20

The Walbrook skulls

The new north road

North of the Roman town, beyond its streets and houses, lay a backland of meadows prone to flooding. A brand new metalled road was driven across this area around the time that London’s new fort was built. ¹ The road surfaces cambered over a clay and brushwood causeway that was carried over the Walbrook tributaries by a series of timber bridges, and at 9.7 metres across was much wider than the city streets. The western stream was redirected into a large ditch flanking the north side of the road, before being diverted beneath. Finds associated with road building included a Trajanic coin (ad 96–117), central Gaulish Samian and black-burnished ware, which date its construction c. ad 130. Burials anticipating the line of the earliest roadworks suggest that the route was established a little ahead of its physical engineering.

The scale of works suggests that the road served an important purpose. Road construction is frequently linked to the needs of the army, and a connection with official supply is implied by the presence of lost or discarded hipposandals along the route. These iron shoes were tied to hooves to improve traction on slippery roads, and are associated with military haulage. ² Other roadside finds included brooches, spearheads and belt fittings of types associated with the army. The faunal remains were dominated by horses that had reached the end of their working lives before death or slaughter, drawing on stock that was distinctly taller than usual and likely to have included cavalry mounts. These features suggest a road traffic including cavalry and heavily laden supply waggons. Timber platforms built where the road crossed the Walbrook may have been stations for watering animals and making offerings to the river. One was built reusing two doors made from timbers felled ad 110–34. Traffic may also have included prisoners and slaves, as suggested by the presence of a Roman shackle in later marsh deposits.

The road’s north-west heading aimed towards a known crossing-point of the river Fleet at Battle Bridge, near King’s Cross. This was where the medieval river was most easily crossed, and the likely site of a Roman ford. First- or second-century building material found near King’s Cross might derive from associated settlement and a legionary tombstone was found at Battle Bridge. ³ Two phases of road surface associated with Roman coins were observed during sewer excavations at Old Street in 1867 and are appropriately located to have been part of the route. ⁴ A road over the Battle Bridge crossing would have carried traffic north-west from Bishopsgate, continuing to join Watling Street just short of its crossing over the river Brent. This route from London to Verulamium navigated difficult terrain requiring substantial engineering, but was significantly shorter than the main road along Oxford Street and the Edgware Road.

Wayside human remains

Human remains were scattered alongside the road, filling adjacent ditches and channels. The unusual burial practices that generated this evidence started before the first road surface was laid. Excavations beneath Liverpool Street found the partially articulated right arm of a young adult or youth, left lying on open ground before it was buried in the road construction. ⁵ It has been suggested that the arm was washed-out of a nearby grave, but since it was found 0.3 metres above the floodplain floor this requires us to assume a major flood. This is unlikely since no river-lain sedimentation was identified, and fact that the arm retained connective tissue suggests that it hadn’t spent significant time in water. The limb is more likely to have reached its location through animal or human intervention, with the absence of scavenging pathologies more suggestive of the latter. Semi-articulated remains from at least three other people were found on the other side of the river channel at Finsbury Circus. ⁶ These too were found beneath the road, where the partly fleshed remains would have been visible to road engineers. At least one other body was buried in a grave here before the road was built. A concern with stray body parts was subsequently evidenced by the excavation of a charnel pit containing parts of seven individuals buried with pottery dated c. ad 140–60. ⁷ These remains were treated in a similar fashion to the disturbed Neronian bodies suggested to have been victims of retributive abuse after the Boudican revolt (above p. 110). The pre-road scatter of semi-articulated body-parts in the upper Walbrook took place within a few years of the Hadrianic fire, suggesting that these too might witness abuse vested on enemies slain in the course of a revolt or its suppression. The landscape reminds us of the sight that Tacitus describes as having greeted Vitellius after the first battle of Bedriacum in the civil war of ad 69: the body of the imperial legate was cremated in customary fashion and a few bodies were buried by friends, but otherwise the battlefield was strewn with mangled corpses, severed limbs, and the putrefying forms of men and horses. ⁸

A burial ground was subsequently established alongside the road west of the Walbrook, where it occupied a strip little more than 20 metres wide. Excavations in Finsbury Circus and Eldon Street between 1987 and 2007 recovered ten cremations and over 100 inhumations, most dated between the early second and early third centuries. ⁹ The graves included an unusually high male:female ratio (at 4:1), and older adults are under-represented. This was a sorry place for a graveyard, the low status of which was suggested by a low incidence of coffins and containers, an absence of animal bones associated with funerary meals, and a range of irregular burial practices including prone and crouched body positions. It is also unusual to find such a high proportion of inhumations at a time when cremation was widely preferred, reminding us that a quickly dug grave was the cheaper way of disposing of a corpse. Several shallow graves belonging to this cemetery were placed within the ditch that channelled the tributary of the Walbrook, following the earlier tradition of ditch burial. These graves were soon eroded by watercourse migration and flooding, carrying away parts of the body and adding to post-mortem body fragmentation. Some unstratified bones from the site also carried de-fleshing cut-marks, from the butchery of corpses.

The unusual practices also included a decapitation burial, and two instances where iron rings had been forged around the lower legs. One adult male, aged about 18–25, had iron rings welded around both ankles. In another case a corpse appears to have been buried with its hands tied behind its back. Other burials were treated with respect: the handful of cremations found here necessarily involved funeral pyres with attendant ceremony and offerings, and a fifth of the inhumations were accompanied by grave goods. These more ‘ordinary’ burials involved a fairly representative group of deceased, including adults, teenagers, children and babies. Six roadside inhumations were also found east of the Walbrook crossing, dated c. ad 140–60 by associated finds. ¹⁰ Three of these bodies, all of individuals 18–35 years old, had been decapitated from the rear or right side in violent blows made by short-bladed instruments, probably swords. The angles of the wounds are consistent with execution from behind to victims kneeling with head bowed. ¹¹ One of those decapitated was a woman with an isotopic signature suggestive of a childhood in chalklands like those surrounding the London basin, with a high presence of lead in tooth enamel indicative of early life spent in a city. Another roadside burial contained a male aged 26–35, with an iron ring forged tightly around his right wrist. The hot metal had been beaten into shape around the outstretched arm, either in brutal torture or to disfigure the corpse. We can conclude that these burials included victims of military punishment and execution. The northern marshland, on the banks of the Walbrook, within sight of fort and amphitheatre, was an appropriate place for the marginalized dead: including some who were allowed a poor burial but others who were not.

Heads of the dead

Many human crania have also been found around the upper Walbrook valley (Fig. 19.4, p. 246), and early discoveries of these Walbrook ‘skulls’ may have inspired Geoffrey of Monmouth’s twelfth-century story of a massacre of Roman soldiers besides a brook. Immense numbers were found in nineteenth-century sewer digging and construction around Finsbury Circus. ¹² No count was made of these early finds, but subsequent discoveries result in over 300 crania being listed in published accounts. ¹³ Some antiquarian discoveries came from the stream-bed, and many were marked and stained from submersion with organic material in stagnant or slow moving water. Few can be dated, but they generally derived from contexts likely to pre-date the Roman town wall built c. ad 200. Drawing on our understanding of the urban topography, it is difficult to identify a source for large numbers of skulls in the upper Walbrook valley prior to c. ad 120 and most of the more recent discoveries were deposited between c. ad 120 and ad 160. Those that have been examined are disproportionately from young males, and some show evidence of violent treatment. ¹⁴ The exceptional nature of this concentration of crania merits emphasis. Altogether these finds represent nearly a tenth of all recorded human ‘burials’ from Roman London. ¹⁵ Since only a small fraction of the relevant landscape has been archaeologically investigated, perhaps no more than 5 per cent, we can confidently conclude that the river’s environs contained many hundreds of crania and probably several thousands.

Many crania had spent time in the river before burial, but we don’t know how they had first found their way into the water. It has long been thought likely that some were washed out from upstream burial grounds. ¹⁶ Skulls can be river-rolled further than other human remains, carrying them to river bends where they are deposited in slower water. Such remains are also easier to identify than other skeletal parts, encouraging their selective recovery for later reburial. Ease of identification is also likely to have resulted in their disproportionate retention by workmen digging on Victorian sewers and construction sites. This idea seemed to find confirmation in the fact that several graves in the roadside cemetery at Eldon Street were damaged by river erosion. Altogether fifteen burials in this cemetery missing their heads were identified as having been disturbed by flowing water. These missing crania are likely to have been carried downstream, adding to the material within the Walbrook. The cemetery was much too small, however, to have generated hundreds of Walbrook skulls. Although fifteen heads had washed from their graves, nineteen isolated crania were recovered nearby. There was no net loss of crania from this cemetery into the wider river system. The dating evidence is also somewhat contradictory, since some of the fragmentary material predates the known phases of inhumation burial. Nor can we find any trace of a larger upriver burial ground from which thousands of crania might derive. Extensive investigation between Moorgate and Bishopsgate shows that the river catchment area was dominated by pasture and marsh, not cemeteries. ¹⁷ Although cemeteries have been found either side of Bishopsgate these lay on the margins of the Walbrook headwaters. The burials found here are also later than the dated Walbrook skulls and show no sign of water damage. ¹⁸ It is not possible to reconstruct any pattern of flooding that would have carried numerous crania from this area into the main channels of the river. The argument that several hundred skulls were washed into the Walbrook from ill-positioned graves is simply not tenable. It is instead likely that most entered the water by human agency, as was demonstrably the case for the non-fluvial finds. There was a deliberate pattern of disposing of fragmentary human remains, disproportionately the heads of young men, in wet places. This involved the massive intensification of ritual practices previously witnessed on the borders of the Neronian city.

These remains are scattered over a wide area. Second-century drainage ditches investigated at Moor House, half way between the Walbrook and the Cripplegate fort, contained parts of human skulls and long-bones that showed evidence of dog gnawing and post mortem knife cuts, which Jon Butler suggested might derive from the manipulation of skulls and long bones in excarnation rites. ¹⁹ Previous studies of the Walbrook skulls have given particular weight to the possibility that Celtic ideas concerning the veneration of the human head may have contributed to such practices. ²⁰ But the emphasis on the manipulation of select body parts finds parallel throughout north-west Europe, and cannot be characterized as a particularly Celtic practice.

Several more recent excavations have added significantly to our understanding of the Walbrook ‘skulls’. In 2013, the cranial remains of thirty-seven individuals were recovered from reworked gravel deposits on the river bank during tunnelling for the Elizabeth Line ticket-hall at Liverpool Street. Many showed signs of damage consistent with being rolled along the river bed, and the deposits in which they were found may have been dredged from the river c. ad 140/160. All of the deceased were adult, none more than about 45 years old, and four-fifths were male. One presented evidence of a blunt force injury. ²¹ Radiocarbon dating shows that although most of these remains come from people who might have died in the Hadrianic period, two were certainly earlier, dated cal ad 55–80. This confirms the reworked nature of the assemblage, and shows that the practices that resulted in crania entering the river had longer ancestry.

Other crania were subsequently found in the fills of the roadside ditches that flanked the Hadrianic road east of the Walbrook. Elements from the skulls of at least twenty-one individuals were placed in clusters at roughly 5 metre intervals along the southern side of the road, with others along its north side. ²² Other body parts were also present, but in smaller numbers. Radiocarbon dating indicates that the crania came from individuals who died after c. ad 90, and all were buried in contexts dated no later than c. ad 160. All were adults, with about twice as many men as women. One had a depression typical of healed blunt force trauma. Several were polished and scratched, suggesting that they had been exposed to running water, whilst others were brown-stained from lengthy submersion in stagnant water or bog. Whilst the roadside ditches were sometimes water-filled, this doesn’t account for the extent of water discolouration and wear. It is probable that the crania had been recovered from some other wet-place, such as the adjacent river, before being placed in the ditch. None was intact at the time of burial, and the mandibles had probably become detached as a consequence of natural decomposition or river damage before being collected for reburial. A fragment of woven textile indicates that at least one cranium was wrapped in a bag or cloth when buried.

One of the most curious discoveries made during this study was that two individuals had teeth with highly unusual isotopic profiles. ²³ Both were males, aged 26–35, whose limited childhood exposure to lead is a sign of a rural upbringing and who had grown up in an area of granitic geology. These geological conditions cannot be found in England, but occur in the Scottish Highlands. Sources in Scandinavia or Ireland are also possible, along with a handful of other remote locations. ²⁴ These two men might have been Scottish highlanders, or perhaps seafarers from Ireland or Scandinavia. Since only four crania from the ditch were analysed the fact that two shared such an exotic origin is wholly remarkable. It leads us to imagine a circumstance in which a British rebellion included highlanders from beyond the wall amongst a host that met its end on the banks of the Walbrook: an imagining prompted in part by the fact that this was the path incompletely trod by the Stuart pretender’s Jacobite army in the 1745 uprising.

Another of the studied crania was from an immigrant to the province, having grown up enjoying a diet found around the Mediterranean. The presence of foreigners from different parts of the Roman world is consistent with the cosmopolitan nature of London, but reminds us of the dangers of assuming a simple fault-line between Roman and Briton. The working hypothesis adopted here is that the unusual concentration of fragmented human remains in the upper Walbrook valley was in part the consequence of corpse abuse vested on British rebels. There are many reasons, however, why some may have come from people who were not British by birth. The disaffected communities drawn to revolt against Rome often included the urban poor, slaves, brigands, and deserters. We are reminded that when Goths descended on Rome in ad 409 they were met by crowds of slaves streaming out of the city to join them and that many of Rome’s most successful enemies were deserters from its own armies. ²⁵

Further valuable information derives from thirty-nine skulls found in waterlogged pits further downstream, on the west bank of the Walbrook close to another bridge over the river. ²⁶ These were also deposited c. ad 120–60. Dog gnawing and puncture marks shows that some had been scavenged when soft-tissue was still present. Almost all came from young men, 28–35 years old, and most carried injuries inflicted around the time of death, including one individual whose cheek was horribly shattered by a violent blow. At least one of these heads had been decapitated with a sword, and many had healed earlier wounds. The aDNA analysis of one suggests that the deceased, a black haired and brown eyed male, was an immigrant whose mother’s family came from eastern Europe or the Near East. ²⁷ Rebecca Redfern and Heather Bonney suggest that some of these crania were trophy heads, perhaps deriving from contests and executions in the nearby amphitheatre or brought to London from wars elsewhere. ²⁸ Whilst it is unlikely that large numbers of expensively trained gladiators died in the amphitheatre, the arena was also a place of legal execution. ²⁹ The governor’s judicial duties included condemning criminals and enemies to death, and hence the presence in London of speculatores such as Celsus who were responsible for beheadings. ³⁰ The unwanted corpses of such noxii risked being denied burial and disposed of in the river. ³¹ In Rome, the decapitation and display of severed heads was closely associated with the punishment of treason, and histories describe how the heads of defeated emperors and usurpers were paraded in proof that they no longer threatened public order. These heads were sometimes brought to Rome for display in the forum, before being disposed of in rivers and sewers (see p. 111). ³² By the end of the first century ad headhunting was also associated with auxiliary troops, illustrated by scenes on Trajan’s column that show Gallic troops presenting the severed heads of slain Dacian warriors. ³³ The tombstones of two auxiliary cavalrymen buried in Britain in the late first or early second century also showed severed heads being presented, celebrating the importance of such trophies to the status and identify of soldiers stationed here. ³⁴

We don’t know if any Walbrook crania were discarded trophies, but Hadrianic wars offered opportunity for the practice. Whether they had been displayed or not, the unwanted remains needed expurgation. At Rome such remains were flushed away by the Cloaca Maxima and Tiber, and in London the Walbrook and Thames were ideally suited for the same purpose (above pp. 163 and 231). The comparison would not have escaped the attention of commanding officers some of whose careers had embraced responsibility for Rome’s sacred rivers before service in Britain. ³⁵ Each and every Walbrook skull boasts an individual history of post-mortem treatment, but collectively they show that the upper Walbrook was an appropriate location for the disposal of disturbed human remains. At the end of the first century the river may have helped define a city pomerium, an area favoured for the execution and burial of criminals. This marshy area lay north-west of the city, reinforcing an association with deathly endings at the setting of the sun. ³⁶ The looming presence of the amphitheatre added to the mortal aspect of an almost Stygian landscape. Crania littered the ditches and pits dug close to the river, concentrated in greatest numbers near places where roads bordered or bridged its tributary channels. Such places provided practical and meaningful platforms from which river-offerings could be made. In some cases, the crania had rolled along the river, before being dredged from its bed or banks for reburial nearby. It is possible that this dredging became necessary because the river was too slow-moving to flush the remains away, leaving them banked at bridges and bends. Might the river have been choked by hundreds of skulls?

The process of reburial shows that heads were accorded special treatment—not necessarily as trophies, but because these elements were more easily identified and recovered, and privileged for treatment as the seat of the soul. The other elements of the dismembered corpse were perhaps more widely dispersed, and it is possible that in some cases, bodies were burnt on collective pyres and only the heads retained.

The fort and its associated landscape suggest that London witnessed an increased military presence after the Hadrianic fire, coinciding with a significant increase in the ritual deposition of human crania in the upper Walbrook. Auxiliary cavalry would have been at the forefront in policing exercises, and London may have been influenced by headhunting practices normalized within the early second-century Roman army. Whilst the Walbrook skulls may have been obtained in reprisals that continued over several decades, and drew on practices that continued for the better part of two centuries, the exceptional scale of second-century evidence hints at a major Hadrianic slaughter.

Other irregular burials

Smaller numbers of irregular burials were found along London’s other borders, including three crania in the second-century fills of the ditch thought to form the Flavian town boundary. ³⁷ Excavations at Duke’s Place, beside the road to Colchester, uncovered two skeletons ‘unceremoniously dumped’ in a drainage ditch with pottery of c. ad 120. Skeletal fragments from at least three people were found in the infilling of the waterfront revetment on the Thames foreshore along with finds dated ad 80–130. ³⁸ At Paternoster Square, on the western borders of town, a small pit dug after the Hadrianic fire contained an upturned cranium with two mortaria. ³⁹ Disarticulated remains were also found in quarries at 16–17 Old Bailey near a small inhumation cemetery of c. ad 70–120 beside the Fleet. ⁴⁰ Others have been identified at sites with religious and ritual connotations In Southwark, including crania from two people of possible African origin from a second-century ditch beneath the cemetery at Lant Street. ⁴¹

The skulls placed in boundary ditches may have drawn on the Janus-like association of heads with gateways. ⁴² The circumstances of their deposition are consistent with the earlier disposal of the fragmentary remains of uncremated corpses in wet-places on the settlement margins. These practices favoured sites where intact bodies were subsequently buried, and some cemeteries may have come into existence at places known to receive the dead. There are grounds for supposing that the disarticulated bones came from bodies denied normal burial, which had provoked a necrophobic response. It is alternatively possible, however, that some bones were carried to site with soil removed from places where human remains were exposed, or that rites involving recycled body parts removed from earlier burials were involved. Richard Hingley has suggested that rituals of this kind might be linked to the acculturation of land through the selective use of anthropogenic material. ⁴³ This would not sit well with the idea that unburied human remains required expurgation, although we should not assume that all dead were equally threatening.

The rituals associated with trophy heads may have a bearing on finds of head-pots made in the upper Walbrook. There is a marked concentration of these vessels, shaped like human heads, in contexts dating from the late first century to c. ad 160. ⁴⁴ Gillian Braithwaite has shown that they were often linked with the presence of auxiliary troops recruited from the Rhine delta and northern Belgium, and potentially used as ritual offerings influenced by Bacchic ideas. Some resembled death masks, where the eyes were shown closed, accentuating links between natural and supernatural realms. This would suit rites of transition where the lifeless head represented the dead trapped in the world of the living. The process of violent exorcism once practised on skeletal remains may have transferred to these surrogates.

Some statues may also have been ritually decapitated and their heads cast into watery places, accounting for the presence of the bronze head of Hadrian in the Thames (see pp. 112, 163 and 306–7). We don’t know when this statue was dismembered and this is sometimes attributed to iconoclasts of late antiquity, but the desecration might equally have been associated with the events of the mid-120s. The imperial image was a target for attack during riot and revolt, and their heads thrown into rivers in violent desecration analogous to the abuse vested on enemies of flesh-and-blood. ⁴⁵ The decapitation of the imperial image could have symbolized a rejection of Hadrian’s authority.

These diverse strands of evidence buttress the argument that London’s destruction c. ad 125/6 was the consequence of hostile action. This was followed by swift and emphatic restoration, including the construction of a new fort, the building of roads to aid supply and communication, and a spike in the number of irregular burials involving features suggestive of military retribution. These details mirror those associated with London’s military re-occupation after the Boudican revolt.