Notes on the Probable Course of the Roman Road from Lympne to Dover

NOTES ON THE PROBABLE COURSE OF THE ROMAN ROAD FROM LYMPNE TO DOVER By IVAN D. MABGARY, F.S.A. THE existence of a Roman road connecting Lympne with Dover is attested by its actual appearance upon the diagrammatic map known as the Peutinger Table. No traces of the road had, however, been identified, and the growth of Folkestone and its outskirts has now put much of the probable route beyond direct investigation. Some notes were put forward by the late S. E. Winbolt in his book Roman Folkestone (Methuen, 1925) as a tentative approach to the subject, and it was with a view to testing these on the ground that the present investigation was made. There is general agreement that the existing road along the old cliffs at Lympne represents the Roman road. East of Shipway Cross it bends a good deal and is probably an old ridgeway track rather than an engineered road, but there seems no reason to disregard it as a part of the route on that account. We thus arrive at the crossing of the Brockhill Stream, just at the western entry to Hythe, and it seems clear that the trackway is directly continued by an old lane, now in part only a footpath, straight up the hill north-eastwards to Saltwood, making no doubt for the hills inland. Consideration of the eastward course of a Roman road from this point is very largely determined by the topography, which here shows marked features some of which would entirely preclude the making of a direct road. The Greensand belt that forms the hills near the coast is dissected by three streams running to the sea in almost gorge-like valleys at the eastern sides of Hythe, Seabrook and Sandgate respectively. We can probably discount the possibility of a road along the shore below the cliffs in those days, and any direct road on the hills above would have to keep well inland, east of Hythe, if it were to avoid needless climbing. However, the Hythe and Seabrook Streams rise well inland too, near Beachborough Park, and so cannot be altogether avoided, though it should be noted that the former flows in a particularly deep and steep-sided valley near Saltwood Castle, which would preclude any reasonable route for a road across it in that part of its course. Farther inland again, at the foot of the Chalk escarpment, the belt of G-ault clay makes a wet zone which is drained eastwards by the Pent Stream, running behind Cheriton and Morehall to Foord and then through another rapidly deepening valley to Folkestone Harbour. 87 PROBABLE COURSE OF ROMAN ROAD FROM LYMPNE TO DOVER Thus we are left with a fairly narrow strip of suitable ground for the Roman road : it should keep near the coast through Hythe till across the Hythe Stream (to avoid the Saltwood obstacles), it should then move inland to avoid the coastward valleys, but must make for a suitable crossing of the Pent Stream before gaining the Chalk, presumably towards Dover Hill. This route is almost exactly represented in its first stage by North Road, Hythe, which runs very directly along a terrace on the hillside above St. Leonard's Church. It is an old road, for it appears on Andrews and Dury's Map of Kent (1777), as does also the next section up Blackhouse Hill and by a very direct lane, through Scene Wood and the north side of Underhill House park, to a point near St. Martin's Church, Cheriton. The crossing of the Hythe Stream is relatively easy, and the climb out of its valley by Blackhouse Hill is somewhat eased by a curve of the valley at this point, as is also the case at the Seabrook Stream farther on. Substantial remains of large stones and metalling remain in Scene Wood and on the steep descent from Dibgate Plain to the Seabrook Stream which look ancient. The Seabrook valley forks below St. Martin's Church, and a straight footpath from the church runs east-north-east towards Cheriton, just turning the head of the eastern branch, a likely continuation of the route which would reach the railway some 460 yards east of Cheriton Junction. This point is the most southerly that would enable the awkwardlysteep eastern side of the Seabrook valley, and the Enbrook or Sandgate valley, to be passed on level ground. From it we can now only infer a direct course eastwards to the most probable crossing-place of the Pent Stream at Foord, near St. John's Church. High ground comes close to the stream on each side at that point, and the marshes which exist higher up on its course are absent; further, the valley has only just begun to deepen there on its way down to the harbour and the old centre of Folkestone. We may reasonably assume that the name Foord refers to an early crossing-place there. The most probable course, then, is a direct line from a point close to the railway near Cheriton Junction to St. John's Church Road, Foord, running near Cheriton Road through Morehall and then just north of Radnor Park. A continuation by the course of Archer Road and Green Lane to the lower elbow of Dover Hill seems very probable, though it is now only an inconspicuous " back lane " among houses. At the top of Dover Hill development has smothered everything, but the most likely course eastwards is along the ridge of the downs north of the present main road, by a row of hedgerows from Abbot's Land Farm, through Hougham Court, Hougham village, and along the ridge of West Down, Sunny Hill and Stepping Down, descending to the forking-point of the valleys where Elms Vale Road joins Folkestone 88 PROBABLE COURSE OP ROMAN ROAD PROM LYMPNE TO DOVER Road, pointing direct to the Roman centre of Dover. This route is very straight, lies high throughout, and on the open ground east of Hougham takes the form of a distinct agger some 15 feet wide, heavily metalled with large flints. Between Parsonage Farm and Hougham Church, where the present lane, Doctor's Lane, curves a little to the north, there is also a distinct ridge continuing the line through the field south of the lane. The lane on to Hougham Court and the hedgerow line near Abbot's Land Farm are, however, very featureless. So far we have only been considering the geographical suitability of the route. What evidences of Roman occupation support it ? They are quite considerable. A Roman villa was found at Harp Wood1 half a mile north-west of Hythe, in the valley of the Brockhill Stream, 540 yards north of the road. In Hythe Roman burials were found in a quarry at the corner of Hill Crest Road and Castle Road, just above North Road.2 In Cheriton an urn-field of some twenty Roman burials has been found quite recently on a housing estate,3 only 180 yards north-east of Cheriton Junction and some 200 yards north of our probable route. (The road over the site, between the railway and the Ashford Road, has in consequence been named " Roman Way," but should not be confused with our route which here lies just south of the railway.) Other Roman burials had previously been found near Radnor Park, west of Julian Road,4 as recorded on the Ordnance Map, also very near the route Roman villas have been recorded near Folkestone Junction Station5 and on Wear Bay.6 Roman burials have also been recorded at Hougham,7 though the site is not stated. The villas would probably not be close alongside the road, being approached by private branches, but the burials are very likely indications of its near neighbourhood. This route is practically that of Winbolt's " strongly supported alternative line " (Roman Folkestone, p. 159), under his route (ii). The other alternative to this route (north end of Coolinge Lane, Shorncliffe Camp, across Seabrook Stream at Horn Street village, and then over Hythe Golf Course) leads the road into several of the disadvantages that have been discussed above, and thus seems relatively unlikely. There is also his route (i), by the Creteway from Dover Hill to Cheriton Hill, along the crest of the downs, and then descending by 1 G. Payne, Coll. Cant. (1893), p. 199. 2 Roman Folkestone, p. 159. 3 Report on p. 21 of this volume. 4 Arch. Cant., XXXIV (1920), p. 156. 6 Arch. Cant., X (1876), p. 173, and XXIX (1911), p. 236. 0 Roman Folkestone. ' Arch. Cant., XXXIV (1920), p. 156. 89 PROBABLE COURSE OE ROMAN ROAD PROM LYMPNE TO DOVER Newington to Pedlinge and Lympne, mainly on the line of field footpaths. The Creteway itself has the undoubted distinction of being the final section of that ancient ridgeway which from the earliest times gave access along the line of the North Downs all the way from the English Channel to the centres of early Man in Wiltshire. The name " Creteway " (now regrettably modernized as Crete Road and Crete Road West in truly suburban style !) applies to the ridgeway from the Valiant Sailor Inn on Dover Hill to its junction, just west of the Canterbury Road, with Coombe Lane coming from the Alkham Valley ; and it is to be noted that the western continuation to Cheriton Hill, although so prominent upon the escarpment, is not the main ridgeway, which quite obviously crosses Coombe Lane, making straight northwestwards to Paddlesworth and Etchinghill where the Elham Valley is crossed. The reason for this is the formation of the downs at the mouth of the Elham Valley which preclude a direct continuance of the ridgeway there. It was easier to cross the high ground by Paddlesworth to Etchinghill, and then by Postling Wood to the edge of the escarpment again above Monks Horton. Here the wide coombe in which Stowting lies necessitated another descent and crossing as at the Elham Valley, and the route clearly did this direct to Brabourne Downs, thence following the summit to Wye and the west. As in so much of this route through Kent and Surrey, there are also traces of a Lower Road (so well known farther on as the Pilgrims' Way although prehistoric in origin) that hugs the foot of the escarpment as a terrace. This is a clear roadway all along from Wye through Brabourne, Stowting and Postling village, and then as a bridle way round the southern skirts of Tolsford Hill to Frogholt and Pean, where it leads up to the ridgeway on Cheriton Hill. But these routes are alternative courses of the ancient western ridgeway and do not really concern us as a direct Roman road to Lympne. The field paths from Saltwood may well be ancient tracks to the hills inland, but such a wide deviation from the direct route is on the whole very improbable for the Roman road. The course for this road described above is certainly the most suitable and likely, and a watch should be kept for further traces of it that may yet be found. 90