Death and Social Division at Roman Springhead

DEATH AND SOCIAL DIVISION AT ROMAN SPRING HEAD MALCOLM DAVIES The cemeteries at Springhead have not been well studied; exploration of the site, much damaged in recent years by road construction, has concentrated upon the sanctuary, most likely a Roman development of a native site. In connection with the presence of wealthy burials, apparently of early third-century date, it should be noted that the sanctuary was systematically rebuilt in stone in the late second century. 1 While this rather succinct observation by Walker fairly summarized the imbalance of the archaeological activity at Springhead prior to 1990, it requires review in the light of three developments that have taken place in the last twelve months. An interim account has now been published of the Romano-British cemetery, about 0.5km southeast of the temple site, excavated by the Oxford Archaeological Unit in 1998-9 in advance of the CTRL.2 Related to the same project, Wessex Archaeology has been carrying out a large-scale excavation of the area to the north of the A2, part of which is adjacent to the temple site, which will surely throw considerable light on the origins and surroundings of Roman Springhead. 3 Lastly, a small team, carrying out a geophysical survey of the Scheduled Ancient Monument sites at Springhead in 2000-1, supported by English Heritage, has identified the precise location of the enclosed cemetery which was excavated by Rashleigh in 1799-1802. 4 The Revd Peter Rashleigh, rector of Southfleet, uncovered the Romano-British cemetery that lay within a stone enclosure, about 18m square. It proved to be a high status burial ground containing two inhumations and at least six cremations with exceptionally rich grave goods which included a stone sarcophagus, two lead coffins, two large glass urns, gold armlets and a ring, a gold forehead-pendant and a pair of elegant woman's shoes. This report describes the original discovery of the enclosed cemetery, the subsequent confusion about its precise location and recent 157 MALCOLM DAVIES MAP OF 5PRINGHEAD SHOWING MAIN ROMAN FEAT!JRES .... "DMAN WAI.LED Cl:!/\1E'TE.􀀜y (.SOL.&'. FIE.J..t>) Map I Sketch-map of Springhead locating the main Roman sites in relation to modern-day features. 're-discovery' just south of Pepper Hill roundabout; examines the dating evidence for this burial ground in relation to the chronology of the main cemetery and the settlement itself; and considers the implications of two neighbouring, but separate, cemeteries at Springhead which reflect a clear-cut social divide at this Romano-British settlement. Map 1 shows the locations of the two cemeteries in relation to the main settlement. The location of the enclosed (walled) cemetery When Rashleigh reported the discovery of the enclosed cemetery and outlined the results of his excavation in letters to the Society of Antiquaries in 1801 and 1802, he stated that the site was located in 'Sole field'. Details of the exact position were lost subsequently. 158 DEA TH AND SOCIAL DIVISION AT SPRINGHEAD Mortimer Wheeler erroneously placed Sole field to the west of the Gravesend railway line and the temple site on his map of Springhead. 5 As a result, when in 1964 W. S. Penn, fresh from his discoveries of the Roman-Celtic temple at Springhead, was called to Pepper Hill roundabout where road-widening had revealed Roman wall foundations, he identified these as part of the temenos of a further temple. 'It was extremely large, measuring some 133.5 x 119m with 3ft walls of flint rubble, faced with Kentish rag. The temple, which was centrally placed, was identified by Penn after a brief excavation' .6 Subsequently, a sketch map was produced and the field was declared a Scheduled Monument site. However, when preparations for a geophysical survey were begun in January 2001, Wheeler's error became clear from the County monuments record. A check on the Tithe map confirmed that a 'Soul field' was located south of Pepper Hill roundabout, just to the east of New Barn Road. The resulting resistivity plot was unusually clear and indicated a walled enclosure, about 18m square, surrounded by another much larger walled enclosure which extended across New Barn Road (Figs 1 and 2). This was clearly the feature that Penn had identified in 1964, but the inner enclosure was, as far as could be Fig. 1 Resistivity plot of the inner and outer enclosures of the walled cemetery straddling New Barn Road, surveyed in January 200 I. 159 - -- ovor-.,..___d pOVeY li 'IWe$ Scale 1:12so #,,# --#-...... , ,... , . , . . , , ' ' ' , ' . ' ' • I ,. I ., , ,, ,.., -......... _____ _ 0 Pig. 2 Plan of the W•lled cemerery at Peppe, HHJ b,sed oa the re,i,t;vity survey. 160 ... Cl, :s C, DEATH AND SOCIAL DMSION AT SPRINGHEAD 􀀃 A • ! . •.... ..... - •--􀀻•-••!-·"' -••·••••--•••••·•··--•,$7 lO-. -••-•-•w-•••···•􀀼·-- .-·••••·•-•· .. -.. . ·􀀽-􀀾- Stone tomb enclosing 2 lead coffins. Ragstone pavement containing glass amphoraa etc Fig. 3 Rashleigh's plan drawing of the (inner) walled enclosure (to which some descriptive notes have been added). judged from the resistivity plot, the shape and dimensions of the plan of the enclosed cemetery drawn by Rashleigh in 1803 (Fig. 3). It was in the named field and had the same dimensions and shape. It was not particularly surprising that Rashleigh had failed to find the outer enclosure, given the fact that the two were at least 30m apart. Nor was it surprising that Penn should misinterpret the feature in his brief investigation. He had no reason to believe that Rashleigh' s cemetery was to the east of Springhead. The inner and outer enclosures which Penn found do resemble the plan of a RomanoCeltic temple where a square cella is frequently set, slightly offcentre, within the enclosing walls of the temenos. To remove any doubts about the matter, it was agreed with English Heritage to expose the south-western corner of the inner enclosure. If 161 MALCOLM DAVIES this showed the presence of a large stone buttress, as featured in Rashleigh' s plan (Fig. 3), then it would prove beyond reasonable doubt that the feature in Sole field was indeed the Romano-British enclosed cemetery excavated in 1799-1802. In March 2001 a resistivity meter was used to relocate the corner of the inner enclosure and an area of approximately two square metres was excavated. At a depth of 34cm the corner buttress was revealed. The interior of the walls and buttress was composed of chalk rubble, flint and mortar. The inner and outer faces of the walls and buttress were of dressed Kentish ragstone. The shape and dimensions of the buttress and walls corresponded very closely to those of Rashleigh's drawn plan and left no doubt that the feature uncovered was the walled enclosure which he had excavated (Plate I). Consideration of the dating evidence Having established that the monument feature in Sole field is the PLATE I The south-west buttress of the walled cemetery, excavated in March 2001. The near measuring pole is laid across the 5ft buttress. The further pole is laid across the south wall of the cemetery and points along the line of the western wall. 162 DEA1H AND SOCIAL DMSION AT SPR.INGHEAD walled cemetery excavated by Rashleigh and that it is sited no more than c. 200m from the communal cemetery for Springhead, recently excavated by the OAU, it is important to take a fresh look at the dating evidence for the two cemeteries and the Springhead settlement itself. The ceramic evidence from the communal cemetery indicates activity in the first century, rising to a peak in the second, with a reduced presence in the third and no usage in the fourth.7 As for the settlement, it needs to be said at the outset that there is a dearth of secure dating evidence even after Penn and Barker's extensive work there between 1951 and 1973. This is an area that has been heavily ploughed, robbed for stone and greatly disturbed in places by post-medieval field drains. There is also a total absence of sealed stratified evidence of occupation for the fourth and fifth centuries. This uncomfortable fact is interpreted differently by Penn and Harker. Penn writes of the fourth century: 'it is difficult to be sure how intensive occupation was during this century' and of the fifth he says, 'what is certain is that life carried on ... '. 8 Harker, for his part, strikes a more cautious note: 'numerous coins and a considerable quantity of pottery can be firmly dated [to the fourth century] but in the area so far excavated, buildings seem to have been confined to insubstantial wooden structures' .9 In assessing unstratified finds of late coins and pottery Penn may have given insufficient weight to the presence of Watling Street as a possible source. A semi-ruined roadside settlement would have had attractions for both inquisitive visitors and squatters. The very limited evidence for the fourth and fifth centuries falls a long way short of suggesting continuity of occupation, while the presence of squatters inside Temple I around the middle of the fourth century clearly demonstrates that Springhead had ceased to fulfil its previous religious function by that time. 10 However, this evidence does not tell us when Temple I ceased to function as a temple. The latest, secure evidence of occupation from the temple site comes from two separate locations. Firstly, within 'Temple V' six coins, datable to the last half of the third century lay on a clay floor, on which plaster had fallen. 11 Secondly, just inside the porch of Temple I, 'a loose layer of tiles, including many tegula, was used to repair the porch floor. This sealed a single coin of Carausius' .12 There is, thus, a case for believing that the temples were continuing to function at some level up to the last quarter of the third century, although the rough and ready repairs to the floor suggest much reduced standards. There is, however, no secure evidence of continued temple usage after this time. Thus the evidence from both the communal cemetery and the 163 MALCOLM DAVIES excavated settlement indicate a Romanised settlement which was established in the first century, reached a peak of prosperity in the second century and declined sharply in the third, probably ceasing to exist as an organized religious centre by not later than the last half of the third century. (This assumes, reasonably it is believed, that there is no late settlement cemetery yet to be discovered). In accord with this interpretation is the fact that the Springhead settlement, which lay open to approach from the Thames estuary, shows no evidence of late third-century defences against the increasing threat from Saxon raiders. 13 The dating evidence for the walled cemetery is rather more complex. Firstly, we have the evidence of a small number of artefacts associated with the burials. These were deposited with the British Museum (in 1837 in most cases) and have been expertly studied subsequently. Secondly, we have the stone sarcophagus, which contained two cremation vessels, and the stone tomb which enclosed the lead coffins. Thirdly we have the character of the stone enclosures which suggests that the cemetery was planned and built as a unified structure, at one time. Among the artefacts was a samian plate stamped 'Graniani', which can be dated to the middle of the second century when the potter Granianus was working at Lesoux, Central Gaul. The finer of the two large glass ampboras is firmly dated to the early years of the second century. The other glass urn is first- or second-century. The pair of shoes, purple-dyed and worked with gold-leaf, can be dated to the third century on grounds of style, use of colour and gold leaf and its complex system of fastening.14 The two gold snake-bracelets, gold forehead-pendant and gold ring, all found within the one lead coffin, do not allow of close dating. The bracelets were in fashion from the first to the early third centuries, while the pendant, the only example found in Britain, 'would appear to have been current over a long period of time in the Empire' . 15 The stone sarcophagus, which contained the two glass amphoras and the third-century shoes has been dated by Walker to about the beginning of the third century and the likelihood is, therefore, that this burial was early third-century, which would mean that the two glass amphoras were either antiques or were reburials from an earlier cremation. There is one final piece of evidence to be considered in assessing the chronology of the walled cemetery. Rashleigh noted the type of stone used in carving the sarcophagus was also used in constructing the tomb in which the lead coffins were laid, a material he described as 'roe' stone. This sarcophagus was later classified as a carved 164 DEA TII AND SOCIAL DNISION AT SPRINGHEAD oolitic limestone chest, 'likely to be a provincial version of metropolitan work of about Severan date' . 16 Quite separately, Blagg examined pieces of Roman decorated stonework from the Springhead temple precinct which included fragments of column shafts and a Corinthian capital, both of oolitic limestone. His conclusion was that the column fragments 'were part of a free-standing votive column, probably erected in the second half of the second century or early in the third' and that the piece of Corinthian capital 'should probably be assigned to a mason from Eastern Gaul' _17 Thus we have evidence of the extensive use of stone, including carved oolitic limestone, in the rebuilding of some of the temples at the close of the second century, being paralleled with an extensive use of stone and carved oolitic limestone (the latter dated to the end of the second century) in the construction of the enclosed cemetery. Without being conclusive, it certainly suggests the strong possibility that both developments took place more or less concurrently at the end of the second century when a skilled mason, the Kentish ragstone and the oolitic limestone must all have been present on site. This would certainly not be incompatible with the artefacts in the enclosed cemetery either being roughly contemporary to the Severan period or earlier. The clear symmetry of the entire cemetery and the burials also lends support to the interpretation that the enclosed cemetery was laid out at one time and not added to in the later third century or subsequently. The walled cemetery seems to have been in the nature of a family/group memorial, set up in the early third century, in contrast to the communal cemetery which gives every sign of having been in regular use over several centuries. To summarize the chronology of the two cemeteries and the settlement: the communal cemetery faithfully reflects the growth and decline of the Springhead settlement from the late first century to about the last half of the third century. The walled cemetery, however, appears to have be built in that form in the early third century but may contain some re-burials from the second century. The dissimilarities of the two cemeteries Comparing the two neighbouring cemeteries, what is striking is their extreme dissimilarity. If we look at the density of their respective populations, the communal one contained 561 burials (326 inhumations and 235 cremations). 18 The walled cemetery contained upwards of six burials (2 inhumations and at least four cremations). But in terms of land usage the picture is quite the reverse. The walled cemetery, based on Penn's survey, covered 15,886 square metres 165 MALCOIM DAVIES while the communal cemetery occupied a mere 1,748 square metres. Again, looking at the plan of the communal cemetery the impression gained is of a largely unplanned burial ground, squeezed in between a road and some boundary ditches. where later graves are cut into earlier burials and where the orientation of the graves broadly follows the line of the road. This contrasts sharply with the planned symmetry of the walled cemetery, where the inner and outer enclosures are part of one whole, where the sarcophagus is laid exactly in the centre of the inner enclosure and where both the lead coffins and the sarcophagus are oriented east-west. The sites of the two cemeteries are also markedly different. The enclosed cemetery, as Rashleigh observed, adjoined Watling Street and, with its prominent and extensive walling, would have caught the eye of every traveller (Map 1). The communal cemetery, on the other hand, is situated on a secondary road, and was probably not visible from Watling Street. 19 An examination of the grave goods from the two cemeteries reinforces the picture which is emerging that they served two different social groups. In the communal one, the bulk of the 628 pottery vessels were locally manufactured, mainly from Thameside and north Kent. Other grave goods included coins, traces of hob-nailed footwear, 20 copper-alloy brooches, 3 finger rings and other individual copper and iron artefacts. Gold and silver, however, were conspicuously absent. Luxury goods were limited to 12 samian vessels and one colour-coated vessel from Cologne. Inhumation was mainly in wooden coffins.20 In the walled cemetery the large proportion of high-status goods and burial chests among so few burials sends a very emphatic message about the power and prestige of the occupants and of the surviving family. The separation of this enclosed cemetery from the main burial ground surely tells us something about the relationship of this family/group to the rest of the settlement, in life, as well as death. For this was not a rich burial within a community's burial ground. The enclosed cemetery was constructed by a family/group that chose to distance itself from the rest of the community and to set up its own memorial in its own style and on its own land. The comparison gives a very clear picture of two separate social classes existing side by side, the one apparently rich in land, possessions and trade connections, the other possessing the merely comfortable life of a small, prosperous Romano-British settlement. Questions raised by the walled cemetery The special character of the enclosed cemetery raises a number of 166 DEATH AND SOCIAL DMSION AT SPRINGHEAD points. Why did it occupy such an eye-catching location? Why did its architect seek to make such a conspicuous display of wealth, with extensive use of metre-thick stone walls? In this connection, it is worth recalling that there are at least thirteen other examples of walled cemeteries in Britain, all of them south-east of the Fosse Way. The majority are in Kent, with four sited along Watling Street.21 Some of these enclosed cemeteries, like the one at Borden, appear to be high status burials and may also be family burial grounds. Their conspicuous siting and structure suggests that they may have been a way in which leading families chose to display their wealth and power. The re-discovery of the precise site of this enclosed cemetery, so close to the Springhead temple site and to the main cemetery for the settlement, also raises a number of practical questions. It would be reasonable to assume that the family/group that built the private walled cemetery would have lived in a villa (in the vicinity of Springhead) or town house of a quality reflecting their status. To date, no candidate for this home has been found, unless the Northfleet villa is such. Secondly, if as seems likely, this dominant family/group died out or moved to a more secure area at the beginning of the third century, what effect may this have had on the settlement? Finally there are some puzzling aspects to the construction of the enclosed cemetery. Why was the outer enclosure built? Was this a concept borrowed from the structure of Romano-Celtic temples in order to indicate the presence of sacred ground? Why did the walls of the inner enclosure require external buttresses? Did they support roofing for a walkway or were the 'buttresses' actually bases for funerary stonework? Why does the southern wall on Rashleigh's plan show a pronounced thickening at its eastern end? Was the approach to the cemetery from Watling Street or via the hollow way on the west side? Regarding the question of the possible relationship between the family/group that built the walled cemetery and Springhead itself, the following is suggested. We know that the Springhead temple or sanctuary site underwent a major rebuilding in stone in the later second century, which would have required considerable planning, organization of resources and provision of capital to convert it to what became one of the most important religious centres in Roman Britain. As Alec Detsicas himself commented: The temple area is striking for the symmetry of its internal arrangements, at least from the middle of the second century. The traveller or worshipper, on entering the compound through the entrance building on its higher level, would have been confronted by two altars or columns, then two temples, almost identical in size and appearance, flanked by ancillary buildings.22 167 MALCOLM DAVIES Is it not likely that the leading local family built both the largest walled cemetery site in Roman Britain and provided the wealth and entrepreneurial spirit behind the success and expansion of Springhead in the second century? ACKNOWLEDGEMENTS An archaeological report is made with many hands. Without the support of Martin and Matthew Gedney, the landowners and Peter Kendall of English Heritage, the investigation could not have been undertaken. All the resistivity work was done by a small team consisting of Robin Grimes, David Pendleton, Karen Roberts and Bill Wyatt, a volunteer from the Gravesend Historical Society. Much of this work was done under the unfriendly conditions of the winter of 2000-2001. Victor Smith helped with the preliminary survey work and the site was finally surveyed and mapped by Siriol Davies and Robert Jackson. Robin Grimes gave able support in the excavation of the inner enclosure corner. Chris Hasler supplied the software know-how for the resistivity plots and also proof-read the report. The writer received very considerable support from Helen Glass of Rail Link Engineering and also Lis Dyson of KCC Heritage Department. NOTES 1 Walker, S., 1990, Catalogue of Roman Sarcophagi in the British Museum, London, 57. 2 Glass, H., 1999, 'Archaeology of the Channel Tunnel Rail Link', Archaeologia Cantiana, cxix, 205-8; Union Railways (South) Ltd, 2000, 'Springhead Roman Cemetery', Current Archaeology (May), 458-9. 3 The possibility of Springhead having Iron Age origins is a subject which apparently received little attention from the original site director, W. S. Penn. His successor, R. S. Harker, restricted himself to a one-sentence comment which makes a significant claim but he does not support the statement with archaeological evidence: 'A complex system of eady ditches and pits of votive character has been found to underlie the Roman levels, clearly pre-dating them and strongly suggesting that here was an important religious sanctuary of the Belgic Britons'. Harker, S. R., 1980, 'Springhead: a brief appraisal', in (Ed.) Rodwell, W., Temples, churches and religion in Roman Britain, 285-8. However, an unpublished paper by French provides the detailed archaeological evidence justifying Harker's claim and confirms that Harker had indeed found extensive evidence of Iron Age settlement on the Springhead temple site dated to between c. 50 BC and the Conquest. The substantial quantities of pottery recovered from the ditches and post-holes unequivocally indicated a late Iron Age chronology. It was also clear that his excavations had revealed a possible Iron Age precursor to the Roman temples on the same site. French, D. A., 1984, 'Springhead: an examination of the pre-Roman origins and earliest occupation' (unpublished report for BA degree, University of London). 168 DEATH AND SOCIAL DNISION AT SPRINGHEAD 4 Rashleigh, P. A., 1808, 'Account of antiquities at Southfleet', Archaeologia, xiv, 37-9; 'Account of a further discovery of antiquities at Southfleet', ibid., 221-3. 5 Wheeler, R. E. M., 'Romano-British Kent', in Kent (Victoria County History, iii), 90. 6 National Monuments Record, TQ 67SW6, 13, KCC. 7 The report by the OAU concludes: 'On present ceramic evidence, the overall chronology of the group extends from the late 1 st century AD to about the mid 3rd century, or possibly a little later but there are very few vessels which need have been of third-century date'. Union Railways (South) Ltd, 1999, Waterloo Connection, Northfleec ARC NBR 98: Detailed Archaeological Works Interim Report. 8 Penn, W. S., c. 1968, The Roman town ofSpringhead, Gravesend Historical Society, 9. The reference to the fifth century seems to be based on unstratified coins and the discovery of a Valentinian I coin, AD 364-375, sealed beneath road R9 which branches north of Watling Street towards Thameside. Penn clearly sees the road as evidence of fifth-century occupation but this is not proven. The road may well have had some other function. Penn, W. S., 1968, 'Springhead: miscellaneous excavations', Archaeologia Cantiana, lxxxiii, 166. 9 Harker, S. R., c. 1973, Vagniacae - The Roman town at Springhead, Gravesend Historical Society, 8. JO Penn, W. S., 1959, 'Excavation of Temple I', Archaeologia Cantiana, lxxiii, I I. 11 Penn, W. S., 1962, 'Springhead: Temples II and V' ,Archaeologia Cantia11a, lxxvii, 117. 12 Penn 1959, op. cit. (see note 10), 10. 13 The double ditch, which may have surrounded the entire settlement, has been shown to be of first-century origin. Penn, W. S., 1965, 'Springhead - map of Discoveries', Archaeologia Cantiana, lxxx, I 16. This point is also supported by French (see note 3). The absence of any defences at Springhead contrasts sharply with the construction of the town walls at Canterbury and Rochester during the third century. The combination of Springhead's vulnerability to a seaborne attack and its limited resources as a minor settlement would probably have led its inhabitants to seek the security of nearby, walled Rochester. 14 Walker, 1990, op. cit. (see note I). 15 Johns, C., 1996, The Jewellery of Roman Britain, 135-7. 16 Walker, 1990, op. cit. (see note 1). 17 Blagg, T. F. C., 1979, 'The Votive Column from the Roman Temple Precinct at Springhead', Archaeologia Cantiana, xcv, 223-29. 18 Glass, 1999, op. cit. (see note 2), 205. 19 It is noteworthy that an early, small cemetery, datable to the first century, was uncovered alongside this same road, but closer to the settlement (see Map I). This was perhaps a predecessor of the main communal cemetery. Philp, B. and Chenery, M., 1996, A Roman site at Springhead (Vagniacae) near Gravesend, KARU Special Subject Series, No. 9, 8-12. 20 Glass, 1999, op. cir. (see note 2), 207. 21 Jessup, R. F., 1959, 'Barrows and walled cemeteries in Roman Britain' ,Journ. Brit. Archaeol. Assoc., xxii, 1-32. 22 Detsicas, A., 1983, The Canriaci, Gloucester, 75. 169