An Early Boundary Probably Anglo-Saxon Associated with Roman Sites in Benenden Pollard

301 AN EARLY BOUNDARY, PROBABLY ANGLO-SAXON, ASSOCIATED WITH ROMAN SITES IN BENENDEN ernest pollard and neil aldridge Pollard and Strouts (2005) showed that dens in the north of Benenden parish belonged to manors in Wye lathe and those in the south of the parish to manors in Lympne lathe.1 They suggested that the boundary between these early lathes was preserved in hundred boundaries, which appear on nineteenth-century maps. The suggested lathe boundary is likely to be pre-Domesday, on the basis that the dens which it divides are Saxon creations and that Benenden Manor, which would have shared this boundary for part of its length, was in existence at Domesday. Everitt (1986) observed that the early Saxon estates, including forerunners of the lathes, were ‘almost everywhere based on Romano-British settlements like Maidstone, Milton, Reculver, Faversham and Otford’. The aim of this paper is to bring together evidence suggesting a different type of continuity in the Weald; an association between an early boundary in Benenden and Roman sites. Roman sites in the Benenden area Two Roman roads pass through the parish (Fig. 1), meeting at a road junction which cannot be located precisely because it is overlain by the buildings of Benenden School (NGR 801 338). The roads were first shown on Ordnance Survey maps c.1870 and their lines slightly revised by Margary (1948) and by Aldridge (2001, and this paper). The northsouth road runs from Rochester to iron-working areas near Hastings: the east-west road runs from the Benenden junction eastwards to Ashford, passing through a contemporary Roman settlement at West Hawk and on to Canterbury. A stone-paved ford at Stream Farm on the north-south road at Iden Green (Fig. 1) has been identified as Roman by Crawford, Margary (1948) and Lebon (1984).2 The identification as Roman is based largely on the location of the ford on the Roman road; a later date for the stone-paving is possible, but it was in any case a Roman crossing place. An ironworking site (Fig. 1, NGR 801 352) was discovered at Little ERNEST POLLARD AND NEIL ALDRIDGE 302 Fig. 1 Roman sites and hundred boundaries in the Benenden study area. The hundreds are Selbrittenden (extreme south), Cranbrook (west), Rolvenden (largest area, including Benenden village), Barclay (north-east). Numbered sites are as referred to in the text: all of the numbered sites are marked by hundred boundary stones and all lie on the suggested lathe boundary. The OS map (copyright Ordnance Survey) is shown faintly in the background to help with the location of the Roman sites and boundaries. AN EARLY BOUNDARY, PROBABLY ANGLO-SAXON, IN BENENDEN 303 Farningham Farm, Cranbrook, in 1956 and investigated by Lebon from 1956-1959.3 The site was re-examined and reassessed by Aldridge (2001). Apart from the furnace and associated ironworking area, there is little indication of domestic or other substantial buildings there.4 A settlement in the grounds of Benenden School (Fig. 1, NGR 8018 3352) was discovered in the 1950s by girls of the school Classical Society, led by a teacher, Miss C. Wilson. The first excavation was in 1954, with further work between then and 1960. An extensive collection of marked pottery shards and other material from the dig has survived, including a small amount of the concrete opus signinum and iron slag. The excavations were reported in the school magazine for 1955-56, but not in the archaeological literature. In 2006, the authors together with Mr Giles Lock, who as a child had visited the open excavation, relocated the site. Roman pottery was found on the ground surface, disturbed by rabbit activity, and the location of a trench confirmed by a contemporary aerial photograph.5 No site plan has been found, but the school magazine describes extensive metal-working areas, with evidence of burning and iron slag. The finds suggest that there was a small settlement here, raising the possibility of a more extensive Roman site in the general area. Hundred boundaries and a suggested lathe boundary The origins of the Kentish lathes and their subdivisions, the hundreds, are uncertain. Everitt (1986) suggested that the lathes evolved from the ‘estates’ of the earliest stages of Anglo Saxon settlement. Brooke (1989) considered that the lathes of east Kent, including Wye and Lympne, may go back to the earliest days of the Kingdom of Kent, but in what form is not known. The hundreds are thought to have been imposed later, within an existing structure of lathes (Jolliffe, 1933; Witney, 1976), but this is open to question and they may have earlier origins (Thorn, 1992). At Domesday, or a little later,6 parts of four hundreds lay in Benenden parish; Rolvenden and Selbrittenden hundreds (Lympne lathe); Cranbrook and Barclay hundreds (Wye lathe). The boundary between Wye and Lympne lathes, separating the dens of the two lathes, is thought to have been preserved in hundred boundaries (Fig. 1). This boundary is referred to here as the ‘lathe’ boundary, to emphasise the provisional nature of the identification. The hundred boundaries are shown on the first large scale Ordnance Survey maps of the late nineteenth century. Pollard and Strouts (2005) suggest that in the Benenden area these boundaries are in essence the original boundaries, although some changes over the centuries are to be expected. Sections of boundary which could not be mapped in detail by the nineteenth-century surveyors were classed as indeterminate, and are shown on the maps as straight lines between known points. A boundary ERNEST POLLARD AND NEIL ALDRIDGE 304 could be indeterminate in the sense that it was through a ‘wilderness’ area without features, but here, in an area of old settlement, it is more likely that the exact line of the boundary was once known but has been lost. There are six hundred-boundary stones in Benenden parish, of unknown age but probably eighteenth- or nineteenth-century. All these stones are around Rolvenden hundred. Coincidence of the ‘lathe’ boundary with Roman sites in Benenden The stone-paved ford, on the Roman road at Iden Green (1) (key sites are numbered on Fig. 1) is at the meeting point of three hundreds, Rolvenden, Cranbrook and Selbrittenden, marked by a three-sided stone. The ‘lathe’ boundary follows the stream to the south-west of the ford. At the ford, the boundary turns northward and the (undefined) line here is almost parallel to, and close to, the line of the Roman road. A later sunken road, now disused, is on, or very close to, the line of the Roman road here. After crossing the modern Benenden to Cranbrook road, the boundary leaves the Roman road and runs north-eastwards along a medieval road, shown on a map of 1599 but now evident only as a worn holloway crossing parkland.7 At site (2), within the school grounds, a stone marks where the boundary turns sharply east (at an angle 105o). The turning point and stone are about 7m beyond (north-west of) a water-filled pit, now known as Paddy’s Pool, shown on the 1599 map, and 39m south of the excavation site of the newly relocated Roman settlement. Eastwards, the boundary meets another old road at a bridge (Hemsted Bridge) shown on a map of 1777:8 this road and bridge were replaced by the modern road in the early nineteenth century.9 The ‘lathe’ boundary continues north-east, along a disused road, streams, field boundaries and another disused road to meet the Roman road at Goddard’s Green (3). This site is another meeting point of three hundreds, Cranbrook, Rolvenden and Barclay, again marked by a threesided stone. From Goddard’s Green, the ‘lathe’ boundary follows a disused road south-east for a short distance before turning north and shortly east along a long undefined line to a boundary stone on Pump lane (4). As is the case between sites (1) and (2), this stretch is close to, and more or less parallel with, the Roman road, the exact line of which is here uncertain.10 Goddards Green is on a ridge which carries a road, assumed to be a drove road, running north-east along the ridge to Biddenden and then on to manors near to Ashford (Pollard and Strouts, 2005). The Roman road and the hundred boundary turn away from the ridge to the south-east, and continue on its southern slopes until the boundary turns sharply south along Stepneyford Lane. Margary suggests that there may have been a AN EARLY BOUNDARY, PROBABLY ANGLO-SAXON, IN BENENDEN 305 branch Roman road near here ‘leading to the head of what would have been an estuary running inland from the Isle of Oxney’; he suggests that the departure of the main Roman road southwards from the ridge, which is the more obvious line (and is the line of the drove road), is to meet this branch road to the estuary. There is a further site, west of Hemsted Forest and the Golf Club (NGR 802 357), at which an acute angle in a hundred boundary (not the ‘lathe’ boundary) is close to a Roman road (30m) and fairly close (350m) to the Little Farningham ironworking site. We believe that the association of this site with the Roman features is probably just chance, although the road here was heavily metalled with iron slag and may have remained a prominent feature when the boundary was formed. Discussion We draw attention to the coincidence of a suggested early lathe boundary and Roman sites in Benenden. Two sites where three hundreds meet lie on Roman roads: two undefined stretches of the boundary are more or less parallel with and close to lengths of Roman roads: a recently relocated Roman site is close to a point where the boundary turns sharply. A boundary needs to be easily recognised, and sites where boundaries meet need to be especially conspicuous. The ford (or the stream crossing, if the ford is not an early feature) is such a conspicuous site. Goddard’s Green, the other site where three hundreds meet, has no similarly obvious feature, but lies on a ridge carrying the Roman and later road; five old roads, two long disused, radiate from the green. Examples of parish and manor boundaries following Roman roads are common and indeed Margary used parish boundaries as clues for tracing roads (Margary, 1948; p. 21). In at least one case in Kent, a parish boundary runs parallel and close to a Roman road (Margary, p. 217) in a way similar to that described here.11 In such cases, road and boundary may originally have coincided. Between sites 3 and 4 (Fig. 1), the line of the road is uncertain and it remains possible that road and boundary coincide here. The location of the newly rediscovered Roman settlement, close to a sharp change in direction of the boundary, seems significant. Some distinctive feature must have marked this turning point; it is possible that the boundary was formed when a Roman building (or buildings) was conspicuous, perhaps in ruins. If the old pit had been the key feature, one would expect the change in direction of the boundary to be at the pit rather than a short, but distinct, distance from it. East of Goddards Green, the departure of both Roman road and boundary from the ridge, the more natural line for the road and the line of the later road, also seems significant. ERNEST POLLARD AND NEIL ALDRIDGE 306 There are a variety of ways in which Roman activity may have left traces in the landscape to be used as markers many years, even centuries later. The association with Roman sites, we believe to have been demonstrated here, does not therefore prove an early Saxon date for the boundary although it makes an early date more likely and raises questions about the nature of the relationship with Roman sites. acknowledgements We thank Mark Gardiner for comments on a draft of the paper and Giles Lock for his help in locating the site of the Roman excavation at Benenden School. We are also grateful to Pat Thoburn and to Benenden School for permission to conduct studies on their land, to Hazel Strouts and Wayne Coomber for help with fieldwork and to Judith Johnson of Benenden School, who provided us with the articles on the Roman excavations in the early school magazines. bibliography Aldridge, N., 2001, ‘Little Farningham Farm, Cranbrook, revisited’, Archaeologia Cantiana, 121, 135-156. Brooks, N.E., 1989, ‘The creation and early structure of the Kingdom of Kent’, in Bassett, S. (ed.) The Origins of Anglo-Saxon Kingdoms, Leicester. Everitt, A., 1986, Continuity and colonisation: the evolution of Kentish settlement, Leicester University Press. Jolliffe, J.E.A., 1933, ‘The origins of the hundreds in Kent’, in Edwards, J.G., Galbraith, V.H. and Jacobs, E.F. (eds), Historical essays in honour of James Tait, Manchester. Lawson, T. and Killingray, D. (eds), 2004, An Historical Atlas of Kent, Phillimore. Lebon, M.C., 1984, ‘The Roman Ford at Iden Green, Benenden’, Archaeologia Cantiana, 101, 69-82. Margary, I.D., 1948, Roman Ways in the Weald, Phoenix House. Pollard, E. and Strouts, H., 2005, ‘The dens of Benenden and a possible early lathe boundary’, Archaeologia Cantiana, 125, 43-65. Thorn, F.R., 1992, ‘Hundreds and Wapentakes’, in Williams, A. and Martin, G.H. (eds), The Kent Domesday, Alecto Historical Editions, London. Witney, K.P., 1976, The Jutish Forest, a Study of the Weald of Kent from 450 to 1380 A.D., Athlone Press, London. endnotes 1 The lathe structure was revised in the thirteenth century (Lawson and Killingray, 2004), after which the whole parish lay within the lathe of Scray. 2 Many stones of the ford are shown in situ on a photograph c.1930 (Margary, 1948, p 225; Lebon, 1984) but they are now mostly dispersed or buried. The photograph was first published in Kent, Victoria County History, iii (1932). AN EARLY BOUNDARY, PROBABLY ANGLO-SAXON, IN BENENDEN 307 3 Excavations described in Archaeologia Cantiana, 71 (1957), 224; 72 (1958), lx-lxii; 73 (1959), xlvi. 4 There may be a connection with the Classis Britannica or Roman Fleet, as CLBRstamped tiles were found here in the earlier excavations. 5 National Monuments Record, V.58/RAF/2423, Frame 0066, date 23/04/1958, Benenden. 6 The hundreds in Wye lathe are not mentioned in the Domesday book, but may have been in existence then; the two in Lympne lathe have just one (Selbrittenden) and two (Rolvenden) mentions respectively and could easily have escaped attention. 7 The medieval road was mistaken for the Roman road by Margary (1948); the Roman road lies further to the west (Fig. 1). The 1599 map is in the Suffolk Record Office, Ipwich, HA43 T501/242 8 CKS P20/27 1. 9 This road is shown on the first, 1801, Ordnance Survey map (Mudge’s map), but had been replaced by the modern road by the time of the tithe map in 1836. 10 Margary’s line is based on his assumption that a strong agger in Uppergate wood marked the line of the road (Margary, 1948; p. 243). Recent investigation (Aldridge, unpubl.) suggests that this may be a woodland bank; scant evidence for a road was found in a dig at NGR 8284 3472. 11 The parish boundary between Chart Sutton and Linton, west of the Rochester to Benenden Roman road, is parallel to the road but 70m west of it.