Romano-British West Kent

ROMANO-BRITISH WEST KENT A.D. 43-100 By BRIAN J. PHILP INTRODUCTION IN attempting to place the Romano-British farmstead at Eastwood in its proper context a short study of similarly dated sites in the area of present-day West Kent was undertaken. As this revealed a number of interesting points the publication of a map (Fig. 1) and a comprehensive list (Fig. 2) of these sites was both apposite and desirable.1 The area studied lies west of the river Medway, excluding the Maidstone district and land north of the Thames. This covers about 300 square miles in which 30 sites have been recorded.2 Doubtless others have been found, but are unrecorded or obscurely published ; others probably have been missed. Twenty three of the listed sites were found and/or excavated in the last 15 years (1946-61). This largely reflects the increase in the number of competent local archaiologists working in the area in postwar years and in particular their more enlightened approach to the subject. CONSIDERATION OE THE EVIDENCE The Pre-Roman Iron Age Evidence of pre-Roman Iron Age settlement in the area of West Kent is not abundant. Hillforts of this date certainly occur at Charlton, Keston, Oldbury, Westerham and perhaps Lullingstone, but traces of occupation sites are few. Crayford3 and Dartford4 are two of the better-known sites and others seem to have existed at Greenhithe (No. 8) and Ramsden5 (No. 18). An Early Iron Age farmstead of the Little Woodbury type may exist at Fawkham.6 A few other, less significant, finds have been recorded. Belgic coins and pottery have been found in the area, but not in any quantity and it seems unlikely that West Kent came under direct Belgic influence much before about A.D. 10. It was then that the 1 The need for local surveys of this type was stressed at the Conference on Romano-British Villas, in 1956. A.N.L., 6. No. 2. 2 Single finds have not been included. 8 P.P.S. (1938). 4,161. * P.P.S. (1941). 7,134. 6 Arch. Cant. (1956). LXXI, xlvii. » Arch. Cant. (1941). LTV, 74. 74 ROMANO-BRITISH WEST KENT A.D. 43-100 region probably came under the control of the Catuvellauni, expanding southwards from Hertfordshire and Essex. Indeed the available evidence suggests that the region was not settled to any large extent much before the Roman Conquest in A.D. 43. Even Caesar's normally detailed chronicle7 dispenses with his journey through West Kent in 54 B.C. in a short sentence and this has been held by some to reflect the lack of noteworthy settlement at that time. The picture of East Kent in Belgic times is, however, quite different. Certainly coins and pottery of the period are much more abundant there. The cemeteries at Aylesford and Swarling and the occupation sites at Bigberry, Canterbury and at many other places suggest a concentrated area of Belgic settlement in East Kent. The distribution of coins8 dating c. 15 B.C.-A.D. 25 certainly suggests an important political division between East and West Kent at that time. Considered with the comparative lack of Belgic material in West Kent, this supports the view that Cantium was entirely East of the Medway. Indeed this view better accords with Caesar's description of it as a " purely maritime district ".9 The Geographical Distribution Whilst acknowledging the limitations of distribution maps certain general conclusions can be drawn from a study of the 30 sites in West Kent. It seems, for instance, that the high ground was generally avoided, the majority of sites being situated below 200 feet. Only five sites are higher than 300 feet and only one of these above 450 feet in an area where land reaches to a height of over 750 feet. 16 of the sites lie along the Cray and Darent valleys and suggest easier natural routes and better land. Another five (Nos. 7, 8, 10, 26, and 27) approximate to the man-made Watling Street, one at least (No. 7) being a road-settlement. It is clear that the well-drained gravels, sands and chalk areas were selected in preference to the heavy clay lands supporting dense forest. Clearly with farming as a main industry well-drained, easily-turned soils on sites near major routes and a reliable water source were to be preferred. The sites at Cliffe (Nos. 4-6) probably existed to exploit the marsh clay for the manufacture of pottery. Roads and Centres Two main Roman roads are known to exist in the area. The more important, Watling Street, certainly dates from the Conquest in A.D. 43. It crosses the region as part of a major route between the Channel ports 7 The Conquest of Gaul, V. 2 8 R. P. Mack, The Coinage of Ancient Britain (1953), 86, map 14. 9 The Conquest of Gaul, V. 2. 75 ROMANO-BRITISH WEST KENT A.D. 43-100 and London and the rest of Roman Britain (as it still does today). Certainly the existence of this important liighway must have helped greatly the development of the region. Little is known of the other road, running from London to Lewes, Sussex, though a first century date is possible. The prehistoric North Downs Ridgeway10 was in use as a track for most of the Roman period. Doubtless other minor roads and tracks existed too. Recent excavations11 have shown that Springhead (No. 7) became established as a road-settlement at the time of the Conquest and that there was abundant activity there during the second half of the first century A.D. Similar settlements may have existed at Crayford and Dartford where the Watling Street crossed the river Cray and the river Darent ; perhaps at Greenwich too ? These road-settlements, as also the rapidly expanding township of London, would be focal-points for the trade of much of the region. In their markets produce from the farms could be sold or exchanged for a variety of goods, both local and imported. Agriculture and Industry The evidence suggests that 23 of the listed sites were small farmsteads and from this we may assume that the main activity of the region was agriculture. Little is known about the type of farming though doubtless it was, to a large extent, of a " mixed " nature. Certainly at Eastwood oxen and sheep were reared and the presence of querns on at least three sites suggests that corn was grown. It is likely that each farm was self-supporting with any surpluses being sold in local markets. The large amount of coarse native pottery found in the area suggests that local kilns existed. On the marshes at Cliffe evidence has recently been found of industrial activity, very probably relating to potterykilns. Close by, at Higham, other probable kiln-sites were found in the nineteenth century.12 The site at Otford (No. 16)13 appears to have had a small kiln operating during the first century. That pottery was ever manufactured at Patch Grove (No. 17) has yet to be positively estabhshed. Population An attempt to estimate the population of the region during the second half of the first century A.D. must be largely based on conjecture. It seems almost certain that the farmsteads were worked by family 10 I. D. Margary, Roman Ways in the Weald, 23. 11 Arch. Cant. (1957). LXXI, 53., Arch. Cant. (1958). LXXII, 77. 12 Jnl. B.A.A. IV, 393. 18 Arch.Cant. (1930), XLVH, 157. 76 SPRINGHEAD tm KEY ROMANO-BRITISri SITES ROMAN ROADS ' LAND OVER 750 FEET LAND OVER 500 FEET LAND OVER 250 FEET MARSH (PRESENT TIME) COUNTY BOUNDARIES (PRESENT TIME) 1 "> 30 ROMANOBRITISH WEST KENT A.D. 43-100 0 1 2 3 SCALE MILES FIG. 1. ROMANO-BRITISH WEST KENT A.D. 43-100 units. Parents and sons, the latter with their wives and children, probably constituted the typical farm family, numbering perhaps only some 8-15 persons. Certainly these figures tend to be supported by the small size of the family cemeteries attached to these sites. Of the 30 listed sites perhaps 23 represent farmsteads. Allowing for a similar number having been already destroyed or missed and a slightly larger number yet to be found, some 100 such sites may originally have existed in the area. An average of 12 persons per farmstead then produces a total of 1,200 people. The populations of the roadsettlements (certainly at Springhead and perhaps also at Crayford, Dartford and Greenwich) and industrial sites must then be added. Thus, with due allowance for itinerant traders and others not classified above, an estimate of about 2,500-3,000 people results. Even supposing the whole basis of this calculation to be incorrect a total of no more than 5,000-6,000 would apply. Certainly, on the evidence, the figure of 10,000 seems grossly excessive. Structures—Native So far as native structures are concerned very little evidence has survived at any of the sites. Certainly structures must have existed. At only five sites (Nos. 13, 23, 24, 26, and 30) have traces of huts been found. At Kings Wood (No. 30) two circular huts about 22 feet across were noted, but details of the other sites are not yet available. It seems clear from this fairly negative evidence that these structures were not of a substantial nature, merely huts of wood and clay following closely the Iron Age pattern. The general lack of building rubble and tile on these sites supports this view. Structures—Roman At only three sites in the region have buildings of Roman type of first-century date been identified. At Springhead the buildings relate to the road-settlement whilst at Cobham (No. 2) and Lullingstone (No. 15) early villas have been found. At Cobham a rectangular structure, about 92 by 32 feet, comprising of a corridor and five rooms was built about A.D. 100. The footings were flint and chalk set in clay under a timber-framed superstructure. At LuJlingstone another reotangular struoture, about 92 by 39 feet, comprising perhaps two corridors and several rooms (one deep) seems to have been built at about A.D. 90. The lower parts of the walls were flint and mortar, but the superstructure was probably timber-framed. In each case native pottery, suggesting earlier farmsteads, was found beneath the floors. Perhaps here we have evidence of official policy encouraging the native population to adopt the Roman way of life.14 14 Taoitus, Agricola, 21. 77 ROMANO-BRITISH WEST KENT A.D. 43-100 At nine other sites Roman buildings eventually replaced the farmsteads, but at 16 others nothing more pretentious was ever built. Pits and Ditches At the majority of the sites pits and ditches are constant features. The pits were probably for storage or rubbish and their dimensions vary considerably. The ditches, normally between two and six feet deep, often enclose the living and working areas. At Eastwood (No. 1) Kings Wood (No. 30) and Bedens Field (No. 24) areas of between one and one and a half acres were enclosed. These ditches frequently contain quantities of rubbish and sometimes the only stratified deposits on the sites. In this respect their contents are of the first importance. Cemeteries I t seems clear that the dead were cremated and buried in small, flat cemeteries, the burnt bones being placed in urns. This practice was a continuation of a Belgic tradition. Seven of the thirty sites have produced such cemeteries. All are small and none contains more than a dozen burial-groups. Five of these were certainly associated with occupation-sites and at Kings Wood (No. 30) five burials were found just a few feet outside the farm enclosure. It seems clear that these were private cemeteries where members of the family were buried. From this each farmstead site must have had its own small cemetery which, in the majority of cases (including Eastwood), have not been found. At two sites (Nos. 4 and 28) only the cemeteries seem to have been found, but not the related settlements. Each burial normally consists of several pottery vessels ; the usual number being three. Of these one, usually large, contains the cremated bones. A platter and a small cup or flagon, with food and drink for the after-life, often completes the group. Samian vessels frequently occur in these burial-groups. Animal Bones Most of the Hsted sites must have yielded several, if not many, animal bones. In very few cases has any attempt been made to record these. Clearly had these been studied our knowledge of the livestock of this agricultural region would have been considerably enhanced. Only at Eastwood have the bones been studied15 in detail. There the bones of oxen, sheep, pigs, horses and a dog were recorded ; the majority being of oxen and sheep. Pottery Coarse pottery has been found on all of the listed sites and is the 18 Under the Bone Research Scheme. 78 ROMANO-BRITISH WEST KENT A.D. 43-100 most prolific of all the finds. The majority of this probably represents native or Romanized products. The native pottery is often soft and under-fired like the so-called Patch Grove wares. Some is harder or has shell or grit(?) mixed in the paste. The Romanized pottery is normally of a hard, sandy texture. The bead-rim is popular in many types, and cordons, reflecting Belgic traditions, are common. Finer wares were certainly imported into the region. Gallo-Belgie vessels and even a little St. Remy ware have been found in the area. Samian, again imported, has been found on at least 21 of the listed sites but, other than at Springhead, not in any quantity. Springhead has produced the only Samian of Claudian date. Coins, Brooches, Glass etc. Again, with the exception of Springhead, comparatively few Roman coins of first century date have been found on the sites, just as might be expected of an agricultural region. Brooches have been found on several of the sites and provide useful dating evidence. Glass vessels are infrequent and window-glass does not appear to have been recorded. Quern-stones have been found on only a few sites though it seems likely that most farms would have had them. Many probably came from not too distant quarries, but certainly one at least, from Eastwood (No. 1), had been imported from the Rhineland. Oyster shells are recorded from Eastwood (No. 1) and Keston (No. 29). Doubtless these have also occurred at a number of the other sites. These shells provide early evidence of the oyster industry, still active today. Tacitus,16 writing in A.D. 97-8, records that in Britain oysters were collected from the open shore. Large quantities of these probably came from the Thames estuary. SUMMARY A study of the 30 sites reveals the preference for well-drained and hght, easily-turned soils. Generally the higher ground was avoided. The natural valley routes of the Cray and Darent and the man-made Watling Street all attracted settlement. The majority of the sites were probably farmsteads and it seems that this was essentially an agricultural area. Farming was most likely of a ' mixed' nature ; at one site oxen and sheep were reared. Pottery was probably manufactured at Cliffe on the Thames estuary and perhaps also on a small scale at Otford in the Darent valley. The population of the region during the second hah0 of the first century A.D. has been estimated at about 2,500-3,000 people ; certainly a total of more than twice this number seems most unlikely. 16 Agricola, 12. 79 ROMANO-BRITISH WEST KENT A.D. 43-100 One major route, later known as Watling Street, ran through the district from the time of the Conquest as may have the less well-known London-Lewes road. A road-settlement soon became established at Springhead which, with others probably developing at the same time, would have served as trading centres for much of the area. Markets in these centres would promote the economic development of the region by absorbing farm surpluses and providing access to industrial products, both local and imported. Pits and ditches have been found on most of the sites, the former being for storage or rubbish. The ditches often enclose the farmsteads and sometimes delimit an area of one to one and a half acres. The ditches frequently contain important stratified deposits in the form of domestic rubbish. At least seven sites had small cemeteries attached to them and it seems likely that most of the others would have had them too. In effect these seem to have been small, private cemeteries where the cremated bones of the members of the family were buried when they died. Pottery is usually the most prolific of all the finds from the sites. Much is coarse, soft native-ware with some affinities to earlier Belgic pottery. Harder, Romanized products and, to a lesser extent, imported Gallo-Belgie, St. Remy and Samian wares also occur. As might be expected of an agricultural region comparatively few Roman coins have been found. Brooches, glass, quern-stones, oysters, ironwork and other small-finds have been recorded. Of native structures on these sites there is little evidence. The apparent lack of tile and building materia] of specific first-century date suggests that these were not of a substantial nature. Probably they resembled farms of the pre-Roman Iron Age. Only at Springhead, Cobham and Lullingstone have first-century Roman buildings been found. The former relate to the road-settlement, but the last two were early villas which had replaced native farmsteads on the same sites by A.D. 100. Both were simple rectangular structures comprising corridors and several rooms. The footings were masonry and the superstructures timber-framed. The evidence suggests that the region was opened up and developed in the decades following the Conquest. The comparative lack of Belgic occupation-sites of pre-Conquest date, so far as have been recorded, seems to support this though some of the farmsteads could have originated a decade or so before A.D. 43. Certainly the advent of Roman rule would have proved a stimulant to the native population, with Watling Street and its road-settlements providing the all-important trade links. Roman pottery, brooches, glass and coins began to appear on native farmsteads and by the end of the century villas of normal 80 ROMANO-BRITISH WEST KENT A.D. 43-100 Roman type were being built. Certainly by A.D. 100 the officially encouraged Romanization of the whole region, inevitably a slow process, was well under way. FUTURE WORK We have seen that 30 sites, of second half of the first-century date, have so far been recorded in West Kent. Judging from recent work it seems that this number may be doubled in the next 20-30 years. Certainly the need will be for a more thorough examination of such sites and better records. Huts, pits, cemeteries and other features must be sought, the ditches followed and their contents studied. The native pottery, its manufacture and distribution, needs studying in detail and surely the animal bones can no longer be ignored. Oystershells, quern-stones, glass, iron-objects and other small-finds all need be given full consideration. Publication is then an obligation. Much can still be learnt about these sites, their layout, economy and development. In particular the complete plan of a farm-hut has yet to be recovered in this area. Further details of the transition from native farmstead to Roman villa, much sought, could also be recovered. It is hoped that this short survey may provide at least a rough basis for future research.17 ACKNOWLEDGEMENTS In addition to published18 accounts a considerable amount of information on several of these sites was supplied by the archaeologists who undertook the work. In this respect many are to be thanked for their help and assistance. In particular Mr. A. J. Parsons gave information on no fewer than eight sites in the Cray valley ; Mr. P. J. Tester on Cobham (No. 2) ; Mr. W. S. Penn on Springhead (No. 7) ; Mr. R. E. Chaplin on Cliffe (No. 5) ; Mr. R. I. Little on Kings Wood (No. 30) and Mr. A. Detsicas on Greenhithe (No. 8). Mr. E. W. Salter has very kindly prepared the map of West Kent and Mr. P. J. Woods has read the proofs. To all these friends the writer is greatly indebted. ABBREVIATIONS Ant. Jnl. The Antiquaries Journal. Arch. Archaeologia. Arch. Cant. Archaeologia Cantiana. Arch. Jnl. Archaeological Journal. 17 The writer hopes to undertake a similar survey for the rest of Kent and parts of Surrey. Information on any new sites will be gratefully acknowledged. 18 Only ten of the sites have been fully published; nine appear in inadequate form and eleven remain substantially unpublished. 81 ROMANO-BRITISH WEST KENT A.D. 43-100 A.N.L. Archaeological News Letter. Cam. Hawkes and Hull, Camulodunum, 1947. Jnl. B.A.A. Journal of the British Archaeological Association. Maiden Castle R. E. M. Wheeler, Maiden Castle, Dorset. 0. Oswald, Index of Figure Types on Terra Sigillata, 1936-7. 0. and P. Oswald and Price, Terra Sigillata, 1920. P.P.S. Proceedings of the Prehistoric Society. Rich. Excavations of the Roman Fort at Richborough. S.A .C. Surrey Archaeological Collections. 82