11
The Flavian forum
At the time of Vespasian’s death in ad 79, London was the most important seat of Roman power in Britain, but it remained architecturally impoverished when compared to leading cities elsewhere in the empire. Many ancient cities were richly endowed by wealthy patrons following the practices of civic euergetism described by Arjan Zuiderhoek. 1 Britain lacked an equivalent tradition of using monumental architecture to frame political life, and social competition in London may also have been stunted by the arrangements made for the government of a city of uncertain legal status.
London’s urban status was perhaps acknowledged in a writing-tablet found in a context dated ad 65/70–80 at the Bloomberg site. This registered a financial agreement affecting one Atigniomarus who came to the city, termed civitatem, on the eighth day before the Kalends of January (25 December) of an unknown year. 2 The reference implies that London was a place with autonomous urban status. London’s identity as a civitatem did not, however, require it to have the political institutions normal to peregrine cities. Military officials and imperial legates sometimes retained the powers elsewhere invested in city magistrates, perhaps routinely so during the transition from military to civilian rule in conquered territories. 3 Another writing tablet found at the Bloomberg site suggests that this may have been the state of affairs in London ad 76. It preserved the preamble to a preliminary judgement, dated 22 October ad 76, made by a judge appointed by the emperor. 4 The appointment would have been made on the emperor’s behalf by the governor, Julius Frontinus, and implies that the city lacked its own elected magistrates. Whilst it was ultimately in Rome’s interests to empower the political institutions of local self-rule, these came slow to London and were apparently still absent in the autumn of ad 76. London’s public architecture has been used to suggest that it gained the instruments of a self-governing community soon afterwards, perhaps during the term in office of the governor Agricola c. ad 77–85, although we have no certain evidence that this was the case.
London’s first public works had been utilitarian engineering projects associated with moving goods through the port and forum, followed by the construction of heated baths similar to those favoured by military communities on the town’s borders. The projects put in hand during Vespasian’s reign introduced new concepts of urban order around the city, as represented by the formal ceremonies of the amphitheatre and a controlled network of roads and quays administered from sites such as the Southwark mansio. This early Flavian architecture enveloped the Neronian city and attracted new areas of elite housing around its borders. Initially the town centre was little changed, but this area was transformed by ambitious building projects later in the Flavian period starting with the construction of a new forum.
In most Roman cities political and social life revolved around the forum, where a principal basilica housed the offices of the civic administration. The decision to build London’s first forum and basilica was consequently an important matter. Our knowledge of this complex, built opposite the T-junction at the city centre, rests on Peter Marsden’s synthesis of the results of excavations along Gracechurch Street (Fig. 11.1). 5 The forum was twice as long as wide, at 104.5 by 52.7 metres, and dominated by a basilica at its further (northern) end. Its walls were built of concreted ragstone blocks separated by horizontal string-courses of red tiles, raised over flint foundations. External buttresses probably supported engaged columns for architectural emphasis. The basilica was a large aisled building, some 44 metres long by 22.7 metres wide, raised on a 1-metre-high platform, with a central nave some 8.38 metres wide flanked by aisles of unequal width. Cross-walls at the east end of the nave might have supported a raised dais to seat magistrates, and a sunken room at the west end of the north aisle could have provided a secure treasury. 6
Fig. 11.1 The Flavian forum and aisled building at Fenchurch Street (after Marsden 1987 and Hammer 1987). Drawn by Justin Russell.
The other sides of the forum were enclosed by narrow ranges of rooms suitable for use as offices, stores, workshops, and shops, replacing the earlier tabernae around the forum hardstanding. Those in the southern range were set behind a portico facing the street, while the east and west ranges probably looked inwards. The southern façade incorporated a portico of columns supporting a tiled roof, with a main entrance at the junction of Gracechurch Street and Lombard Street. The enclosure of the forum adds to the sense that this was a closely managed space. 7
The forum was probably built in the late 70s or early 80s. Excavations at 83–7 Gracechurch Street in 1934 found that the basilica’s south wall cut through a pit containing Flavian pottery and a coin of ad 71, and rubbish beneath the street east of the forum contained Southern Gaulish Samian no earlier than ad 75. 8 This probably pre-dated the forum’s construction, although this is not certain. These finds suggest that London’s forum was roughly contemporary with the basilica at Verulamium, dated to ad 79 or ad 81 by a dedicatory inscription, and Silchester’s timber basilica which has a terminus post quem provided by a coin of ad 77–8. 9 These buildings must all have been erected with official sanction but were differently designed, indicating that different architects and builders were involved. A political context for the works is found in Tacitus’ description of how the governor Agricola gave private encouragement and public aid to building temples, fora and houses during his second winter in post (i.e. ad 78/79). 10 London’s Flavian forum may have been a product of this Agricolan policy, and if so is likely to have been commissioned c. ad 79 onwards.
Scholars have long been attracted to the idea that the building of London’s forum accompanied the grant of an urban charter, perhaps elevating the city to the status of a municipium. 11 The argument is consistent with wider evidence for political reform during Agricola’s administration, but there is no certain relationship between the construction of a basilica and the award of chartered status. 12 Although Flavian London was the largest city in Britain its forum was unusually small. Verulamium’s extended over three times the area, with a basilica almost three times as long. 13 There are several possible explanations for this. Peter Marsden has suggested that London’s forum was the first to be built, establishing a prototype improved-on by neighbouring cities. 14 London’s forum was planned after tabernae had already been set along London’s waterfront, reducing the scale of storage facilities required. We can also question whether it shared equal status to the buildings in neighbouring cities. John Wacher argued that it might be better identified as the seat of the procurator or a market building. 15 Although usually considered a civilian administrative building, it could have been dedicated to the management of official supplies. In this case the courtyard and surrounding rooms might have been modelled on the offices of trading associations found in port cities. The Piazzale delle Corporazioni at Ostia presents a similar layout to the London forum, although lacking the dominating basilica, where the large rectangular courtyard was enclosed by a colonnade concealing rows of small rooms which housed the offices of guilds, merchants and shippers. On balance it seems likely that London’s basilica was built to house administrative facilities, and the forum to accommodate business interests around the open piazza.
Public architecture in the forum district
In 1934, Frank Cottrill found the foundations of a building likely to have been a small classical temple, 20.7 metres long, set within its own precinct west of the forum. 16 Its walls were formed of roof tiles laid with the flange outward to resemble brick, perhaps for a stucco render, set over flint foundations similar to those used in the forum. Peter Marsden has suggested that the building was approached by a low flight of steps, leading to a south-facing façade of two substantial columns forming a porch (pronaos) into the chamber (cella) at the back of which a polygonal apse could have housed a cult statue. These are the characteristics of the distyle in antis façade used in temple architecture. Gravelled surfaces around the building were enclosed by a precinct wall. 17 This enclosure was clearly separated from the adjacent forum, in a layout that faintly echoed arrangements in fora in Gaul and Spain, although on more modest scale. 18 This provided a setting for bargaining with the gods through vows and sacrifices made at altars on the temple steps. 19
We don’t know who was worshipped here, but a temple to the imperial cult was probably established hereabouts. An incomplete monumental inscription found at Nicholas Lane in 1850, half way between the forum and Thames, appears to have included a dedication to the divinity (numen) of the Emperor by the province of Britain. 20 The formula used in this inscription is an unusual one, and its expansion uncertain, but it implies a Flavian date. Flavian patronage emphasized the imperial cult, and collective practices organized around the figure of the emperor would have reinforced London’s identification with the colonial project, solidifying the hegemony of the Roman state. 21 Loyalty to the imperial person and symbol offset the risks of fragmentation posed by the power of Rome’s frontier armies. The temple dedicated to Claudius at Colchester has been identified as a meeting place of the provincial assembly (concilium provinciae), where leading members of Britain’s different communities could celebrate the sacred games associated with the imperial cult. 22 But by the 80s it was possible for more than one city in a province to perform such a role. We have already noted the importance of the imperial cult to the celebration of games in the amphitheatre. The forum would similarly have been dominated by the imperial presence, reinforced by sacrifices on days set aside for the veneration of the imperial house in front of statues and images of the imperial household. Prayers would have been organized for the guardian spirit of the emperor (the genius Augusti) and his divine essence (numen). The religious processions organized on these occasions are likely to have started by collecting the images of the emperor from the temple where his divine spirit resided, and culminated in the games held at the amphitheatre. 23
The construction of London’s forum and temple prompted improvements to the surrounding streets. In addition to a new road built against the east side of the forum, the ‘decumanus’ was re-established c. ad 85 after the vestiges of the Neronian fort were finally cleared. 24 The line of this road was eventually extended eastwards, with traces of gravel metalling observed in Pepys Street. 25 New properties were also established alongside a north-south street set over the site of the fort. Several important buildings were built south of the forum. Excavations in Fenchurch Street uncovered the corner of an unusual aisled building with internal pier bases and external buttresses (Fig. 11.1). The timber and brickearth walls were built over distinctive foundations formed of bands of gravel and crushed mortar laid horizontally in a deep trench. 26 Partitions subsequently divided the narrow aisles into smaller rooms, one of which contained a latrine set over a brick-built drain flushed by running water. Friederike Hammer’s interim report suggests that the building was Flavian, although an earlier date has also been proposed. 27 The unusual features of the building have resulted in suggestions that it might have been a meeting place for a guild (collegium), or a market hall (macellum) involving booths set within the aisles. Traces of another important building were found a short distance to the south, at 55–8 Gracechurch Street. We know little about this large south-facing winged building, but it was supposedly a late first-century construction containing a hypocaust of tile pilae set between parallel ragstone foundations. 28
An even more imposing public building was built over terraces overlooking the Thames on the east side of the Walbrook (Fig. 11.2). Peter Marsden’s report on excavations at Bush Lane between 1961 and 1972 argued that these were the remains of a Flavian palace, possibly that of the provincial governor. 29 Its foundations employed such strong cement that explosives were needed to clear the site when Cannon Street Station was built in 1868. 30 A distinctive feature was an enormous apsidal-ended pool, over 10 metres wide and up to 55 metres long, which could have held 900,000 litres of water. Its base was set over concrete foundations over 1.8 metres thick, and a small apse on its northern side contained a rectangular masonry base where a statue may have stood. This pool occupied an open terrace in front of a monumental building dominated by a massive central hall with 3-metre-wide walls and a concrete floor. East of this hall was another massive room with a projecting apse, north of which hypocaust floors were found. Foundations found beneath Cannon Street Station in 2008 may derive from a similar apsidal chamber to the west. 31 The complex may have extended 140 metres along the river terrace towards the Walbrook, as implied by the presence of a hypocaust floor and monumental wall faced with thin slabs of Purbeck marble at 3–7 Dowgate Hill. 32
Fig. 11.2 The monumental building complex in the area of Cannon Street station (the ‘governor’s palace’) at around the time of its construction c. ad 84/85 (after Marsden 1978, Spence and Grew 1990, Milne 1996, and Brigham 2001a). The difficult circumstances of excavation leave several areas of uncertainty over the location and date of individual features. Drawn by Justin Russell.
A separate range of masonry rooms occupied a lower terrace south of the pool. Marsden suggested that these were the palace’s residential wing, but subsequent work shows that they belonged to a later and probably unrelated development. Excavations beneath Cannon Street station in 1988–9 found that the tile-built terrace wall included curved and squared buttresses forming a decorative river façade. 33 The terrace below this wall was originally an open quayside. Excavations to the east, at Suffolk House, found that these quays were built using timbers felled spring ad 84. 34 Since the river façade was an integral part of the original design these timbers may date the entire monumental complex. Further high-status rooms lay east of the monumental pool, where Marsden suggested that the administrative offices of the palace were located. These too are likely to have been part of an unrelated development.
The argument that this public architecture was part of the Governor’s palace rested on the assumption that the building contained high-status residential quarters as well as monumental state reception rooms. This no longer appears to have been the case, and Gustav Milne’s review of the evidence found little to recommend the palace hypothesis. 35 Several alternative interpretations have been advanced, but the most convincing is that this was a bathhouse. 36 If so, the ornamental pool would have been the focal point of a palaestra, or exercise court. 37 The massive Cannon Street ‘hall’ might have been a large cold or changing room (apodyterium), giving access to separate suites of heated rooms either side. These would have been laid out in the monumental ‘imperial’ style involving cross-axial symmetry that came into fashion at this time. Altogether the complex occupied an area in excess of 10,000 square metres, not much smaller than the Baths of Titus built in Rome a few years earlier.
The vast quantities of fuel required to heat the baths would have arrived by river, landing at the new riverside quays. Water was piped to the site from the north and a sunken masonry structure found at 119–21 Cannon Street may have been a reservoir built for this purpose. 38 A supply-pipe, represented by a row of iron connecting collars, was set into the western roadside ditch of the Flavian road at The Walbrook and Cannon Place. 39 These supplies were probably arranged in the mid-80s drawing from the upper Walbrook where an oak water-pipe was recorded at the Bank of England in 1933–4. 40 Excavations at 6–8 Tokenhouse Yard, just north of the Bank of England, found a narrow gravel-surfaced alleyway built to gain access to the Walbrook outside the northern boundary of the Flavian town. This was flanked by fences built with timbers dated to ad 83 onwards and contained a drain built with a timber felled in the spring of ad 85. 41 These alterations may have been associated with the hydraulic engineering that supplied the baths and other buildings between the forum and Thames. Water engineering of a broadly contemporary date is evidenced by the presence of a wooden water-pipe set alongside the ‘decumanus’ re-established over the site of the Neronian fort, recorded at Plantation Place where it was dated ad 70–80 from associated ceramics. 42 This supply was aimed in the direction of the town’s east gate, perhaps supplying a replacement of the Neronian bathhouse (above p. 77).
Although it is difficult to sustain the argument that the monumental remains at Cannon Street were those of the governor’s palace, it is plausible that palatial residences occupied the river terraces either side of London Bridge where buildings included luxurious mosaics, hypocausts and wall-paintings. Some houses saw significant Flavian investment, such as a masonry structure with flint foundations found at Suffolk House decorated with white marble inlay. 43 This was a busy period for rebuilding around Cannon Street, indicated by the use of construction timbers felled in ad 83 in buildings east of the bridge across the Walbrook. 44 Moulded stones reused in early second-century foundations next to the bridge, at Regis House, included part of a stone conduit that may have come from a nearby public fountain or bathhouse. 45
The procurator as patron
Trevor Brigham has drawn attention to the fact that although London’s earliest masonry buildings were usually built over flint foundations, the monumental northern range of the ‘governor’s palace’ used ragstone and tile. 46 The early use of these materials may have been associated with the opening up of new supplies of brick and tile to meet the elevated needs of building programme in the mid-80s. At about this date local kilns also started producing brick and tile stamped by the office of the procurator of the province of Britain at London. 47 Discards of overfired and underfired tiles carrying this stamp have been found near Gresham Street on the western margins of the Flavian town, indicating that the kilns lay nearby. 48 Similarly stamped material was manufactured near the roadside settlement at Brockley Hill, north-west of London (below p. 183). None of these distinctive tiles has been found in pre-Flavian contexts. The earliest stratified example was found at Newgate Street, in a context thought likely to be early Flavian but where a later Flavian date cannot be discounted. 49 They were used in several important buildings from the mid-Flavian period into the second century, but there is no record of their presence in the first phase of the Flavian forum and other building programmes dated to the 70s. This might indicate that they were first produced to supply construction programmes in the mid-80s, possibly commencing with the construction of the monumental complex at Cannon Street c. ad 84. Tile kilns could be built to supply particular construction projects where capacity was lacking, and bathhouses had a particular need for specialized building materials.
Procuratorial involvement in this tile production may have made particular sense at this time. Whilst army engineers were involved in some earlier phases of building, particularly after the Boudican revolt, this would no longer have been feasible. The forces commanded by the governor Agricola were heavily engaged in campaigns aimed at completing the conquest of Britain. 50 Agricola’s crowning victory, the battle of Mons Graupius, was won in the Scottish Highlands in the summer of ad 83. Any support the Roman army might have lent to earlier programmes of building would have been diminished by these competing needs. The absence of the governor would, however, have left the procurator in a commanding role.
The fact that the procurator’s office was involved in local tile manufacture, using official stamps to manage production and supply, doesn’t mean that it was responsible for the buildings where the tiles were employed. The tileries could have been commercial enterprises, exploiting imperially owned assets to meet market opportunity and generate income for the imperial treasury. 51 London’s civic administration, perhaps newly installed in the Flavian forum, could have taken on increased responsibility for public works funded from local resources and employing architects and engineers brought in for the purpose. The patronage of London’s leading merchants and aristocrats, now competing for social status using the time-honoured tools of civic euergetism, might also have contributed resources for the building of the baths and other public monuments of Britain’s premier city. On the other hand, the Roman emperors are known to have endowed various cities with public baths. 52 If Domitian had wished to make a gift of a new bathhouse, drawing on the spoils of northern conquest, then the procurator would have been tasked with bringing this about. In the absence of any building dedication claiming credit for the works, we cannot know whether this was a private, civic, or imperial project. But the involvement of the procurator’s office in supplying building materials, as well as the style and date of the building, weigh in favour of seeing this as the product of imperial patronage.
London’s Flavian growth started in the public sector, anticipating rather than reflecting on the town’s success. Whichever agencies or individuals were responsible, these projects required the approval of the provincial government. 53 It has consequently been suggested that London’s public works were enabled by imperial patronage and aimed at the creation of a suitably impressive provincial capital. 54 The completion of the conquest of Britain in ad 83 warranted official celebration, and was the likely inspiration for the construction of a monumental marble-clad arch at Richborough. 55 London, the most important city of the province, might also have merited commemorative architecture. The programme of Flavian advance may have been initiated in London after Vespasian’s accession to power, perhaps marked by festivities that accompanied the building of the amphitheatre. A decade later, Domitian’s general finished the task and London ruled all of Britain: a status soon to be lost and only regained 1600 years later with the 1707 Acts of Union. This was an appropriate moment for the erection of an imperial bathhouse, a symbol of the rewards that Flavian patronage brought to the people of London.
The Huggin Hill baths and town houses
The monumental building at Cannon Street was one of a series of late first-century building projects. Ian Betts has noted similarities in the building materials used in Roman monuments at Cannon Street station, Huggin Hill, and Winchester Palace in Southwark, suggesting a coordinated programme both sides of the river. 56 At its conclusion, large public bathhouses probably stood in all three of London’s main districts.
London’s best-preserved bathhouse occupied a terrace overlooking the Thames west of the Walbrook, where natural springs emerged from the junction of gravel and impervious clay (Fig. 11.3). It was first identified at Huggin Hill in 1964 and investigations in 1987–9 found that its walls survived 3 metres high, culminating in the reburial exercise described above p. 19. 57 The tile-walled building was set over a thick ragstone and concrete raft supported by timber piles. The baths initially followed the row layout favoured in Romano-British forts and towns, similar to Silchester’s late Neronian baths. The entrance was at the east where changing and cold rooms were located. These gave access to a warm room (tepidarium), about 11 metres square, from which the hot room (caldarium) was reached. This apsidal-ended vaulted space measured 13 by 5.5 metres with a bath set within a smaller apse to one side. These rooms were decorated with polychrome mosaics, only fragments of which survived. Pottery suggests an early Flavian date for their construction. It is an open question as to which of the two monumental buildings that flanked the Walbrook was the first to be built, but the differences in layout and context suggest that different communities were catered for.
Fig. 11.3 The Huggin Hill baths (GM240/DMT88: after Marsden 1976 and Rowsome 1999). Several phases of construction are illustrated. The caldaria on the north side and at the east end of the complex were both later additions, as was the latrine block to the south. The changing line of the waterfront to the south of the baths is not shown. Drawn by Justin Russell.
The Huggin Hill baths served the district west of the Walbrook. Although this residential area was probably planned before the Boudican revolt (p. 84), it wasn’t densely populated until the Flavian period (Fig. 11.4). The character of the occupation hints at the arrival of settlers with a background in military service. Town houses at Watling Court were strikingly similar to contemporary centurion’s quarters in the fortress at Gloucester (Fig. 11.5). 58 One of these building’s earth walls were built over masonry foundations formed of ragstone rather than flint, favouring a construction date after c. ad80/85 which is consistent with the sparse pottery dating evidence. At the time of its destruction in the Hadrianic fire the building housed a bronze diploma recording a grant of citizenship and marriage rights of a type issued to members of the Roman auxiliary army on completion of 25 years’ service. 59 The names of witnesses to the document indicate that it was issued between ad 98 and 108, which means that its recipient probably started his military career between ad 73 and ad 83. He might have been in service when the diploma was issued, since it was only after ad 110 that such documents were restricted to veterans. We don’t know why this individual chose to settle in London, but he may have served here in an administrative role. 60 Auxiliary centurions included personnel recruited from provincial aristocracies commanding the status and wealth implied by the design of this house. 61 It contained an important series of late first-century mosaic pavements with black-and-white designs showing close stylistic parallels to pavements at Reims in Gaul. The floor in one room, decorated with small mosaic crosslets inset into a highly polished crushed brick terrazzo (opus signinum) floor, slumped into an unconsolidated quarry and required replacement with an entirely new floor of the same kind. These repairs showed that client and contractor retained a continuing commitment to the project over some years. Although the design of this house may have drawn on skills drawn to London for unrelated reasons, perhaps to work on public contracts, the client’s tastes and contacts may also have been a factor. If so, this may have been the house of someone recruited to the army from northern Gaul, who had commanded auxiliary troops under Agricola and was employed on imperial service in London.
Fig. 11.4 A proposed reconstruction of London c. ad 85. Drawn by Justin Russell.
Fig. 11.5 The Flavian town houses at Watling Court (WAT78: after Perring and Roskams 1991). Drawn by Justin Russell.
A comparable town house was found at Gresham Street, west of the Cheapside baths, next to a road built c. ad 80. Debris from the structure included Purbeck marble wall veneers, hypocaust flooring and tesserae from a disturbed mosaic. 62 A copper pipe suggests that it was supplied with running water. A carnelian intaglio depicting Mars found on the opposite side of the street hints that someone associated with the military lived hereabouts. Nearby, at Gutter Lane, a black and white mosaic panel set within a red border was found inside an apsidal-ended reception room in a clay-walled building beside the road to the south gate of the Cripplegate fort. 63 These formed part of a cluster of at least six houses flanking Cheapside that contained impressive late first- and early second-century mosaics. 64
High-status town houses were also introduced to other districts in the late first century, if in smaller numbers. Houses containing mosaics were built either side of Bishopsgate (Ermine Street), on the north side of town. This investment in luxurious display was reflected by a stone inlay (opus sectile) floor at 28–32 Bishopsgate and a large terrazzo-floored cellar reached by a flight of stairs at 7–11 Bishopsgate. 65 The presence of mosaic pavements indicate that another high-status property was found over the site of the Neronian fort at Plantation Place. 66 These houses were probably built soon after ad 85, when the materials that they employed came into common use in London’s domestic architecture. Most were built of timber and unfired clay, and the best had solid walls of rammed earth, mud brick, or clay slab set over masonry footings. Only monumental public buildings were wholly stone-built at this time, although brick was increasingly used for string coursing, and for quoining around doors and corners. Most pavements survive only as small fragments with geometric designs, the best of which were of a quality only surpassed in Britain at a handful of palatial villas on the Sussex coast. Terrazzo floors (opus signinum) of cement and crushed tile were also a late first-century innovation following the increasing availability of brick. The houses were commonly decorated with painted walls, where red panels with decoration on a black background were particularly popular. The distribution of these higher-quality town houses witnesses an influx of wealthy settlers into districts on the fringes of earlier housing, concentrated west of the Walbrook. They afforded space for sophisticated city living. Morning attendance to business in the forum could be followed by an afternoon with comrades at the baths, before closing the day with supper-parties in rooms fitted out with mosaics and wall-paintings. Some of the new residents were connected to the military administration, likely to have been brought here by the opportunities arising from the Flavian advance.
On the south-bank
Contemporary improvements south of the river involved the redevelopment of the riverfront site at Winchester Palace c. ad 80 (Fig. 11.6). 67 Reclamation dumps behind piled revetments incorporated a range of expensive building materials, suggesting that an earlier building had been refurbished or replaced. The earlier Flavian building included wall paintings of unparalleled quality in Roman Britain, incorporating cupids, dolphins, sea-monsters and a head of Medusa. 68 The new development included two unusual buildings associated with early Flavian pottery that encroached onto a metalled road. A large masonry building fitted out with a raised floor over parallel masonry sleeper walls, about 1.2 metres apart may have been a granary, as the excavator Brian Yule proposed, or perhaps a hypocaust floor of unusual design. An adjacent curved wall of massive posts, with tile noggin infill, formed a circular structure about 7 metres across. This tower-like structure may also have provided storage space. It consequently seems likely that the mid-Flavian development involved the addition of new monumental facilities, including bulk storage, to an establishment that included luxurious residential quarters and heated rooms. Associated food waste included high proportions of game, poultry, and pig, indicative of wealthier households. 69 This was an appropriate site for an official residence, facing the urban community across the Thames, and the suggestion that it was the residence of the procurator has already been noted (above p. 79).
Fig. 11.6 The high-status Flavian building complex found at Winchester Palace, Southwark (WP83: after Yule 2005). Drawn by Justin Russell.
The building tentatively identified as a mansio at 15–23 Southwark Street was also rebuilt in the late first or early second century, perhaps after brief abandonment. It is not clear if it retained its earlier function, although the presence of a hypocaust floor demonstrates its high status. 70 This floor was unusual both for being set within a clay walled building and for the use of circular tiles made in kilns at Radlett on the Verulamium road. 71 Finds of metal junction collars used to join wooden water-pipes found nearby suggest that the reconstruction involved the addition of a small bathhouse. 72 Another public building may have been built or restored on Southwark’s southern island. Demolition debris from a high-status structure was found over Flavian reclamation dumps at 33 Union Street. 73 The material included imported marble and coloured stone from floors and wall veneers. Loose tesserae included unusual examples made in glass, forming the largest group of such tesserae from Roman London. Associated ceramic building material included fragments of ceramic water pipe. Some of the earlier tile fabrics support a first-century date, although the coloured marbles probably reached London in the early second century, perhaps in a phase of refurbishment. These construction materials were different to those associated with the high-status buildings at Winchester Palace and 15–23 Southwark Street, and derived from a different building.
In addition to these higher-status buildings, the road to London Bridge was crowded with Flavian shops and workshops (Fig. 11.7). One had walls of roughly worked ragstone and tile when such materials were rarely employed in London’s domestic buildings. 74 It consisted of three ranges set around an inner courtyard, occupying a street frontage at least 14 metres long. The building lacked the expensive interior decoration and high status finds one would expect to find in a town house. The excavators suggest it to have been a market building or macellum, drawing parallels with a similar structure found in Cirencester. 75 This is a type of building better known from the cities of Roman Italy, where small units for retailing meat and fish were arranged around central courtyards. 76 Although the Southwark building lacks the architectural complexity of these Mediterranean examples it was a likely place for the distribution of foodstuffs. It was built next to a butcher’s shop, indicated by cattle bone waste, and a bakery containing brick-and-tile ovens associated with burnt cereal grain was built nearby. 77 The building next door may have been a smithy. Other works of this period include the construction of a road leading to new housing between Watling Street and the river at Tabard Square. 78 Rubbish associated with this development, notably animal bone and glass, suggests a high-status character to the occupation. The earliest architectural embellishment of the ritual landscape beyond London’s southern border may also date to the late first century, when a timber colonnade was built around an open area that may have housed a shrine at Trinity Street. 79
Fig. 11.7 Late first-century street-side buildings and possible market-hall (macellum) found at Borough High Street in Southwark, with nearby baths and a building that might possibly have been a temple (in large part from BGH95, LBN08 and BVK11: after Drummond-Murray and Thompson 2002 and Ridgeway et al. 2019). Drawn by Justin Russell.
This was one of the busiest periods of building activity on both sides of the river. The variety of this architecture suggests the operation of a developed housing market, drawing on supplies of imported materials and the skills of specialist craftsmen. Some of the most important building works may have been commissioned by the procurator, but the entire urban community appears to have enjoyed a confident and busy prosperity.