12

Episodes of renewal (c. ad 90–110)

Later Flavian London

The pace of change slowed somewhat after London’s rapid early Flavian growth. No new roads are known to have been created between c. ad 90 and ad 120. Those that existed were sufficiently busy to warrant repair, but at longer intervals than previously, such that the ‘decumanus’ beneath Cheapside was only resurfaced once between its reinstatement after the Boudican revolt and the Hadrianic period. ¹ This may reflect on changed circumstances. In ad 87, shortly after Agricola’s victory at Mons Graupius, some troops were withdrawn from the province to strengthen the Danube frontier against Dacian invasion. ² Britain is thought to have received new supplies of low-denomination coins of Domitian at about this time. ³ Guy de la Bédoyère identified a concentration of these coins along the Thames foreshore at Billingsgate, which he identified as losses from supplies of coin shipped to London for provincial dispersal. ⁴ Since troops facing redeployment would have needed small-change to arrange transport, lodgings, and local purchases, London may have been one of the places where this coin was most needed.

Some buildings were damaged by fire towards the close of the first century. Properties at One Poultry, beside the Walbrook crossing, were burnt around the 90s, as were houses at opposite ends of the settlement at 14–18 Gresham Street and 60 Fenchurch Street. ⁵ A building behind the forum, at Leadenhall Court, was also incinerated in the late Flavian period. ⁶ South of the river buildings along the Watling Street were destroyed in a late Flavian fire and a coin of Domitian (ad 81–96) was recovered from pre-fire deposits at 22 Borough High Street. ⁷ These traces of destruction are unlikely to have been the product of a single conflagration, since they were widely separated and many houses were unscathed. It is likely that they evidence a series of isolated house fires, largely restricted to single properties, but we cannot dismiss the possibility that some damage was caused by incendiary unrest along London’s principal streets.

What is more certain is that there was a pronounced decline in the volumes of Samian pottery reaching London at the end of the first century. ⁸ Since these imports serve as a proxy for other goods it seems likely that there was a hiatus in port activity. There are also instances of properties left undeveloped around this date and a decrease in building density at Bow Bells House on the west side of town is dated c. ad 100 from associated pottery. ⁹ If there were problems in London towards the end of Domitian’s reign they were neither persistent nor universal. A new portico was built around the forum’s central courtyard using ragstone foundations that hint at a date after c. ad 85, and brick piers witness a later phase of alteration to its south wing. ¹⁰ The addition of covered areas in front of the tabernae-style units around the forum would have benefitted shippers moving goods around the complex. The east gate of the amphitheatre was also rebuilt, using timbers that suggest that work was underway c. ad 91 perhaps involving repairs to the timber superstructure and cavea seating. ¹¹ A new bridge may have been built across the Thames: a timber box-structure, about 5 by 7 metres across, found on the Thames foreshore at Pudding Lane in 1981 was probably a pier-base for a temporary replacement of the fixed crossing. ¹² The timbers used in its construction were dated ad 78–123, but it was in certainly place before adjacent quays were built by c. ad 98. ¹³ The Neronian warehouses and workshops immediately up-river of the bridge, investigated at Regis House, were also restored and enlarged. ¹⁴

Waterfront renewal under Nerva and Trajan

Towards the close of the century, work began on a major overhaul of the port. ¹⁵ Construction started down-stream from London Bridge where a new timber-fronted embankment subsumed the Flavian landing stage (Fig. 12.1). Two tile-roofed masonry warehouses separated by a narrow passage were built facing the river. Each measured about 25 metres long by 6 metres deep, and was divided into five timber-floored open-fronted bays secured by removable wooden shutters. Their design was similar to the warehouses built upriver of London Bridge some 30 years before (p. 101). Gustav Milne suggests that this new quay was built c. ad 90, since it employed construction timbers felled ad 86–98. Timbers used to build associated warehouses had, however, been felled no earlier than ad 94. Since these were an integral part of the new waterfront it is likely that the improvements were all undertaken ad 94–8. ¹⁶ Wider urban renewal at this date is hinted at by excavations at the Bloomberg headquarters, where buildings fronting onto the main road were all replaced during the 90s. ¹⁷ The Neronian waterfront up-river from London bridge wasn’t replaced until a few years later, but a vulnerable section was buttressed with timber felled no earlier than ad 95.

Fig. 12.1 London’s port c. ad 95 after the construction of the new warehouses downstream of the bridge (ILA79, PEN79, PDN81, FM085: after Milne 1985, Brigham 2001b, and Brigham and Watson 1996). The bridge itself is shown in the short-lived arrangement represented by the timber pier-base found at Pudding Lane. Earlier and later iterations of the fixed crossing may have been located immediately upstream of this location. Drawn by Justin Russell.

The decision to rebuild may have been influenced by a reduction in the Thames’ tidal range, evidenced by the fact that later Roman waterside structures were built at levels previously inundated at high-tide. ¹⁸ Waterfront advance would have placed the quays into deeper water. On the other hand, the main effects of tidal regression may not yet have been in evidence. Tides are thought to have powered a mill built at the mouth of the river Fleet within the period ad 76–99, and a colony of acorn barnacles (Balanus improvisus) attached to early second-century revetments at Regis House show that the river remained brackish. ¹⁹ What is certain, however, is that the enlarged waterfront terrace added space for handling cargoes and vehicular access, and new warehouses increased secure storage capacity. The constructions downstream of London Bridge suggest that this capacity was targeted on shipping arriving upriver, seeking to dock close to London’s forum.

These building activities merit considering within their historical context. We have no reliable sources for events in Britain in the decades after Agricola was recalled, but an inscription in Rome indicates that a veteran colony was established at Gloucester during Nerva’s short reign ad 96–8. ²⁰ Archaeological evidence suggests that the colony at Lincoln was also established around this time. ²¹ These new foundations on the sites of redundant legionary fortresses were the gift of imperial patronage. Nerva’s involvement is an interesting one. Like Vespasian before him, he had no dynastic right to succession and relied heavily on the support of the army. Investment in Britain’s garrison was a prudent strategy, and would account for land-grants to veteran colonists and spending on the infrastructure of supply through London’s port. It might also account for the presence in Britain of unusually high numbers of a rare coin-type issued by Nerva showing Neptune holding a trident and a ship’s prow. ²² One of these was found by Charles Roach Smith near London Bridge where a shrine to Neptune may have stood (below p. 230). ²³ Other coins of Nerva, issued c. ad 97, employed the legend annona August(i) and showed annona as the personification of Rome’s grain supply. ²⁴ These rare types followed a model set by coins minted under Nero ad 63–7 and Domitian ad 85–6. All three annona issues were minted at dates when new quays were built along London’s waterfront. There is no suggestion that these coins were minted to celebrate building works in London, but there may have been an underlying connection between the imperial intervention in annona supply celebrated on the coins and the commissioning of new harbour works in London. In a similar vein, it can be observed that the warehouses in London were contemporaries of Nerva’s warehouses in Rome (the Horrea Nervae), one of only two public works in Rome known to have been the product of his benefaction. ²⁵ London’s new waterfront may well have been built as part of a considered programme of investment in annona supplies under Nerva’s administration.

This programme of waterfront renewal wasn’t completed until somewhat later, when the Neronian quay at Regis House was eventually replaced. The new revetments, built 4.5 metres further onto the foreshore, employed a pile foundation with a final sapwood ring dated to ad 101, matching a date of c. ad 102 obtained from alterations to waterfront buildings a short distance down-river. ²⁶ This suggests that the works were undertaken under Nerva’s heir Trajan after his triumphant conclusion of the first Dacian war. Projects initiated at this time included harbour works in Rome, dated ad 101–3, and improvements to the harbour infrastructure at Ostia and Portus to provide safe anchorage for transports associated with the grain dole. ²⁷ London may have similarly benefitted from investment in the annona. Changes in the scale and direction of supply at this time were marked by the arrival of new imports of Samian from the central Gaulish kilns of Les Martres-de-Veyre following a decline in imports from the southern Gaulish kilns at La Graufesenque. ²⁸ Work is also likely to have started on the construction of London’s massive new forum complex, as described in the next chapter.

Other public works

The aisled building opposite the forum, at 5–12 Fenchurch Street, was also refurbished early in the second century. One room was painted with an architectural illusion featuring a standing woman with arms aloft: a pose typically used in prayer. ²⁹ An adjacent room furnished with amphorae and flagons may have been a kitchen. The suggestion that this building was the meeting place of a collegia or guild has its attractions. Roman cities relied heavily on such guilds to function from the early second century onward. ³⁰ In Chichester a collegium of fabri (builders) was sufficiently important to have dedicated a temple to Minerva and Neptune and we can assume the existence of similar institutions in London (see p. 269). ³¹ Collegia functioned as ‘brotherhoods’, establishing an institutional framework for members to further shared business interests. This was important at port cities where transient communities had need of such patronage networks. They also compensated for any under-development of other civic institutions, mediating between the provincial authorities and local community. Collegia scheduled regular feasts and banquets, and were cult associations figuring prominently in public ceremonies and processions. This would have included devotions to the imperial cult. The praying lady painted on the wall of the building at Fenchurch Street would have been at home in such a context.

A few years after the rebuilding of London’s port the Blossom’s Inn waterworks were improved. ³² A new timber-lined well was built using timbers felled ad 104, and another c. ad 108–9. This was a larger feature, some 3.6 metres square, and its fills contained the remains of a water lifting device formed from two heavy wrought iron chains linked to a series of open-topped, oak-board box-buckets that carried water (Fig. 12.2). The bucket-chain was discarded in the well after the destruction of the well-house in the Hadrianic fire alongside several complete box-flue tiles and roofing-tiles, three bearing procuratorial stamps, that may have come from the nearby Cheapside baths. This was a different device to the one found in the Neronian well. The earlier mechanism relied on tread-wheel-powered box-chains with side-ports, but the new one involved bucket chains using capstan and gears. This improved water-lifting technology reached London at the start of the second century. The chain would have held at least thirty box-buckets, each containing approximately six litres of water when full, capable of lifting over 72,000 litres (16,000 gallons) of water over a ten-hour operating day. It is probable that the main demand for this water came from London’s thirsty bathhouses, including those at Cheapside and a proliferating number of private establishments. ³³ The Flavian baths at Huggin Hill also saw several major alterations in the Trajanic period, possibly drawing on the same improved water-supply. Works included the insertion of a large timber drain around the building, the addition of second double-apsed hot room (caldarium) onto the northern side of the complex, and a further eastern suite of rooms built over the probable site of an earlier palaestra which included another large new caldarium (Fig. 11.3, p. 141). ³⁴ The interior was decorated with a range of imported marbles. The new rooms increased the capacity of the building, perhaps allowing for social and gender segregation. Curiously, however, the baths fell out of use soon after these improvements. Interim reports describe a rapid abandonment, involving stripping reusable fixtures and fittings from the furnace rooms and service areas. Trajanic pottery was associated with the subsequent closure of the larger caldarium on the western side of the building. Earth-walled buildings were eventually set within the disused buildings, whilst the main vaulted sections of baths may not have been demolished until the late second century. One can only imagine that the building developed a structural fault that left it unfit for use, or was associated with a community that had departed London.

Fig. 12.2 A reconstruction of the water lifting device from Gresham Street (GHT00), showing the operation of the tread-wheel-driven Type-A bucket-chain (drawing by Bob Spain and reproduced by kind permission of Museum of London Archaeology).

At some point in the late first century, a Romano-Celtic temple was built between the amphitheatre and the Blossom’s Inn pool (Fig. 12.3). It had long been recognized that masonry walls in this area were not orientated on the local street-grid, but not known why. ³⁵ It can now be suggested that this unusual alignment prevailed because these structures were built within a temple precinct established around the hillside springs (above p. 77). The corners of Romano-Celtic temples were commonly aligned on the cardinal points of the compass, disregarding the orientation of neighbouring streets. ³⁶ Although the springs here may have been the focus of earlier attention, the earliest temple was built in the late first century. ³⁷ This consisted of a square central cella surrounded by a colonnade with pier-bases clad in Purbeck marble. ³⁸ Most temples in southern Britain were built to this Romano-Celtic design, and precincts containing clusters of such buildings have been found on the borders of several towns. ³⁹ As has already been noted, the Cheapside baths may have been associated with this temple complex.

Fig. 12.3 Temples and baths, with associated waterworks, to the south of the amphitheatre. The northern part of this figure involves a considerable amount of conjecture and is based on incompletely published interim plans (after Blair et al. 2006). Drawn by Justin Russell.

A bathhouse was also built in Southwark close to the road to London Bridge, near the market hall and bakery at Borough High Street (above p. 80). Four rooms of this masonry structure were discovered in excavations for the Thameslink railway. ⁴⁰ A coin of Domitian of ad 86 was recovered from construction deposits, and the building materials are consistent with a late first- or early second-century date. The structure contained a circular room, about 5 metres across, that may have been a sweat room (sudatoria), dry heat room (laconicum) or sunken plunge bath. ⁴¹ Too little of the building was found to establish its scale, but it is unlikely to have rivalled the larger bathing complexes north of the river. These baths may also have been associated with a temple precinct, mirroring arrangements at Cheapside. A roughly square masonry walled building, found 50 metres east of the baths at 25 London Bridge Street, may have been a Romano-Celtic temple. ⁴² Little information on this building is presently available, and its identification is far from certain, but a statue-base dedicated to Silvanus by Publius Fabius confirms the presence of a shrine in the area. ⁴³ Silvanus was a pastoral deity whose cult appealed to men from the lower economic classes within Roman cities. ⁴⁴ A late first-century flagon found nearby in Tooley Street in 1912 carried a scratched graffiti ‘Londini ad fanvm Isidis’ identifying it as vessel that belonged to a shrine of Isis. ⁴⁵ The cult of this Egyptian goddess was popular with administrators and officials along Rome’s frontiers. ⁴⁶ The survival of the vessel suggests that it was buried intact, probably as a structured deposit within a well or ritual shaft, and that a nearby Iseum shared the ritual landscape with the worship of Silvanus. A sacred site at this location would have faced the incoming tides, where cult practices could have used the rising water to replicate the Nile floods. The topography of this area finds parallel in a shrine to Silvanus at Ostia that was reached through a bakery, and which included images of the deified Augustus, Isis, Harpocrates, Fortuna, and Annona amongst others. ⁴⁷ The Ostian bakers negotiated with these gods to secure the harvests and supplies of annona grain. Similar concerns may have contributed to the development of a religious complex behind the market-hall and bakery on the road to London Bridge.

London may now have had five or more bathhouses, including two public baths flanking the mouth of the Walbrook (at Huggin Hill and Cannon Street), and three smaller roadside establishments alongside routes into town (at Cheapside, Borough High Street, and Fenchurch Street). This was not excessive. London was larger than Pompeii, which had at least four public bathhouses at the time of Vesuvius’ eruption. ⁴⁸

Extending the city

London is likely to have been defined by a formal boundary, separating and sanctifying political life within the settlement and expelling polluting activities. These limits were redrawn as the city expanded. London’s first boundary probably reproduced the slighted defences of the Claudian army encampment, and early Flavian expansion may have resulted in the construction of earthworks to define a northern perimeter (above p. 127). At some point in the early second century, the ditch that marked the Flavian limits was levelled. ⁴⁹ The western limits of the city are not so easily recognized. It is likely that London was undefended for much of the time, and boundary markers are difficult to trace on the ground. The topography suggest that an early boundary existed close to the east end of St Paul’s Cathedral since burials are not found east of this line and the street grid did not extend to its west. ⁵⁰

By the end of the first century, London’s western boundary might have followed canalized stream channels between St Paul’s Cathedral and the river Fleet. Much remains uncertain about the history and purpose of these features, which diverted small streams that flowed into the Fleet south to the Thames. ⁵¹ Two large channels near the source of these streams were recorded north of Newgate Street in 1907–9. ⁵² They appeared to predate the masonry city wall, suggesting that they were here before c. ad 200. This finds confirmation in the presence of late first- and second-century pottery in the fills of a ditch at Paternoster Square into which they were directed. ⁵³ The eastern channel was described as a boundary ditch, with artificially cut sides 12.2 metres wide and 5.3 metres deep. A 2.45-metre-thick monumental ragstone wall was inserted along its east side. There is little doubt that ditch and wall were Roman, and a late first- or early second-century date is likely. The scale of these features suggest the presence of an important public construction close to the settlement boundary. Despite its location, the structure doesn’t appear to have been a town wall, and it remains one of Roman London’s more enigmatic features. The engineering involved in redirecting the streams towards the Thames must have been born of some significant need, resulting in the excavation of channels that were both larger but more irregular than the other boundary ditches known from Roman London. One possibility is that the water was being channelled towards mills by the Thames.

A large ditch may have marked the northern boundary of the settlement west of the Walbrook. In excavations at St Bartholomew’s Hospital, some 25–33 metres to the north of the later town wall, a 6-metre-wide U-shaped ditch a little over 1 metre deep, was traced east-west for a length of c. 100 metres. ⁵⁴ This was apparently dug after trees had been cleared in the early Flavian period, and contained Hadrianic and Antonine pottery within its fills, before being recut as a smaller feature, probably in the late second century. This differed from the large V-shaped ditches that defined major boundaries in London prior to ad 85, possibly indicating that these earthworks were later and differently engineered.

Cemeteries and tombs

Most Claudio-Neronian cemeteries remained in use throughout the Flavian period. New cemeteries were also established further into the countryside flanking roads radiating from the city. Several of the main graveyards of the later city started in use at this time. Four areas of burial can be described, each the result of later expansion agglomerating smaller Flavian plots. ⁵⁵ The cemetery east of town developed around the Flavian road to Shadwell (p. 123), especially after c. ad 120. These graveyards eventually extended over some 12 hectares, from which more than 140 cremations have been recovered. ⁵⁶ The western cemetery expanded to cover an area of 24 hectares, in an arc from Holborn to Aldersgate. Roadside burials also flanked Roman Watling Street, extending more than 1.5 kilometres west of the city to Endell Street where a cylindrical leaden cist containing cremated bones and coins of Vespasian was found. ⁵⁷ Smaller burial plots included a group of cremations found west of Fleet Street in Shoe Lane in 1927. ⁵⁸ Roman Ermine Street (modern Bishopsgate) also attracted cemeteries, although perhaps not until later. Several other cemeteries were located on the mainland to the south of Southwark.

Cremation was initially the preferred burial rite. Ashes were placed in vessels, cists and casks similar to those used in earlier cemeteries, sometimes accompanied by other vessels with food to sustain the deceased on their journey. ⁵⁹ A couple of pits within which bodies were cremated (bustum) have been found, as well as spreads of redeposited pyre material. ⁶⁰ The pyre, rather than the grave, was the main focus of funeral rites. The debris suggests that corpses were often clothed, wearing dress accessories, and had sometimes been cremated on nailed wooden biers. ⁶¹ Finds from these and other cemetery assemblages indicate that meals took place at the funeral and visits to the grave. Offerings, mainly pig and chicken, are present within half of the cremations: both as cremated materials from the pyre and unburnt. ⁶² Handfuls of lentils, a token of mourning, were thrown onto some pyres. ⁶³ A preponderance of drinking vessels, rather than cooking or serving dishes, suggests that libations were more widespread than feasts. ⁶⁴ These funerals were important occasions, leaving burial grounds well-tended. ⁶⁵

A few tombs were set within imposing masonry mausolea. A sculpture of Cotswold limestone buried in a late first- or early second-century roadside ditch at 24–6 Minories portrayed a half life-sized eagle with spreading wings that held a writhing serpent in its beak. ⁶⁶ This personification of Jupiter was a fitting subject for a funerary monument. It declared the victory of celestial power over the chthonic forces of death represented by the serpent, whilst alluding to the supremacy of Roman imperial power. The piece was shaped to fit into an alcove or niche within a tomb and its style suggests a Flavian or Trajanic date, indicating that the monument only survived for a generation or two before demolition. It would have been removed from a tomb in London’s eastern cemetery where the foundations of several mausolea have been found. ⁶⁷ Fragments of wall veneers, including Purbeck and Carrara marble, dispersed amongst later burials in this area would have decorated these structures. Other fragments of architectural stonework from monumental tombs were reused in fourth-century towers added to the town walls and include the effigy of a Flavian soldier and a lion-and stag sculpture found at Camomile Street. ⁶⁸ The female head of a small sphinx, also carved in Cotswold limestone, was found in a second-century pit at Drapers’ Gardens. ⁶⁹ Blocks of rusticated purple sandstone reused in the footings of the west gate of the Cripplegate fort are likely to have derived from a monumental Flavian tomb that stood in the western cemetery area. ⁷⁰ A female figure pouring water from an urn, probably Venus or a nymph, found at Skipton Street in Southwark may have derived from a Hadrianic funerary statue on the southern approaches to the settlement. ⁷¹

These funerary monuments along the roads that converged on London proclaimed the status and Roman identity of successful individuals whose achievements mattered after death. ⁷² Those who merited extravagant commemoration tended to be outsiders, particularly with a background in military service. It is also notable that some tombs were conspicuously more monumental than contemporary houses. This suggests that we are looking at the graves of a few persons of rank, whose heirs were anxious to retain a status that could not be taken for granted and whose marginal status limited other options for political participation. This uncertain inheritance might explain why some tombs were demolished relatively soon after their construction.