Part 7

The Late Antique City

26

Restoration (c. ad 270–85)

The riverside wall

London was diminished by the changes that closed its port, but these problems were not terminal. Some 20 years later, the city was emphatically restored. This revival was announced by the construction of a monumental masonry wall along the riverfront. The quays at Billingsgate were briefly rehabilitated beforehand, using timber baulks salvaged from the earlier waterfront to re-establish a waterfront to which a new stave-built section was added. ¹ Revetments at Swan Lane and New Fresh Wharf were similarly refurbished. These works might have been designed to improve transport infrastructure in the later years of the Gallic Empire, although the brevity of their use might indicate that they were built to take deliveries of building materials for the construction of the riverside wall. In either case, these quays were soon blocked from the river when the wall was built (Fig. 26.1).

Fig. 26.1 A proposed reconstruction of London c. ad 295. Drawn by Justin Russell.

The wall’s existence was established by discoveries at Baynard’s Castle in 1974 and has since been confirmed by numerous observations along Thames Street. ² These trace the line of a 2.2-metre-wide tile-coursed concrete construction, faced with small ragstone blocks and incorporating a basal plinth of Reigate stone, that was set over a rammed chalk raft laid on five neat rows of squared oak piles. Although most of the riverfront was enclosed, a short length at its marshy western end was probably left open until later (p. 387). An earth bank may have been raised inside the wall, although this is not proven. The wall matched the landward wall built half-a-century earlier, but incorporated a significant amount of reused building material, witnessing an increasingly pragmatic approach to procurement. ³ Although the wall saw later alteration, the consistent design of its distinctive foundations shows it to have been planned as a single construction. Exact dating eludes us, since none of the squared-off timbers retained bark. Piles at Baynard’s Castle included one with a final growth-ring of ad 255, and the heartwood-sapwood transition of others indicates that some came from trees felled no later than ad 270/275. ⁴ Those at New Fresh Wharf were felled after ad 268, providing a terminus post-quem for the wall’s central section. ⁵ Piles beneath the eastern stretch of the wall at Sugar Quay were felled ad 279–301 and others from Three Quays are dated ad 251–87. ⁶ We can therefore conclude that parts of the wall were in place c. ad 275 but others weren’t built before ad 279. The building programme probably started at the western end in the mid-270s, drawing on supplies brought down-river, and progressed eastwards until completion around the end of the decade.

The wall severed the city from the river but was penetrated by water-gates. One marked the entrance into town over London Bridge, and others gave access to places where rivercraft beached. Billingsgate was probably the main port of entry: lead seals found on the foreshore included an example of the fourth-century emperor Constans suggesting that imported cargoes were being taxed at a nearby custom point (p. 301). ⁷ There is no evidence that any quays were built to receive these imports, although warehouses associated with the earlier waterfront survived. ⁸ In places a low bank was set in front of the wall to define the foreshore, but little other than rubbish dumping took place between wall and river. The building of the wall involved extensive reclamations of low-lying areas around the mouth of the Walbrook. Excavations at 14–16 Dowgate Hill found a 2-metre depth of rubbish dated after c. ad 270/275, containing large assemblages of pottery from Oxfordshire, Colchester and Nene Valley, with more than 100 leather shoes and sandals and several hundred ‘barbarous radiate’ coins. ⁹ This was a rubbish of urban wealth.

The wall’s construction kicked off a period of renewal, not dissimilar to London’s Severan restoration (Table 26.1). It relied on advanced engineering skills, new supplies of building materials, and a dedicated construction workforce. If the mid-third century had witnessed failures in production and supply, then part of this lost capacity was restored. The dating suggests that we should credit this to the years following the collapse of the Gallic Empire in ad 274, and Aurelian’s reassertion of Roman authority offers a likely political context for this exercise. ¹⁰ At a general level, if not in details of design, inspiration may have been drawn from Rome’s Aurelian wall, the building of which began in ad 271. ¹¹ As Hendrik Dey has observed, Rome’s walls witnessed the emperor’s commitment to his capital city, testifying to the permanence and authority of his regime. This was a language of imperial control that was readily translated into the architecture of provincial cities.

Table 26.1 A suggested timeline of events affecting London in the period ad 270–410

The completion of London’s defences also echoed programmes of urban fortification elsewhere in the north-west provinces, particularly Gaul and Spain. We lack sufficiently precise dating to know how these parallel initiatives influenced one another, but some are thought to have been centrally directed. ¹² Several places in southern Britain also gained masonry defences at this time, most notably amongst the east-coast sites that subsequently formed the command of the Saxon Shore. ¹³ Richborough was part of this group, where coin evidence indicates that the masonry defences were built between ad 273 and 285. ¹⁴ Defensive circuits also encircled London’s neighbouring cities: the stone walls of Canterbury are dated c. ad 270–90 from stratified coins and pottery, and those at Chichester, Silchester, and Verulamium were broadly contemporary. ¹⁵ This proliferation of defended sites would have placed competing claims on available resources, and their maintenance may have drawn on taxes and rents previously rendered to London. If so, some estates may have been removed from London’s orbit and assigned to new commands, accounting for changes in patterns of supply within the region.

The restoration of London appears, therefore, as part of a concerted building programme involving a large-scale mobilization of resources. It also assumed a significant ongoing commitment of forces to man the 5-kilometre-long defensive circuit. James Gerrard has drawn a useful parallel with requirements described in the tenth-century Burghal Hidage, which imply that some 2,700 men would have been needed to defend Roman London properly. ¹⁶ Defensive walls both projected and protected imperial interests, and were planned at the instigation of the ruling authorities. Whether or not they were built by the state, or by cities responding to impositions and expectations, is harder to judge. ¹⁷ We should not underestimate the symbolic importance of town walls. Places where power and wealth were concentrated were made secure, and their authority and status enhanced. The walled cities formed a network of fortified strongholds at strategic points, housing the military and bureaucratic apparatus of a rejuvenated state. The completion of London’s town wall also helped to control of movements of people, securing the collection of port duties on goods entering through the gates. Such taxes would have underwritten the costs of defending the city, supporting the garrison placed on its walls.

The emphasis placed on the display of strength can be seen as a reaction to earlier insecurity. Recent studies of the Saxon Shore forts have tended to downplay the importance of barbarian threat as a prompt for their construction, emphasizing their role in controlling shipments along coastal supply routes. These are, however, different sides to the same coin since Rome’s investment in the logistics of supply was part of its military strategy. An increased sense of vulnerability might plausibly have contributed to the decision to set a wall along London’s waterfront. Carausius’ subsequent appointment to pacify the sea from Franks and Saxons (below p. 350) shows that such fears were taken seriously. ¹⁸

The fact of London’s late third-century recovery is remarkable. Despite the perils of collapse and fragmentation, Rome re-established political authority and revived its cities. ¹⁹ This restoration, perhaps instigated by Aurelian and taken to completion by Probus, may have involved a new conceptualization of London’s role, setting it at the command of a network of sites that were as much a frontier as a coastal supply route (below p. 367). Following the closure of its harbour facilities London became a very different place marked by a different kind of archaeology. The abandonment of the waterfront quays deprives us of a principal source of archaeological evidence, and the late Roman city relied extensively on recycled and local resources. The story of later Roman London is therefore drawn from a slighter body of material offering fewer secure dates.

New patterns of procurement

The building of the wall coincided with a marked change in the range of pottery used in London. A new ceramic phase can be described from an influx of wares manufactured in southern Britain. ²⁰ In particular, London started to receive significant quantities of pottery made at Alice Holt near Farnham c. ad 270. These kilns were located on the western margins of the Weald, in an area that benefitted from access to river transport. Alice Holt potters produced ‘Romanized’ products from the Flavian period onwards, but these only reached London in small quantities. ²¹ A suddenly increased volume of importation shows that this part of Surrey became more closely integrated with London’s consumers, meeting some aspect of urban demand previously met from elsewhere. The pottery can be seen as a proxy for movements of other goods, perhaps the bulk supply of grain or timber. Since Alice Holt’s ancient oak forest was an important source of wood for eighteenth- and nineteenth-century shipbuilding it may have filled a similar role in late antiquity. ²² The wood used in later Roman London was obtained from smaller faster-grown oaks than previously, and although this may have been a response to earlier deforestation it is more likely to reflect on changed supply networks involving different woodlands. ²³ Some pots brought to London from Alice Holt kilns are likely to have been used as containers for local produce, such as honey and mead. Regional sources of alcohol would have been sought after when the bulk importation of continental wine ended. The value of these crops, allied to the improved transport infrastructure developed around the movement of timber or grain, might explain why London imported large quantities of Alice Holt pottery instead of stimulating the establishment of kilns closer to town.

London also started to import distinctive types of roof and flue tile made at Harrold in Bedfordshire some 87 kilometres north of London. ²⁴ These must have been carted along Watling Street through Verulamium. There is no evident reason why London needed to procure building materials from such distant kilns, and this traffic must also have been integrated with some other aspect of supply. Few villas are known around Harrold, where production may have been geared towards more distant consumers. This too may have been dominated by grain, although several estates in the region turned to an intensive form of horticulture involving close-trench lazy bed cultivation some of which may have been planted with vines. ²⁵

The development of new patterns of agricultural and industrial production might have followed changes in ownership. The collapse of the Gallic Empire is likely to have provoked confiscations, interrupting and reconfiguring the relationship between London and its hinterland. The restoration of London did not result in a return to former patterns of economic exploitation, as evident in the reduced intensity of production in the Weald. Some lost capacity was never replaced, perhaps because the provincial economy had been reorganized in ways that no longer relied so heavily on London and its region, and perhaps because labour remained in short supply. It is also possible that estates that formerly sent tax and rent to the procurator in London were reassigned to other commands, such as those of the Saxon Shore. The office of procurator was itself soon to disappear and its assets, as well as its responsibilities, must have been redistributed. Although the office was still functioning in the early third century, there is no direct record of procuratorial involvement in Britain beyond the middle of the century, and the post didn’t survive later reform (p. 367). ²⁶

Property redevelopment

There was a small-scale building boom at the close of the third century, accompanied by repairs to some city streets. ²⁷ The extensive Lloyd’s Register site, vacated in the mid-third century, was comprehensively developed with new timber and masonry buildings set out on a new layout c. ad 270 onwards. ²⁸ The data reviewed by Peter Marsden and Barbara West (discussed above p. 282) also implies a wider demographic rebound. This is reinforced by the evidence of improved pottery supply, encouraging Robin Symonds and Roberta Tomber to describe the late third to early fourth centuries as a period of revival. ²⁹

The distribution of tiles imported from Harrold at this time indicates that the better-appointed town houses followed the terrace behind the riverside wall and along the Walbrook, with sparser distributions west of the Walbrook and in Southwark. Buildings were placed in areas where running water was easily obtained, and access to fresh water remained an increasingly important factor in the distribution of aristocratic housing. Several second-century masonry houses were restored and enlarged. At Plantation Place alterations c. ad 270 onwards included the addition of a small heated bathhouse, and this is the date of the renovation of heating systems in the house at Lothbury. ³⁰ Elsewhere earlier structures were demolished and replaced, as at 25–30 Lime Street where a large apsidal-ended building was built using the close-piled foundations typical of the period. ³¹ Several houses included extended enclosures and colonnades, and some incorporated mosaics and stone-walled cellars. ³² This domestic architecture was rooted in the fashions of the previous century, but executed on a slightly grander scale. Earlier vernacular traditions of timber-framing and mud-brick walling continued, but chiefly in workshops and outbuildings. ³³ Problems of survival make it impossible to estimate building densities, but there is no reason to believe that the city was any smaller at the end of the century than it had been at its beginning.

One of the masonry houses built at this time, at the Lloyd’s Register site in Lime Street, incorporated a small coin-hoard of c. ad 275 beneath the floor, perhaps as a foundation deposit. ³⁴ It consisted of thirty-two of the small copper issues known to numismatists as ‘barbarous radiates’. This irregular coinage copying issues of the later Gallic emperors c. ad 268–73 circulated widely throughout the north-west provinces. ³⁵ The Lime Street hoard included several coins from the same dies, suggesting that they were manufactured locally, perhaps testifying to the continued presence of authorities concerned with facilitating transactions using small change. The production of these ‘radiates’ expanded on earlier attempts to produce a token coinage, possibly compensating for the decreased purchasing power of the debased coin in circulation. The readier availability of currency may also have facilitated a shift towards cash taxation, encouraging the monetization of Britain’s later Roman economy. Despite this, these coins remained proportionally more common on urban and military sites. Roman officials relied extensively on cash to pay for food, goods, and transport, especially when travelling on official business, and such uses are sufficient reason for London to have needed additional small change. ³⁶

The forum basilica was repaired following its earlier fire destruction (above p. 332), and new cement floors laid within its nave. ³⁷ The masonry amphitheatre may also have been refurbished in the late third century. Pottery associated with a drain built reusing timbers felled ad 243 suggests that it was cleaned and repaired after c. ad 270. ³⁸ The arena was resurfaced and seating rearranged, perhaps in rehabilitation for games in the busy late third century (below p. 350). The refurbishment of these two great public buildings shows that the city sustained public life modelled on earlier arrangements, although we have no measure of the intensity of their use (see p. 356). In any case, the monumentality that mattered was expressed by the walled circuit which was more important than the public buildings within. Attention was otherwise redirected into smaller worlds of domestic luxury, following patterns that can be traced back to Hadrianic-Antonine fashion. That earlier ‘golden age’ may have framed aristocratic ambition towards the end of the third century.

Later Roman cemeteries and people

The late antique city was surrounded by its dead. The main burial grounds continued to lie beyond the town gates at Aldgate, Bishopsgate, and Newgate, and in Southwark (above p. 322). These grew significantly in the late third and early fourth century, as in Southwark where cemeteries expanded to the south as well as encroaching on peripheral parts of the southern eyot. ³⁹ These later burials outnumber earlier ones because of changed burial practices not population growth.

Only a minority of the dead was cremated. One from London’s eastern cemetery at Prescot Street deserves mention for the splendid glass vessels used as grave goods. ⁴⁰ These included a polychrome millefiori dish formed from hundreds of translucent petals of blue glass bordered with white, embedded in a bright red glass matrix, imported from the eastern Mediterranean. As time progressed, fewer people were buried with property. John Pearce has also described a shift in emphasis from the burial of items associated with feasting and care of the body towards those that emphasized rank and appearance. ⁴¹ This included jewellery, dress accessories and expensive clothing indicated by scraps of silk and gold thread. ⁴²

Most bodies were placed within oak coffins in well-ordered graves. A lack of intercutting shows that each was marked and respected. Their orientation was influenced by local topography, with graves aligned on roads and property boundaries, but with a distinct preference for laying bodies east-west with the head at the western end of the grave. ⁴³ This cardinal arrangement, where the rising dead could face the sun, mattered in many of the belief systems of late antiquity.

London adopted a late Roman fashion of adding chalk or gypsum plaster around the corpse, following practices adopted in North Africa and the Rhineland. ⁴⁴ Calcium carbonate from marine chalk was the preferred material. Some of London’s earliest chalk-lined burials were set behind the second-century mausoleum within a walled enclosure alongside Watling Street, south of the city. These probably dated to the third century and may have formed a family group consisting of a young man and woman, aged 26–45, flanking a child who had suffered from rickets. ⁴⁵ In the cemetery at Lant Street in Southwark the chalk burials are all likely to date after ad 270, where one in eight burials were of this type, rising to 30 per cent of all burials in an early fourth-century peak. Chalk burials were found in similar numbers in the eastern and northern cemeteries, where they made up 12.4 per cent of all burials and over 20 per cent of those dated to the late third and early fourth centuries. ⁴⁶ Since the chalk was imported this was an expensive treatment, perhaps used to counter decomposition in preparation for the bodily resurrection promised by the soteriological cults of late antiquity.

In sharp contrast with the earlier predominance of male burials, which was a feature of the Severan revival of London, London’s later cemeteries had a more normal male: female ratio. At Spitalfields men had outnumbered women by more than five to one in the late second and early third centuries, but from the late third century onwards this ratio had fallen to around 1.4:1. ⁴⁷ At Lant Street female burials of this period slightly outnumbered male ones. In all of London’s cemeteries children were under-represented, and most of those buried had died in their 20s and 30s, with few surviving beyond 50. ⁴⁸ The evidence from Spitalfields also shows an earlier peak in deaths, in the 26–35 age group, than in the early third century. ⁴⁹

Although London’s cemetery populations show a high level of skeletal stress compared to neighbouring towns, and pathologies indicative of dietary stress are common, Londoners were healthier than contemporary rural populations. Since London wasn’t densely occupied, and enjoyed a high-protein meat diet, it didn’t suffer excessively from the problems of infectious disease and inadequate sanitation associated with pre-modern cities. ⁵⁰ It was a place of wealth and luxury for a ruling minority.

Recent studies by the Museum of London’s Centre for Human Bioarchaeology reveal fascinating detail from a sample of London’s Roman dead. Drawing on the evidence of carbon and nitrogen isotopes to study diet, oxygen isotopes to examine migration, and macromorphoscopics to assess ancestry, this research confirms London’s continued demographic dependence on migration. The skeletal morphology of individuals buried in the late antique cemetery at Lant Street showed the likely African ancestry of four individuals and Asian-Chinese ancestry of two others, with oxygen isotopes suggesting that five had grown up around the Mediterranean. At least five, and probably eight, of the nineteen individuals studied in this research can be identified as migrants. ⁵¹ One was a young woman, probably about 14 years old at the time of her fourth-century death. ⁵² She stood about 1.6 metres tall, and her aDNA indicates that she had blue eyes and her mother’s ancestry was from eastern Europe or north-east Africa. Isotope analysis shows that she grew up in a Mediterranean climate, but was eating things likely to have been produced near London by the time she was nine. Although her diet included meat, fish, vegetables, and cereals, her teeth were in poor shape and she had suffered from rickets. The cause of her early death is unknown, but she was laid to rest on a bed of chalk, with two glass vessels set beside her head and an inlaid wooden casket at her feet. The bone inlay was carved with an image of Vesta, the goddess of hearth and home. She also took with her a rare folding knife with an ivory handle carved as a leopard or panther with paws outstretched, to which a bronze key had originally been attached by a chain. This valuable object was made to an exotic design that finds closest parallel in Carthage. It seems likely that this girl was born in North Africa and had travelled to Britain as a child. Her journey from childhood malnutrition to high-status burial suggest a change in social status of the sort that might follow marriage or slavery.

Studies of the isotopic signature of tooth enamel from twenty individuals buried in London’s cemeteries found that twelve had teeth consistent with a Romano-British origin, but four were unlikely to have grown up here. These immigrants included a female, probably 36–45 years old and possibly of continental origin, buried in the eastern cemetery in the mid- to late fourth century with distinctive ‘Germanic’ jewellery and an Alice Holt flagon at her feet. This jewellery included a composite triangular antler comb worn in her hair, and a pair of silver disc tutuli brooches linked by a glass-bead chain that fastened her clothes. ⁵³ This lady went to her grave dressed in the fashion of a German of high rank. This confident expression of barbarian identity is in keeping with a rising German influence on custom and dress, perhaps accelerated by settlements in Britain late in the third century. ⁵⁴ The emperor Probus is reported to have settled defeated Germans in Britain after victories of ad 278, where they are reported to have served Rome loyally. As with the earlier settlement of Sarmatians, above p. 293, it is tempting to see this as compensating for population decline. Another skeleton from this cemetery was a man of black African ancestry, with brown eyes and dark hair who was probably over 45 years old when he died and whose stable-isotope signature indicates a childhood spend in the London region. ⁵⁵

Although the sample is small, between one-fifth and two-fifths of this buried population were migrants. The establishment that ruled from London was always commanded by foreign-born elites, some only stationed in the city for a few years at a time. These individuals drew on the support of a network of clients and dependents who travelled in their wake. The transient nature of London’s elite society might explain why the British were so uninvolved in the patronage networks of the wider Roman world. Unlike their Spanish and Gallic contemporaries, Britons failed to make a mark on Rome’s Senatorial aristocracy or enter the social circles written about in ancient sources. Most Londoner’s of account were immigrants for whom service in Britain was but a staging point in their careers. Ancient city populations were, in any case, rarely self-sustaining. Episodes of contraction encouraged corrective flows of newcomers, and London’s late antique revitalization relied in part on immigration. Differences amongst cemetery populations suggest that London’s late third-century restoration depended more on the settlement of entire households within which women were well represented and less on transient communities of economically active males that previously dominated.