publications—although erroneously labeled as Roman.⁹⁶ Its new life has just begun and it has many secrets yet to reveal.

As this piece was being made and transported, a new factor came into the equation that was to have an impact on the use and value of glass in China, namely the use of glass in Buddhism.⁹⁷ This is discussed further in chapter 4.

1. I am indebted to many scholars for this chapter, but especially Julian Henderson of Nottingham University. The work and comments of Cecilia Braghnin and Shen Hsueh-man have also been invaluable. All mistakes, misunderstandings, and omissions are my own.

2. Whitfield (2009: cat. 48). An Jiayao (2004: 58) also described it as Roman.

3. See below, note 76. Conventionally the Roman period begins in 27 BC. Late Hellenistic refers to the period preceding this.

4. Borell (2011).

5. As Henderson points out, “Glassy slags can be produced in virtually any high-temperature environment” (2013: 6).

6. Buck (1982).

7. See Henderson (2013: 5–6) and “Tut’s Gem Hints at Space Impact,” BBC News, July 19, 2006, http://news.bbc.co.uk/1/hi/sci/tech/5196362.stm. Manmade glass is also extensively used in Tutankhamun’s tomb.

8. Hodges (1992: 125); McCarthy (2008: 915). For a discussion of faience and glass, see Henderson (2013: 14–16).

9. Its name in Egyptian means “artificial lapis” (Pagès-Camagna 1998). On the trade, see Tosi (1974).

10. In 1824, the Societé d’Encouragement in France offered a prize of F 6,000 to anyone who could produce a synthetic variety of lapis lazuli pigment at a price not to exceed F 300 per kilo. The prize was not awarded for four years. Finally, in 1828, it was awarded to Jean-Baptiste Guimet. Guimet’s ultramarine was sold for F 400 per pound (lapis lazuli cost between F 3,000 to F 5,000 per pound at that time). For preliminary observations on the economics of lapis as a pigment on the eastern Silk Road, see Whitfield (2016).

11. Moorey (1994).

12. For a history of beads, see Dubin (2009).

13. Henderson (2013: 134).

14. Also glass from Mesopotamia that has been found in northern France.

15. Basu, Basu, and Lele (1974); McCarthy and Vandiver (1991). For example, faience bangles from Gola Dhora are discussed in “Gola Dhoro (Bagasra),” n.d., accessed September 14, 2017, www.harappa.com/goladhoro/faiencemaking.html.

16. Lal (1987). The trade in lapis and in other semiprecious stones such as carnelian is material evidence of these links (During Caspers 1979), but whether there was any transfer of glass technology from Mesopotamia to the Indus valley is more difficult to establish.

17. Henderson (2013: 134–45).

18. Moorey (2001: 4) argues that the technology was developed earlier and refined under the Mitanni.

19. Pulak (1998). For a brief description of the ship and for a general introduction to ships across the Silk Road, see McGrail (2001: 123–25).

20. Bass (1987: 699; Cline 1994: 100).

21. Jackson and Nicholson (2010).

22. Ingram (2005).

23. Pulak (1998); Muhly (2011); Hauptmann, Madding, and Prange (2002).

24. Although there is no firm evidence that it was fused in Europe.

25. Trowbridge (1930: 95–96).

26. Kowatli et al. (2008) and Jennings (2000). For a discussion of the tank furnaces discovered in Beirut, see Henderson (2013: 215–22).

27. Jennings (2000) distinguishes between slumping, slagging, and casting.

28. Jackson-Tal (2004: 19).

29. Jackson-Tal (2004: 19n22, 22–23 for a list).

30. Jackson-Tal (2004: 17, 27).

31. Henderson (2013: 207), quoting from Petronius’s Satyricon.

32. Jackson-Tal (2004: 27).

33. Henderson (2013: 212). There was an additional major development in the next century in glass production, the invention of glassblowing.

34. Henderson (2013a). For his introduction to the various techniques, see 8–23.

35. Fan and Zhou (1991).

36. Henderson (1995: 62).

37. For a simple introduction, see Hirst (2017).

38. Henderson (2013: 238–40).

39. See Henderson (2013: 240, 326–34) for the importance of the environmental approach in isotopic analysis.

40. A. Oikonomou et al. (2016).

41. Kenoyer (1998: 176)—although the finds are very weathered and there has been discussion about whether they are glass.

42. Brill (1999, XIII 335, sample 443).

43. “From Mali to Bali,” in the words of Peter Francis. See M. Wood (2016) for an update on Francis’s conclusions about trade to Africa.

44. Francis (2002: 41).

45. Lankton and Dussubieux (2006). See Borell (2011) for the argument about the Southwest China production site.

46. Gan (2009b: 56–57) and Wang Bo and Lu (2009). Faience is also found further east from around the same period, and since there is no evidence of faience production in this region and the find-site this also suggests that they came by land routes from regions further west (Brill 1995: 270). Soda-lime glass eye beads have been discovered in the Xu Jialing Tomb, Xichuan County, Henan Province, in central China. For images and analysis, see Gan, Cheng, et al. (2009).

47. Li Qinghui et al. (2009: 343); Q. Li et al. (2014).

48. Paynter (2009).

49. Kerr, Needham, and Wood (2004: 464) note the “puzzling example” of a high-potassium glaze found on a Chinese vessel from the second or first century BC and point out the affinity with high-potassium glassmaking in South and Southeast Asia, the latter including South China.

50. Kerr, Needham, and Wood (2004: 59–60).

51. West FitzHugh and Zycherman (1992). A fourth-century textual source records the probably much older legend of the goddess Nuwa, who “smelted stones of all five colours to patch up the flaws” in the sky when it had been damaged by the collapse of a supporting pillar. This is often cited as a reference to an earlier glassmaking tradition. The “five” or various colors becomes a common motif for glass; see Shen Hsueh-man (2002).

52. Gan (2009a: 8).

53. Easthaugh et al. (2007: 36).

54. Brill, Tong, and Dohrenwend (1991: 34).

55. The preference in Chinese culture for hot drinks—in terms of both temperature and their effect upon the body—is well evidenced from a later period, although is not clear in this early period. A thousand years later, Islamic glass was praised by the Chinese for being able to contain hot liquids (Shen Hsueh-man 2002).

56. Gan (2009a: 20).

57. Gan (2009a: 21).

58. “The discovery of the materials, processes, and structures that comprise technology almost always arose out of aesthetic curiosity, out of the desire for decorative objects and not, as the popular phrase would have it, out of preconceived necessity” (Smith 1981: 347).

59. Smith (1981: 347).

60. For example, at Ban Don Ta Phet (Reade 2013; Glover 2004: 75).

61. For a discussion of the role of jade in Chinese culture, see chapter 1. Also see Rawson (2002).

62. This trade might have started by the end of the second millennium BC, since jades from the tomb of Fu Hao (ca. 1200 BC) have been identified by some as from Khotan (Di Cosmo 1996: 90). However, see chapter 1 for doubts expressed by some scholars on Khotan as an early source.

63. For example, a bead necklace unearthed in Suzhou (Gan 2009a: 3, photo 1.2). Shen Hsueh-man has pointed out, however, that this does not mean that the glass bi was considered a cheaper alternative. As she notes, glass was probably as difficult to work as jade, and making a traditional shape in this material might add to its value (pers. comm., January 16, 2016).

64. “The Art of Feeling Jade,” Gemmologist, July 1962, 131–33.

65. Shen Hsueh-man (2002: 72–73). See chapter 4 for further discussion of this.

66. Note that liuli and boli are the modern pronunciations.

67. For a discussion of the names used for glass, see Schafer (1963: 235–36) and Brill (1991–92).

68. Braghin (2002: xi). This became one of the seven treasures of Buddhism. However, Francis asserts that “China was one of the great glass beadmaking and trading nations of the world” (2002: 54).

69. Quoted by Kinoshita (2009: 255)—he suggests the merchant was from Kushan, but this was probably post-Kushan.

70. Quoted by Kinoshita (2009: 256).

71. For example, Lullo argues that glass replicas prepared for tomb objects held more value than other substitutes, partly because of their association with “foreign exotica” (2004: 17, 22). See also chapter 5 for a discussion of “exotica” in tombs.

72. Brindley (2015). (The name Vietnam is a transposition of this—Yue nan.)

73. Lin (2012: 233–44).

74. Nickel (2012: 105).

75. Eleven tombs out of the approximately two thousand excavated.

76. The broad dating of the tombs near Hepu and Guixian to the Han period, i.e., 206 BC–AD 220, does not help in the identification of the glass bowls as either late Hellenistic or early Roman. However, fragments of a bowl made of ribbed mosaic glass from the same region are in a tomb dated to AD 67, and, as Borell (2011: 61) points out, its pattern seems to be emulating a stone—murrhine—that Pliny mentions as being introduced from Parthia in the first century BC. Such bowls were produced in the eastern Mediterranean and exported widely throughout the Roman Empire (27 BC–AD 1453).

77. Analysis has shown it to be natron-based soda-lime glass (Borell 2010: 128).

78. Borell (2011: 59).

79. “A great amount of evidence testifies to the lively trade relations in the Gulf and the Arabian Sea in the late third millennium. It was both direct and transit trade. The main stations were the Sumerian ports in Southern Mesopotamia, then Dilmun, Makan, and Meluḫḫa, or with modern names, Bahrain, Oman with Eastern Iran and the ports of the Harappan civilization” (Karttunen 1989: 330).

80. For an introduction to ships from this time, see McGrail (2001).

81. Thiel (1966); Salles (1996).

82. The chronology of Arikamedu is based on that proposed by Begely (1983), which, as Salles (1996: 262–63) points out, suggests direct or indirect contact between the Hellenistic world and Southeast India by the second century BC.

83. Bellina (1997); Bellina and Glover (2004). Southeast Asia has a long history of maritime activity, but there is more sustained evidence from the second half of the first millennium BC.

84. Francis (2002: 27–30).

85. Borell (2010: 136–37).

86. Ting (2006: 46). For other excavations of ships in China, see McGrail (2001: 360–78).

87. Hanshu, trans. Needham, Wang and Lu (1971: 444)—see also Borell (2010: 136).

88. Loewe (2004: 75–77).

89. Schafer (1967: 76), citing Du Xunhe (846–904). Man was a Chinese term meaning “southern heathens” and was used by this period indiscriminately for all the different peoples of the Southwest; see chapter 10 for a brief discussion of the “other” in Chinese and other cultures.

90. Erickson, Yi, and Nylan (2010: 166).

91. Munger and Frelinghuysen (2003).

92. Francis (2002: 57). Geographical distance is only part of the story. Distance from the potential owner in terms of cost is also a factor: a Gucci handbag is valued in China and Italy, even though manufactured in the former and branded to the latter.

93. For a description of Han tombs, see Erickson (2010: esp. 13–15).

94. Belt hooks are often also made of jade, so this one was probably emulating jade. As mentioned above, belt hooks were not part of traditional Chinese clothing.

95. See van Giffen (n.d.); Craddock (2009: 235).

96. Exemplifying the dichotomous and therefore inevitably simplified interpretation of the Silk Road as linking China and Rome (see Whitfield 2008).

97. See Braghin (2002) for a discussion of this and Shen Hsueh-man (2002) for an inventory of glass in later Buddhist stupas in China.