The rise and fall of a Roman waterfront
In early May the excavation of a riverbank in the heart of Roman London drew to a close. Waterlogged layers preserved here contained timber buildings and almost 10,000 small finds. Sophie Jackson, Sadie Watson, Angela Wardle, and Michael Marshall of MOLA (Museum of London Archaeology) told Matthew Symonds about the breathtaking archaeology at Bloomberg Place.
In the heart of the City, MOLA archaeologists work against a backdrop of construction as the former Bucklersbury House is redeveloped to become Bloomberg Place. The River Walbrook once flowed from left to right across this site. A 4th-century timber-lined Roman well is visible at the bottom left of the picture. St Mary Aldermary church can be seen just beyond the construction site.
Free from any major prehistoric occupation, the greenfield city site was selected because it lay at the lowest bridging point of the Thames. Despite this blank canvas, any visitors familiar with cities of the Mediterranean world arriving during the early years of Roman London could be forgiven for thinking that the settlement expanding across its two hills left much to be desired. Rather than an imperial showpiece of gleaming marble, the simple wooden houses and stock pens crowding the city blocks could have been plucked from a Wild West town. The structures colonising the sloping, sinuous banks of the Walbrook stream that flowed between the two hills were little different.
Today the Walbrook has been reduced to a trickle dribbling into the Thames near Cannon Street Station, but it was a major landscape feature when the first Roman settlers arrived. Rising near Hackney, the stream cut the new city in half. A substantial timber bridge was built to span its steep valley so that the main east—west road could run uninterrupted through the city. But the stream was not just an obstacle to overcome, it also represented an opportunity.
The tidal reach of the Thames engorged the Walbrook at high tide, allowing boats to unload their wares as far upstream as the road bridge. Now, the MOLA team have exposed astonishingly well-preserved structures just south of the bridge. Excavation revealed the changing face of the city over the course of the Roman period, and has already produced a unique collection of around 10,000 small finds, 900 boxes of pottery, and thousands of fragments of industrial waste.
Despite the area’s prime location within Londinium, when redevelopment was first mooted in 2006 it was assumed that archaeological vestiges would be limited. Soon to be known as Bloomberg Place, the 1950s construction of Bucklersbury House, complete with double basements, was believed to have obliterated any substantial remains. One discovery salvaged from that post-war reconstruction was the famous London Mithraeum. Excavated by Grimes in 1954, and only saved from the bulldozer following public outcry, the temple lay on the east bank of the Walbrook and contained a remarkable group of sculptures. The enthusiasm for London’s Roman heritage this generated did not prevent the surrounding area from being unceremoniously machined away and dumped in an Essex landfill site.
A Roman woven straw basket, just one of many fragile organic objects to have survived intact on the Bloomberg site.
Bloomberg Place was dubbed the ‘Pompeii of the north’ in the ensuing media coverage. It is the breadth of finds made on the site rather than its size that bears comparison with the Bay of Naples town. The artefacts are a closer fit with Pompeii’s less renowned sister site: Herculaneum. Unlike Pompeii, where organic remains only survive as voids in the ash that shrouded the town, the superheated pyroclastic cloud that collapsed on Herculaneum preserved a bounty of charred wooden objects. Such artefacts also survive at Bloomberg Place, although here it was a quintessentially British phenomenon that saved them: the damp.
The foundations for timber buildings that were preserved by the damp conditions.
The early buildings on the banks of the Walbrook offered little in the way of Classical creature comforts. By AD 55-60, rectangular wooden shacks had colonised the stream’s east bank. These were simple to assemble: thick oak planks had mortice holes cut into them, creating baseplates to hold upright wooden staves. Withies woven around these created walls around 2m high, for buildings that appear to have stood a storey and a half tall. Loft accommodation was probably tucked under a wooden shingle roof. They were little more than sheds to modern eyes: rather than trying to lure settlers with luxury living, the emphasis was on throwing up housing as quickly as possible to get the new town populated. The problem of the sloping stream bank was solved by installing wooden box revetments. Packing these with soil created waterfront terraces, and a level platform to build on. In one instance the temporary wooden steps leading down into a box revetment and providing access for those filling it with soil still survived in place.
Despite the focus on speedy construction, time was still found to placate the local spirits. The first act during building was often to lay a foundation offering. In some cases a mortarium, used to grind up the salad leaves popular in Mediterranean cuisine, was carefully placed under the timber baseplate. On another, later, occasion the dedication consisted of a beaker with ring and dot decoration that contained a brooch.
A complete mortarium deposited as a foundation offering beneath the oak baseplate of a building. The holes in the timber are mortise joints cut to hold upright staves. Withies woven between these would have created outer walls up to 2m high.
The rough and ready neighbourhood on the east bank of the Walbrook appears to have emerged from London’s first great catastrophe relatively unscathed. In AD 60 Londinium became the second of three towns to be sacked by Boudicca and her band of Iceni freedom fighters. While destruction has been encountered elsewhere in the City, apart from a few charred timbers that can, with the eye of faith, be linked to rebel attempts to torch the town, there is little sign of damage at Bloomberg. Perhaps the damp environment was the area’s salvation, insulating it from Boudicca’s inferno. Over time, the character of the buildings clinging to the bankside gradually changed. The rectangular timber buildings were replaced by two circular structures. One had fine wattle walling that would have risen to a low dome. Lined with clay and brick internally, the absence of a door suggests that this was a bread oven. The other appears to have been a stock pen. Both emphasise the rudely utilitarian character of the area.
In the early 2nd century the area was redeveloped to create an industrial facility on an unprecedented scale. Huge timber piles were sunk into earth, spearing the waterlogged remains of earlier buildings. These piles supported a massive Roman concrete opus signinum floor, covering almost the entire footprint of the former Slug and Lettuce pub. This new complex would have been two or three stories high, and appears to have featured a water mill. Although the building itself was not unearthed — it probably lay to the north of the excavated area — parts of various cogs, pulleys, and a lantern gear were discovered. Such machinery controlled the direction the mill mechanism moved in, shifting it from horizontal motion to vertical turning. The power of the Walbrook was also harnessed, with a system of culverts and leats controlling its flow.
MOLA archaeologists recording the AD 55-60 levels on the east bank of the Walbrook. The team have to work around concrete piles dating to the 1950s, timber Roman piles dating to the 2nd century AD, and modern service pipes. The natural curve of the Walbrook bank can be seen sloping away from the camera. This uneven building surface was countered by constructing the box revetment visible in the photograph. It was filled with earth to create a level surface.
Intriguingly, huge quantities of burnt grain were discovered when the neighbouring Magistrates’ Court was erected in the 1860s, raising the possibility that grain was being ferried up the Walbrook for processing at the new facility into flour to feed the urban masses. This would fit with the 1,000 or so fragments of quernstones discovered in 1994 just over Queen Victoria Street at 1 Poultry.
This mill complex proved to be the industrial high-water mark of the excavated area, with its economic vitality never recovering after the Roman Great Fire of London. This Hadrianic conflagration, believed to have broken out in the AD 120s, succeeded where the Boudiccan blaze had failed and laid waste to the entire area. By the late 2nd century, in a move foreshadowing 20th-century urban regeneration, the former Roman industrial zone was overrun by des res properties. Partially destroyed by the pub basement, fragments of mosaics, flagstone floors, and porticoes testify to the arrival of fine living. This fits with perceptions of a more suburban character for the late city, and a stunted manufacturing base.
Although the 4th-century ground-level did not survive, activity from the twilight of Roman London was represented by wells sunk into the earlier layers. One of the last datable acts to occur on the site was the formal closure of such a well. In the late 4th century four fine pewter bowls and six cattle skulls were sent down the well. The bowls were presumably quite valuable, while the slaughtering and decapitation of six cattle — assuming they were butchered to order — was also a conspicuous sacrifice. Such an offering can be seen as the opposite of the pottery vessels laid under the foundations of new timber buildings over 300 years earlier. Rather than a bribe to ensure future fortune, these brought the well’s working life to an end. Whether it was because the water had gone bad, or the space was needed for something else, the event was marked with suitable ceremony.
Despite the stunning sequence of Roman structures, in many ways it is the finds that are the stars of the excavation. While some artefacts lay in situ within buildings, the majority came from Roman rubbish pits, landfill or the Walbrook channel itself. A rich cessy aroma created by the slow putrification of organic matter over nearly 2,000 years often alerted archaeologists to the presence of a rubbish pit. The landfill was initially used to level off the box terraces that early buildings were constructed on. Over time, further land reclamation aimed at raising living areas clear of floodwaters saw more and more earth dumped on site. Ultimately this created the astonishing height difference between Roman and modern levels.
A miniature amber gladiator amulet discovered during digging at the Bloomberg site.
The showpiece find, a miniature amber amulet in the form of a gladiator helmet, is an import. Although this object has no direct parallels, the raw material would have been sourced in the Baltic, and there were major amber workshops around Aquileia. This raises the prospect that this much-loved accessory — to judge from the wear marks scored into the loop that its necklace was threaded through — travelled from the edge of the Adriatic before slipping into the Walbrook mud. Less ornate imports include oil lamps from Lyon and high-status tableware. Despite such exotica, it is the detritus of more local manufacturing that proved most surprising.
One feature of the area that the 1950s excavations did not really touch on was the quantity of industrial waste. Thousands of fragments of cut copper-alloy sheet were recovered during the recent excavations, indicating a thriving metalworking industry nearby. Various miscast and incomplete metal objects give an idea of the goods being manufactured, such as enamelled mounts, a military belt buckle, and probable armour, including possible lorica segmentata fittings. A broken mould also indicates that ceramic oil lamps were being fired nearby. They are a type that was fashionable in the decades immediately after the Roman Conquest; similar lamps were made in a pre-Boudiccan workshop at Colchester. Once manufactured, these goods could be transferred directly into boats moored in the Walbrook.
A Roman leather carbatina. Made from a single piece of leather with cut openwork decoration and a seam at the heel, this is one of 250 shoes discovered during the excavations.
The excavation also produced the largest set of Roman leatherwork ever discovered in London, including around 250 shoes. These range from the Roman military caliga to flip-flop style sandals and even cork-soled footwear to insulate feet from the blistering heat of a well-stoked hypocaust. The size and style of the shoes provides a window into demography, offering a new approach to an old question. The early cemeteries of Roman London tend to be well-stocked with adult males, fostering the suspicion that the early town was largely a masculine environment. This footwear, however, reveals the presence of children, suggesting that at least some settlers arrived as families. Offcuts also testify to a flourishing recycling industry, with sheet leather panels trimmed of their seams and transformed into footwear. Fortunately, the finest leather panel from the excavation, the mysterious upholstered object representing an armed figure flanked by hippocamps (see CA 279) survived intact.
Small pieces of furniture including turned wooden legs and spindles are proving difficult to identify, as they only survive in exceptional circumstances. Devoid of ready parallels, team members are scouring wall paintings, funerary sculpture, and coin reverses for hints about what they belonged to. Other fragments, including lion-paw table supports and a fancy cabinet door are more familiar. Given the scarcity of such household ephemera elsewhere, it is likely that the Bloomberg furniture will be of major international importance.
Two of the writing tablets unearthed during the excavations. The text was written in ink on the example to the right, with the only part translated so far saying ‘Dear Januarius, to my dear brother…’. All the other tablets originally held a wax plaque to write on. Although the wax has long since vanished, it is sometimes possible to reconstruct the original text from where the metal writing stylus scored the wood beneath.
It is a different class of wooden objects that holds the greatest promise to bring the past to life: 350 fragments of writing tablets were recovered during the excavation, a major haul. All bar one are wax tablets, where text was scratched on with a metal stylus. It remains to be seen how many are legible. The exception, an ink tablet similar to those found at Vindolanda fort, is also the only one to have been partially translated so far. It opens, tantalisingly, with ‘Dear Januarius, to my dear brother…’. The wax tablets may well prove more formal in tone. Examples from elsewhere in London are legal documents recording the sale of a wood in Kent and a French slave girl called Fortunata. A cache of eight tablets, complete with styli, was found dumped in the corner of a room dating to around AD 80. Is this the earliest intact archive from London?
It will be several years before study of this unique assemblage is complete. What is known so far represents only the tip of the iceberg, and we can be confident that these finds will revolutionise our knowledge of Roman London.