The Roman Road from Sutton Valance to Ashford Evidence for an Alternative Route to that proposed by Margary

THE ROMAN ROAD FROM SUTTON VALENCE TO ASHFORD: EVIDENCE FOR AN ALTERNATIVE ROUTE TO THAT PROPOSED BY MARGARY

neil aldridge

In his Roman Ways in the Weald, Ivan Margary describes the network of Roman roads in east and west Kent (see Map 1):

Route I provides direct connexion from Watling Street at Rochester to the Maidstone district, and this part of the route, more accurately aligned than the rest, may well have been laid out at quite an early date. When the iron-working industry in the Hastings area had become well established, at first most likely with seaward communications to the south, it must have become necessary to establish links with the settled areas to the north and north-east which were separated by a wide zone of Wealden forest. Thus the extension of Route I southward from Maidstone was the obvious first step, and this road, aimed directly at the Hastings area, may well have been constructed under more commercial, and less military, direction than the earlier roads which were built on accurately laid alignments. … It is clear, too, that this road lay near enough to the settled area of east Kent for eastward connexions to be needed for traffic to that district, and two such routes are traceable.

For the first 3½ miles [southward] from Maidstone Route I was laid in a south-easterly direction before the main southward alignment was resumed, and it seems clear that the engineer’s intention was to fork this road at this point, route I going south to Staplehurst and Hastings while route II [subsequently numbered 131] continued on an east-south-east course through Sutton Valence, to Kingsnorth and Lympne continuing thence along the coast to Dover. From a point near Kingsnorth route II still remains in use right on to Lympne and Hythe, but nearly all of it to the west of Kingsnorth has vanished, leaving but a few fragmentary traces. This route provided for traffic from the Maidstone area to the Kentish south coast and thus formed a southern parallel to both Watling Street and the Pilgrims’ Way.

Finally, a distinct need must have been felt for an eastward road from the iron-working area of the Weald, and this was provided by route III, which leaves route I at Hemsted, half a mile north of Benenden, and which proceeds past Tenterden, in an east-north-easterly direction, till it cuts route II near Kingsnorth.

The detailed maps provided by Margary in Roman Ways in the Weald show that his proposed route between Sutton Valence and Kingsnorth followed one general alignment but there were some minor curves and bends which took account of terrain. But he admits that ‘the evidence adduced for the lost portion of Route II west of Kingsnorth is admittedly slight, though it is what one would expect in such cases’.¹

The recent fieldwork

Between 1991-1996 the author and colleagues undertook a series of fieldwork investigations in the general area traversed by this section of Road 131. The evidence on the ground is generally elusive but the finding of a number of Roman sites in this stretch of country since Margary’s time provided impetus to carry out fresh fieldwork in order to accurately locate a road which would have provided communication between them.

Map 2 shows the route suggested by Margary together with the alternative route put forward in this paper. It should be noted how the new route is made up of a series of short, straight alignments in typical Roman fashion. Because the greater part of the farmland along the alignment of the road is down to pasture it proved more profitable firstly to examine watercourses and field ditches for any exposed sections of a road surface. The work was made more difficult because most of the road appears not to have been metalled with waste material from iron working in the area, unlike some of the roads sited nearer to the border with Sussex.² Instead, locally outcropping stone was utilised, most often that known as ‘crow-stone’. It is, however, quite likely that much of the route was unmetalled with perhaps only those sections impassable during wet weather being surfaced. If so, it would be virtually impossible to identify those unmetalled sections as being Roman, or indeed, to be a road of any period.

Although a large part of Margary’s suggested route was investigated to a greater or lesser degree, no sections through a road were found where they might have been expected to occur. Although the terrain is relatively flat there are several locations, such as Kingsnorth Wood, in the southern part of the parish of Ulcombe [TQ 843 477], where significant gradients do occur. If an ancient road had existed there features such as a worn hollow might have been expected. It is of course possible that later cultivation may have removed any surviving features although the wood is known to be of ancient origin.³ It does in fact probably represent the southernmost extent of the former royal forest of Kings Wood. The line of the Roman road may have been used to delimit the border of the royal forest in this area.

The westernmost section of the alternative route depicted in Map 3, passing through the parishes of Sutton Valence, East Sutton and Ulcombe, was the first to be located. Importantly, an early settlement site was discovered in this area during the initial phases of the research programme, only 400m south of the route. It is situated close to Jubilee Corner in the southern part of the parish of Ulcombe (No. 7 on Map 3). The site here is principally Romano-British and appears to have consisted of a number of buildings of timber construction sited in a shallow valley aligned north-east to south-west [TQ 841 471]. Three of the buildings have been excavated. Building 1 dates from the late first to mid third century ad and formed a rectangular structure lying north-south. An adjoining building on the eastern side appears to have been used as a workshop. Parts of these structures produced evidence to show that they had plank floors. Ironworking waste occurred in and around this building and several smelting areas were found to the south-east of the settled area. Building 2 was earlier, possibly late first century ad, and was sunken-floored. It was found to partially underlay Building 1 which replaced it. Building 3 was only partially excavated but resembled Building 1 in type and date.

The site seems to be rather more than a farmstead and may have been a rural native settlement with iron smelting taking place alongside agricultural activity. The local geology, although largely heavy clay, also includes areas where a shale-like stone with a significant iron content occurs.⁴ This must have been noticed from the earliest times as the area was occupied in the pre-Roman Iron Age with metal working taking place from as early as the fifth century bc. Pottery from the site suggests that the occupants were native peoples who were becoming increasingly Romanised. Dating evidence from coinage and pottery shows occupation up until the mid third century ad.

A second-century ad cremation cemetery has been identified at the assumed eastern edge of the settlement close to the present day Ulcombe to Headcorn road [TQ 8424 4703]. This may originally have been a branch from Road 131 heading to the south. It is therefore possible that the cemetery was sited alongside a contemporary road and would have been following the typical practice of the time.⁵

Air photographs show that the hillslope above the site at Jubilee Corner was occupied by a series of ditched enclosures dating from the Iron Age and Roman periods. Some of the aerial photographs show traces of a possible road connecting the settlement with Road 131. A linear double ditched feature appears to extend from the line of the road close to Jubilee Hall Barn along the crest of the ridge into the settlement.⁶ The dating of the road may thus be earlier than Margary thought; it is possible that it was built to link a series of rural settlement sites at a period when iron was required in some quantity, from the latter part of the first century ad. Certainly the fact that this site at Ulcombe appears to have been abandoned quite early on in the third century is an exact parallel of what has been found at sites in other parts of the Weald including the newly discovered settlement site at Westhawk Farm, Ashford.⁷

Description of the amended route

In his Roman Ways in the Weald, Margary shows the road approaching Sutton Valence from the north-west. A stone building of Roman date sited to the north of Chart Sutton church may have been a mansio lying close to the road [TQ 804 496]. Excavation of the building took place during 1949-50 consequent to evidence seen on air photographs (since mislaid).⁸ These suggested that the structure lay on the south side of a rectangular enclosure which covered many acres. This is however perhaps more reminiscent of a villa estate rather than a mansio. The excavation has not been published in full.

The road probably crossed the present day A274 slightly below Hilltop Garage, Sutton Valence. It then passes behind Heaven Cottage, the terrace way noted by Margary still exists in the back garden of the house here [TQ 811 493]. The line down through the village of Sutton Valence is difficult to trace owing to the concentration of buildings lying along the hillside. The walled cremation cemetery situated just above the road contained around 100 burials.⁹ This suggests that any Roman settlement here was fairly substantial and possibly lay where the present village lies. Evidence has not been found and indeed may have been largely destroyed by later development. The hillslope here is well watered with numerous springs. The Domesday survey lists vineyards at nearby Chart Sutton only slightly farther to the west on this same hillside. At the oddly named site of Pested Bars alongside the Roman road near Parkwood on the southern edge of Maidstone a further Roman walled cremation cemetery was located.¹⁰

The line south-east of Sutton Valence is uncertain. Margary felt that the road ran along the terraceway below the twelfth-century castle. The alternative alignment now proposed runs more to the south-east (Map 2). The land below the village here is mainly pasture or under intensive orchard cultivation which makes fieldwork rather difficult. The finding of a gold coin of Nero at Henikers, a house just south of Sutton Valence, in 1947 may be significant [TQ 8155 4820].¹¹ It is on the alignment projected north-west from the first confirmed section of the road, at Bells Farm in East Sutton (see below). At the end of South Lane in Sutton Valence is an area known for at least two centuries as ‘The Harbour’. It is a curious name and stories of something Roman in the vicinity still persist locally. It is worth noting that a line projected on from Henikers would pass through this area with the Roman road ascending the escarpment via South Lane.

The first exposed section of the alternative route to be positively identified was recorded in the parish of East Sutton [at TQ 825 479], 300m to the north-west of Bells Farm (No. 1 on Map 3). Fig. 1 shows the section as exposed in the side of the stream which flows south-west from Willow Wood in East Sutton Park. At this location the road metalling is made up of a 20-30cm layer of small stones, mainly crow-stone. The layer is 2.5m in width and there are traces of a possible side ditch on the south-west side. The northern bank of the stream is surmounted by a substantial hedge bank which effectively buries the Roman road under 1.4m of overburden.

There is further evidence (No. 2) for the road in the form of a prominent hump in the lane known as Friday Street [TQ 8285 4780] which lies beyond an arable field containing iron slag to the east of the section recorded in the stream.¹² Friday Street forms the western boundary of Park Wood where slight evidence for a hollow way survives where the line of the road emerges onto the road immediately west of Brissenden House. From here a line of hedgerows (No. 3) appears to mark the course down to the crossing of the stream north of Stickfast Lane (an appropriate name even today!). It is noticeable that some of the field boundaries in this area appear to mirror the line of the now lost road. There are very slight indications in the stream and clear sections of the Roman road which is running west-east still surviving in the banks on either side of the sunken road, Brick Kiln Lane, running north-south, which cuts across its course. The road surface here (No. 4) was found to consist mainly of the local crow-stone, together with ironstone [TQ 8395 4755].

In the stream south-east of Kingsnorth Wood, and adjacent to the north side of the minor road bridge close to the junction of Brick Kiln Lane and Stickfast Lane (No. 5), there is a prominent tumble of stones [TQ 8399 4753], which appears to be the remains of the road surface cut through by the stream. No sign of the road surface could be found in the meadow close to the stream to the east when a trial trench was put in here. It is likely that the stream has moved course during the last 2,000 years. In the field to the west, between the stream and Brick Kiln Lane, a clear agger-like feature remains visible in the cultivated field. After the crossing of the lane from Jubilee Corner the course lies through the field leading up to Jubilee Hall Barn, where a hollow is visible, and then passes directly through the site of Jubilee Hall itself.

The place where the road crossed the present Ulcombe Road (No. 6) is suggested by a ridge across the road, midway between White Acres and Roselands Farm in Ulcombe [TQ 8433 4725]. The hedgerows to the east of here also seem to mirror the course of the lost road. It probably passed to the south of Parsons Wood and then crossed the Lenham Road [TQ 8588 4623], 150m south of Thornden Cottages which are sited directly opposite the junction with Southerden Road.

Southerden (see Map 2) is first recorded in a pre-conquest Anglo-Saxon charter of the eighth-century and is amongst the earliest named sites in the area. The road carries the parish boundary between Headcorn and Boughton Malherbe and must therefore be of some antiquity. The road makes use of a saddleback of land here but, perhaps significantly, it doglegs somewhat towards the eastern end as though the route has deliberately diverged from the original straight alignment (a feature often encountered on roads of Roman origin).

The course of the road beyond Southerden probably lay south-east through Kingsden on Bedlam Lane, Egerton (this area suffered much disturbance during World War II), then north of Stace Wood into Frith Wood, referred to locally as Smarden Woods, in which parish they chiefly lie. The road emerges at Fright Corner north-west of Frith Farm [TQ 9040 4490], where the amended route coincides with Margary’s, the only place where in fact it does (Map 2). There are traces of stone road metalling in the modern road ditch and the Roman road possibly ran along the north side of the unusually wide present-day road for much of the way down to the Pinnock Bridge.

Shortly before the bridge the line of the Roman road crosses the present road and continues on the south side again changing alignment at the crossing of a stream. The line passes through the front garden of Rose Farm, Pluckley. The probable line of the road can be clearly seen as a worn hollow [TQ 91354430] in the pasture south-east of Rose Farm and south-west of the converted chapel. This area is mentioned by Witney as Pirifield,¹³ a rare survival of a place-name preserving the folk memory of a Roman iron working area.¹⁴

The route then crosses the Smarden Road and climbs on to a ridge north of Turner Farm, Pluckley. The line of the road is possibly marked by an existing hedgerow running south-east [TQ 9155 4415]. The road seems to run down the end of the ridge and crosses the road into a former football field in the south-west quadrant of the Chambers Green crossroads. There are signs of a slight agger running from the gateway into the football field [TQ 9208 4368], just to the south-west of a cottage before the Roman road crosses the present road north of Pluckley Station. To the east of the road the landscape is one of large flat almost featureless fields. However, a footpath running close to a watercourse north of the railway line is close to the alignment. After the road enters ‘The Forest’ there are no clear signs until it emerges south-east of Dynes Farm, Bethersden. There are slight traces of a hollow south of the road running east of Belmont Farm, Bethersden.

Clues to the course over the next section are few but a series of hedge-rows lying north-west to south-east appear to mirror the alignment in the vicinity of Etchden and Coldharbour Farms, and Daniels Water, on the edge of Great Chart parish. After Purchase Wood the road probably crossed the A28 at New Street Farm [TQ 9720 4084], south-west of Great Chart. There is a straight section of existing road here that appears to mark another change in alignment of the Roman route which is now directed towards Hamstreet, an area which would have had access to the sea in the Roman period.¹⁵ The road from New Street Farm to south-east of Bartlett Farm is noticeably wide and raised on an agger. Hamstreet would have been considerably nearer to the Weald for the shipment of iron products than Portus Lemanis.

After passing Bartlett Farm the road diverges from the present highway [TQ 9798 3948], and heads across the fields to Stubbcross Farm on the parish boundary of Shadoxhurst and Kingsnorth. Margary recorded pebble metalling and large stone foundations and clear signs of an agger in the vicinity of TQ 9825 3888.¹⁶ There is much iron slag in the area TQ 9815 3898 and it is possible that this indicates that the road heading north back towards Bartlett Farm was metalled with slag. Without doubt the road at Snailswood Farm on the Benenden-Westhawk route is largely made up of iron slag and the agger there is largely intact.

The remainder of the route to Hamstreet has not been examined in detail, however, a line via Coxland Wood and Bromley Green is quite feasible. At this point the line would join the present day road from Kingsnorth to Hamstreet. It is noticeable that Road 130 changes alignment shortly after passing the position of the postulated road junction east of Stubbcross Farm, Kingsnorth. It will be seen therefore that Margary’s Route 131 does not meet Road 130 from Benenden at Westhawk Farm, Kingsnorth at a crossroads. Further confirmation of this has come from the excavations carried out at Westhawk during 1998-1999. Both the excavation and the various geophysical surveys that preceded it clearly showed that the road from Lympne did not carry on beyond the northern boundary of the settlement in the direction of Sutton Valence.¹⁷ The settlement did not lie at a crossroads but at a road T-junction. Iron working was taking place within the occupied area and it has been suggested that iron from the Weald was being processed here on the edge of the forested zone. It is possible that the Westhawk site was just one of many such settlements existing throughout the Weald.

From Margary’s fieldwork and more recent research,¹⁸ it is probable that almost the entire length of the Roman road from Benenden to Westhawk (Road 130), a distance of some 15 miles (22km), was metalled with iron slag. It would not have been possible for this material to have been transported long distances from a handful of settlements within the Weald but rather that many iron producing sites were present throughout the area. Fieldwork is beginning to locate some of these but further research is required. The amount of iron smelting taking place in the Kentish Weald would have consumed significant areas of timber in the form of coppiced woodland to provide charcoal. There is some evidence for the shallow quarries where the iron stone was being extracted.

discussion and summary

Fieldwork has now provided sufficient evidence to support an alternative route for Margary’s Road 131. It is probable that the road did not traverse an unpopulated heavily forested zone. Indeed, the road may well have been intended to link a number of established rural native settlements. If the early settlement pattern that has been shown to have existed in Ulcombe, East Sutton and Headcorn is repeated throughout the Low Weald it would suggest that this general area was not an impenetrable forest in the late Iron Age and early Romano-British period.

Although the underlying geology does not at first appear particularly well suited to iron-working it is evident that smelting was indeed taking place from the Middle Iron Age onwards. There is at present no evidence for the presence of the Classis Britannica operating in this area unlike the corridor alongside the road from Cranbrook south to the assumed port at Bodiam and beyond to Beauport Park (both in Sussex). That road, No. 13, is somewhat easier to trace in that area owing to the large-scale use of waste material from iron working in its construction. Recent excavation undertaken by the writer along the route of the Benenden -St Michaels section of Road 130 has shown that in places the road survives intact as a solid iron slagged surface with metalling up to 23cm thick. The road has been found to vary in width from around 2.3-5.5m with some evidence for side ditches.¹⁹

There is a definite change in the alignment of the newly discovered route of Road 131 at New Street Farm, Great Chart, where the heading is altered towards Hamstreet rather than the Westhawk settlement site (see Map 2) at Kingsnorth. Further work is required in order to begin to understand the nature of the agricultural and industrial land usage in the Weald in the Roman period.

It is gratifying that much of the new development taking place in and around Ashford is being covered by archaeological surveys and investigations under PPG 16. This work when combined with other archaeological fieldwork taking place in the area will, hopefully, enable us to discern the broader picture more clearly.

acknowledgements

The writer is indebted to the following for their assistance during the fieldwork: Mr W. Coomber for practical help in seeing the terrain through the eyes of a Roman road surveyor; Mr R. Sharp for his knowledge of Kingsnorth Wood, Ulcombe; and Paul Booth (Oxford Archaeology) for his description of some of the preliminary findings of the Roman settlement at Westhawk. Thanks are also due to the landowners along the route of the road for all their help and interest.

endnotes

¹ Margary’s pioneering fieldwork led to the publication of Roman Ways in the Weald in 1965. However, the likely former existence of the Sutton Valence-Ashford road was postulated by a number of earlier writers including Donald Maxwell in his book Unknown Kent, published in 1921. He appears to have been the first to put forward the idea (p. 143).

² A. Miles, ‘The Rochester-Hastings Roman Road near Sandhurst, Archaeologia Cantiana, lxxix (1964), 207-8; see also N. Aldridge, ‘Roman Road at Sandhurst/Bodiam’, Romney Marsh Irregular, No. 18, Oct 2001, 16-18.

³ K.P. Witney, 1976, The Jutish Forest, pp. 60-61, 67.

⁴ J. Hodgkinson, former Chairman of the Wealden Iron Research Group, pointed out that some iron-bearing strata outcrops in the base of the Greensand escarpment in Surrey and probably continues through into Kent. This has been noted in watercourses in the part of the Low Weald investigated here and may account for the presence of some of the iron bloomery sites – for example, the extensive Kingsnoad site and evidence for iron smelting found at nearby Jubilee Corner Romano-British settlement, both in Ulcombe. See also WIRG Bulletin No. 20, 2000, p. 4.

5 If the line of the road to Headcorn is extended towards Frittenden it would pass close to the Roman site discovered recently by the writer at New House Farm, Headcorn [TQ 8315 4320], during the course of the fieldwork. Further conclusive evidence for a contemporary Roman road serving the New House Farm settlement is at present lacking. The site there appears to be a Romano-British farmstead with evidence for earlier Iron-Age occupation.

⁶ NMR. Air Photo, 106/UK/1387 FrS149, 11 April 1946.

⁷ Pers. comm., Paul Booth, Oxford Archaeology.

⁸ OS Arch. Records; Roman Building, Chart Sutton.

⁹ Archaeologia Cantiana, x (1876), 166; xv (1883), 88.

¹⁰ Archaeologia Cantiana, xv (1883), 81.

¹¹ OS Arch. Records, Roman Coin from Henikers, Sutton Valence.

¹² If the line of Friday Street were continued south of Bells Farm along the footpaths through Thornden it would provide a link to the Roman site that exists beneath the site of the Medieval priory at Moatenden. This is another of several possible branch roads from Road 131.

¹³ The Jutish Forest, pp. 20-21, 187.

¹⁴ A branch from this point south-west along the wide straight road to Maltmans Hill/Romden is quite feasible as there was a significant iron working site in that area which would have required access to a major road – see N. Aldridge, ‘Three Notes on Ironworking sites in Kent’, WIRG Bulletin, second series, No. 16, 1996.

¹⁵ The derivation of Hamstreet is the ‘watermeadow close to the Roman road’, Greensand Way in Kent, KCC, Maidstone, 1992, 38.

¹⁶ Roman Ways in the Weald, p. 249.

¹⁷ Site plans of geophysical surveys, Westhawk Farm, Oxford Archaeology.

¹⁸ N. Aldridge, ‘Roman Roads around Benenden’, KAS Fieldwork Project 2000.

¹⁹ Ibid.

Map 1 The Roman Road network east of the Medway.

Map 2 Roman Road 131 as suggested by Margary (dot and dash line) and the proposed alternative route (dashed line). The Benenden-Canterbury road (dotted line) passes through the Westhawk settlement (black circle); other Roman sites are shown by an asterisk.

Map 3 Details of the alternative route of Roman Road 131 in the parishes of Sutton Valence, East Sutton and Ulcombe.

1 – section recorded north-west of Bells Farm; 2 – crossing Friday Street; 3 – hedgerow on the line east of Brissenden; 4 – section in hedgebank at Brick Kiln Lane; 5 – metalling visible in stream; 6 – crossing Ulcombe Road; 7 – centre of Romano-British settlement south-west of Jubilee Corner.

Fig. 1 Section of Roman Road 131 exposed in north bank of stream at Bells Farm, East Sutton. (A-B indicates probable original width of road.)