In AD 60 the town of Colchester, about 85 kilometers northeast of London (Fig. 2), was burnt to the ground and its inhabitants slaughtered during a native revolt led by the Oceanian queen Boudica. The remains of this early town are now buried 2 to 3 meters below the modern town, and so are not normally accessible. Redevelopment of large parts of the historic town center over the past 25 years, however, has given the Colchester Archaeological Trust the opportunity to excavate the Roman town, including these mid-lst century levels (Fig. 3).
The excavations have provided a treasure trove of information about how the early Roman settlers lived and the things they used (Crummy 1984, 1992). This article will focus on one small aspect of this material, namely the vessel glass. The glass fragments recovered during these excavations, and from earlier ones just out-side of the town, are an invaluable resource for studying early Roman glass, not only in Britain but elsewhere in the Empire as well. To understand quite how useful they are, let us first place Colchester in its wider setting in the Empire, and then consider the state of the Roman glass industry at the time.
The Roman Invasion
Britain came late to the Roman Empire. Though Julius Caesar had conducted two summer campaigns in Britain in 55 and 54 BC, it was not until almost one hundred years later that Rome, under the Emperor Claudius, decided to incorporate it into the Empire. At that time the Trinovantes occupied lands north of the Thames in the present-day area of Essex, with their capital at Carnulodunum, modern Colchester. The hostile acts of this tribe against others who considered themselves to be Rome’s allies gave Claudius an excuse to invade, or at least as much of an excuse as a Roman emperor ever needed (Fig. 4). Rome invaded in AD 43. Colchester was obviously seen to have symbolic significance in the conquest for Claudius himself came from Rome to enter the town in state and inaugurate Britannia, the new province. Colchester became the headquarters of the 20th legion, and a fortress covering about 20 hectares was built.
In AD 49 the legion moved out, leaving the fortress and its buildings to become the core of a new colonia. A colonial was the highest ranking type of town in the Roman empire. The institution had originated from the need to settle legionaries with grants of land after their period of service in the army had come to an end. The foundation of a colonia at Colchester would have been a good strategy for the Roman authorities. It would have kept many of the soldiers whose tours of duty had expired in the new province, and effectively given the authorities a unit of reservists in an area from which they had formally withdrawn the army.
The new colonia developed rapidly. The fortress buildings, though useful for shops and houses, were less well adapted for the grandiose public buildings that were being planned, and a new area was laid out to the east of the fortress (Fig. 3). In the process of this expansion, the ramparts of the fortress were demolished, but no other defensive measures were put in their place. It was a decision that would be bitterly regretted. In the early 50s, however, there were no worries about the future, and the authorities concentrated on providing the amenities that would turn Colchester into a showpiece. In the eastern part of the town we know that a theater and a Temple to Claudius were erected; no doubt other early public buildings remain to be uncovered. The Temple of Claudius would have been the center of the imperial cult in the province. Its presence is a good indication that Colchester was intended to be the capital city. Colchester supported a population of comfortable means, well able to buy fashionable consumer goods including glasswares from elsewhere in the Empire.
The Roman Glass Industries of the Mid 1st Century
By a fortunate coincidence the time at which Britain became part of the Roman Empire was also one which saw major changes in the industries producing Roman vessel glass. Glass blowing had been invented by the middle of the 1st century BC, probably in the east Mediterranean area (Israeli 1991). The new technique revolutionized the production of glass vessels. Before its invention, most vessels had been produced by techniques that are known collectively as casting. At its simplest, casting involves heating a flat disc of glass sufficiently for it be sagged over a hemispherical shape to form an open bowl, but more complex angular forms could also be produced (Grose 1989).
Though the casting techniques were used to produce a wide range of cups, bowls, and dishes, they did have two great disadvantages. The first was that only open forms could be produced with any ease, so glass was not usually used for such items as jugs, bottles, or flasks. The second was that vessels could not be produced quickly and cheaply. Blowing was a much faster process, and one that could produce open and closed forms with equal ease. It also allowed for much thinner vessels which could make the most of the transparent properties of glass. Cast vessels, by contrast, were rarely transparent. Instead, brightly colored glasses were favored, probably in imitation of semi-precious stones such as agate.
Blown vessels did not immediately supplant those made by casting techniques, and for about a century the two different technologies continued side by side. The mid 1st century AD saw the final flowering of cast vessels and the rapid development of blown vessels; by the later part of the 1st century nearly all vessels were made by blowing. For glass this was a period of great change, and at Colchester we have an assemblage that captures this moment exactly.
The Boudican Uprising
In Book 14 of the Annals of Imperial Rome Tacitus gives us a vivid account of Colchester’s destruction. The uprising started among the Iceni, the tribe that lived to the north of Colchester in the area of modern Norfolk. The Iceni were not at the time part of the Roman province, but were an independent tribe with diplomatic relations with the Empire. In AD 60 the Icenian king Prasutagus (Fig. 5) died and the Romans decided to incorporate the tribe into the province. Tacitus tells how they proceeded to do this.
Kingdom and household alike were plundered like prizes of war, the one by Roman officers, the other by Roman slaves. As a beginning, (Prasutagus’s) widow Boudica was flogged and their daughters raped. The Icenian chiefs were deprived of their hereditary estates as if the Romans had been given the whole country. The king’s own relatives were treated like slaves. (Annals XIV.31)
Outraged, the Icenian under Boudica rose in arms against the Romans and were joined by many others within the province. Their grievances were numerous, including rapacious tax collectors and the arrogant settlers in the colonia at Colchester who had dispossessed the natives of their land and treated them like slaves. The settlers did not seem to appreciate their danger until too late.
They appealed for help to the imperial agent Cams Decianus. He sent them barely two hundred men, incompletely armed. There was a small garrison on the spot. Reliance was placed in the temple’s protection. Misled by secret pro-rebels, who hampered their plans, they dispensed with rampart or trench. They omitted also to evacuate old people and women and thus leave only fighting men behind. Their precautions were appropriate to a time of unbroken peace. Then a native horde surrounded them. When all else had been ravaged or burnt, the garrison concentrated itself in the team ple. After two days’ siege, it fell by storm. (Annals XIV32)
Colchester was destroyed (Fig. 6) and Boudica’s army swept on. They wiped out the ninth legion in battle and destroyed the other two principal towns of London and Verulamium. Tacitus estimates that 70,000 Romans and provincials were killed at the three towns. The uprising was eventually put down by a pitched battle conducted by the governor with a force of 10,000 Roman soldiers. After the battle Boudica poisoned herself and peace, of a sort, returned to the plundered province (Fig. 7).
The Colchester Glass
The recent excavations in the city center have produced over 500 fragments of Roman vessel glass from layers that either pre-date the Boudican uprising or are contemporary with it (Cool and Price 1995). An equally large group comes from the Sheepen area, which lies about a kilometer to the west of the town and which was destroyed at the same time (Harden 1947; Charlesworth 1985). As the use of glass vessels in Britain prior to the Roman invasion was very rare, we can be sure that virtually all of this material dates to the period between AD 43 and 60.
The archaeological value of destruction like the one caused by Boudica’s army is that it disrupts the normal patterns of disposal. In the case of glass vessels, this appears to have entailed collecting broken vessels so that they could be made into new ones. The recycling of glass is not a new phenomenon, but one with considerable antiquity. Martial, writing slightly later in the 1st century, even includes a disparaging reference to the sort of people who went around collecting broken glass.
Yours is the kind that every gutter hatches, Across the Tiber it is bred in batches And trades in broken glass and peddles matches. (The Twelve Books of Epigrarns, Book 1. ALI) For the early Colchester glass there was little opportunity for this recycling to take place. What we have is therefore a good reflection of what was available between the years AD 43 and 60. The material is useful far beyond Roman Britain, because at that time there were very few regional differences in the types of vessels used across the Empire. The information gathered at Colchester is thus equally valuable for all contemporary sites. Had the destruction taken place thirty years later, the value of the Colchester material would not have been so great. By then, regional styles were developing as glasshouses spread out of the Mediterranean area and into the northern provinces. An abiding impression of the pre-Boudican glass from Colchester is how colorful it was. Many vessels were made in a blue-green shade that can be viewed as the natural color of glass that has not been deliberately colored or de-colorized. Many other vessels, however, were made in shades of deep blue, dark brown, emerald green, and even occasionally purple.
Multicolored vessels were also popular (Fig. 8). These effects could be produced by a number of methods. Cast millefiori vessels were made from segments of glass canes with patterns running through them. Blown vessels could be decorated with applied trails or lumps of contrasting colors. Cased vessels with an opaque white interior and a brilliantly colored exterior were also produced. Very occasionally vessels were painted with colored scenes and fired to produce enameled decoration, but such vessels are rare and only one has been found at Colchester.
The colonists’ tables would have been covered with very gaudy glass. Truly colorless glass was not popular; only a handful of fragments from colorless vessels have been found in pre-Boudican contexts. Colorless vessels did not start to become fashionable until after the colonic was destroyed.
A Glass Shop in the Colonia
An intriguing discovery allows us to catch a glimpse of the sort of shop where the settlers may have bought their glass (Hull 1958). Remains of the shop first came to light in 1927 when workmen were digging foundations for a tea-room in the High Street. Alerted by the discovery, the local museum was able to dig two test pits at the site in 1929, when the neighboring building was demolished.
The main stock appears to have been the glossy red pottery known as Samian ware, but other types of fine pottery, pottery lamps, and glass vessels were also sold. The pottery vessels appeared to have been stacked one on top of the other, upside down, on shelves. Glass vessels had been displayed on higher shelves. Because the shop was destroyed by fire, many of these had melted and dripped onto the pottery vessels below before the building collapsed, crushing everything within. It is tempting to view this destruction as a result of the Boudican fire; however, the shop may have come to its unfortunate end a few years earlier. Samian potters stamped their wares with their names, inadvertently providing valuable dating tools. A study of the stamps on the pottery from the shop suggests that the structure is more likely to date to AD 50-55 than 60 (Millett 1987).
Given the partial excavation of the site and the fact that many of the vessels had melted into amorphous lumps, it is not possible to reconstruct the complete stocklist of the shop. From the fragments that were salvaged and the observations that were made, however, it is possible to suggest some of the items that were sold. These are shown in Figure 9. Cast pillar-molded bowls appear to have been common; examples included blue-green, blue, and millefiori glass. Blown glass included a yellow double-handled cup and small cylindrical cups in various shades of green and amber. The latter are known as Hofheim cups after the site in Germany where they were first recognized in quantity. Vessels of a more utilitarian sort were represented by a blue-green cylindrical bottle.
Typical Boudican Glass Assemblagea
As it includes both very common and relatively unusual forms, the stock in the shop is a good place to start reconstructing what a typical set of glass vessels of Boudican date would have been like. Pillar molded bowls and Hofheim cups are ubiquitous throughout Colchester in the Boudican and earlier deposits. The latter were certainly drinking cups, but quite what the pillar molded bowls were used for is unclear. Some are small enough to have been used as drinking vessels, but many are far too large for that to have been a practical proposition. They may have been used instead for the presentation of food at table. Many of the other cast vessels are bowls or shallow dishes similar to those from the shop (Fig. 9b, c). They also appear to have been used for serving or presenting food.
Glass was certainly a popular medium for drinking vessels. In addition to Hofheim cups, tall beakers with very solid feet and the sports cups discussed below were very common. The double-handled cup from the hopis an example of some of the more elaborate drinking vessels that would have been available. Other shapes known to have been in use elsewhere in the town include the elegant canthari with their high arched handles (Fig. 10).
Glass was also used to make jugs and. similar vessels, including amphorisks (Fig. 11). These tall vessels were often made in brightly colored glass and would have looked very striking on the dinner table. One cannot help but feel, however, that they would have had relatively short life-spans as their walls were very thin and the breakage rate must have been high. This drawback with glass tablewares was well recognized in the mid 1st century AD, as a contemporary author has one of his characters say: “I prefer glass—that’s got no taste at all. If only it didn’t break I’d prefer it to gold” (Petronius, The Saolricon, Book XV.50).
Glass vessels were used not just as tableware, but as containers such as jars, bottles, and small perfume flasks. Food can be stored in glass vessels without any fear that it will be tainted by previous contents. In the mid 1st century it is likely that this property was beginning to be appreciated, and storage vessels become increasingly common. Later in the 1st century we find fragments of glass containers suitable for storing food in large quantities. Such fragments are rarer in the preBoudican levels, but utilitarian glass vessels were already being made in some quantity by the mid 1st century. Glass was no longer confined to the realms of ‘special’ or luxury vessels, but had moved into the everyday world of the kitchen.
The development of glass vessels with very specialized functions was part of this broad trend towards using glass for a much wider range of vessels than it had been at the beginning of the 1st century. Three examples of these specialist vessels from Colchester are shown in Figure 12. The funnels (a) were possibly part of drinking sets for the mixing of wine. Double bottles or jugs (c) have the interior divided into two separate containers. To a modern eye this is very reminiscent of some containers intended to hold oil and vinegar for salad dressings. Glass inkwells (b) may also have been in use at this time. Although these specialist vessels form only a small part of the Boudican glass assemblage, a larger range of them appears to have been available in the mid 1st century than was ever to be available again during the Roman period. Experience was to show that some were much less practical than others, and these fell out of the repertoire.
Souvenirs of Sporting Events
To conclude, we can look at the sports cups, as these cast light on what the leisure interests of the settlers may have been. Sports cups are cylindrical vessels with scenes of sporting events such as gladiator fights and chariot races. These cups were produced by blowing a bubble of glass into a mold which had the scene carved into its inner surface. As the bubble of glass expanded, the sides and base of the cup took on the pattern of the mold. The mold could then be dismantled, leaving a glass cup with a scene of gladiators or chariots running around its sides in relief. This mold-blowing process allowed vessels with complex decorative schemes to be produced repeatedly.
Mold-blown vessels in a variety of types were especially popular in the mid 1st century. At Colchester, however, it seems that the sports cups were the most popular, and at least 13 examples have been found. A complete example (see Fig. I) was found with a cremation burial in the cemetery to the west of the town. It shows two scenes from the Circus at Rome. In the top band there are representations of the architectural ornaments that decorated the long central barrier of the Circus around which the chariots raced. In the lower band four chariots, each with four horses, can be seen. On another example the race scene runs all around the cup (Fig. 13).
The names of the gladiators and charioteers are written above the scenes. These may well have been real people fighting and racing in Rome. Certainly some of the gladiators named on the cups are the same as those known to have been fighting during the reigns of Caligula (AD 37-41) and Nero (AD 54-68). These include Proculus, Columbus, Spiculus, and Petraites. All are featured on one very common design of gladiator beaker of which there are three examples from Colchester (Figs. 14, 15). If it is correct to identify the names on the glass vessels with the historically attested people, the time span implied by the names occurring together on one cup would suggest that these are not souvenirs of particular sporting bouts, but of the sport itself. The equivalent in 20th century baseball might be a coffee mug with pictures of Babe Ruth, Joe Dimaggio, and Cal Ripken, Jr. Across the centuries the sport may change, but the eagerness of fans to have souvenirs of their heroes remains the same.
Nearly two millennia ago, Boudica’s armies destroyed Britain’s first town and inadvertently froze in time the way of life of the mid 1st century colonists. The recent redevelopment of the modern town center has allowed archaeologists to explore that moment.
Here we have used one small aspect of the discoveries— the glassware—to examine the colonists’ tastes in topics as diverse as the color schemes they favored for their dining tables, to their sporting heroes. The full range of information from these excavations will provide many new insights into life in Roman Britain. Many of the different categories of evidence, such as the small finds, the animal bones, and the structural remains, have already been published, and the final volumes are expected next year. When this sequence of publication is completed, the archaeologists of the Roman Empire will have a valuable research tool, and the archaeologists of Colchester will deserve much praise.
Acknowledgments
Much of my work on the Roman glass from Colchester was done in conjunction with Dr. J. Price and was funded by English Heritage. The glass itself has been studied by courtesy of Philip Crummy of the Colchester Archaeological Trust, and Mark Davies and Paul Sealey of the Colchester Museums. I am very grateful to them for all of their help over the years. I am also very grateful to John Davies of the Norfolk Museum Service for advising me about Oceanian coins and providing the pictures of them (Figs. 5, 7). I would also like to thank the Colchester Archaeological Trust, the Trustees of the British Museum, and the Corning Museum of Glass for providing the photographs used in Figures 1, 4, 6, and 14.