Notes on Roman Roads in East Kent

NOTES ON ROMAN ROADS IN EAST KENT By IVAN D. MARGARY, F.S.A. THE Roman roads of East Kent are generaUy so obvious and weU known that no further description of them seems necessary. There are, however, a few points at which the line is doubtful or where topographical problems admit of some discussion, and it is in consideration of these that the foUowing notes are offered. They are based upon field observation made during a visit of some days to the district in June, 1947. STONE STREET—LYMPNE TO CANTERBURY The very striking directness of this road makes it clear that its purpose was to link Canterbury with the Roman port at Lympne. This was probably situated below the old cliffs, near the hamlet of West Hythe, to which access is given by a convenient combe in the cliffs at that point from Shipway Cross above. The Saxon Shore fort at StutfaU Castle was, of course, a late Roman construction, much later than Stone Street, and was probably connected to West Hythe by a spur road below the cliffs, for access to it from Lympne, directly above, would have been awkward. It is to Shipway Cross and the head of the combe that the main ahgnment of Stone Street is exactly directed. Although it has now disappeared between the Cross and New Inn Green, there are distinct traces of its scattered stone metalling in the field to the south of the Green, while portions of hedgerows and a footpath mark some parts of its course there. Its existence here is thus reasonably supported, and it seems desirable that the title " Stone Street—Roman Road " appUed by the Ordnance Survey (Kent Sheet LXXIV SW) to the present road from New Inn Green to Lympne vfilage should be canceUed, for this is, presumably, a later diversion to the vfilage from which point the direct access down the chffs to the late Roman fort would have been difficult.. The road runs very straight from New Inn Green through Stanford, aUgned upon a shoulder of the Downs at Monks Horton where a convenient combe (as at Shipway) aUows of an easy ascent to the top of the escarpment. It is noteworthy that the deviation tends to f oUow a series of short straights rather than a curve. At the summit the road i oUows the course of the ridgeway for about half-a-mile before resuming its northward direction ; this is to enable it to avoid an eastward vaUey, and the main ahgnment is not resumed till it has passed this, haU-a-mile 126 NOTES ON ROMAN ROADS IN EAST KENT farther on. A tumulus stands beside the road, upon the west, near Hemsted.1 From this point the road is remarkably straight, even for a Roman road, almost to Canterbury, and the modern road appears to foUow its line very accurately. It is especiaUy weU preserved, a clear raised agger, near SteUing Minnis and Upper Hardres,and it makes clever use of the ridge east of Petham, indeed it seems very likely that it was the position of this convenient ridge that determmed the course of this main ahgnment. Over Chartham Downs the road is now deeply sunken. Upon reaching the descent into Canterbury it appears to f oUow the course of a footpath, in which traces of the metalling are plainly visible, lying somewhat to the east of the present road, Iffins Lane. In any case, it joined the road from Chartham and Ashford, Wincheap Street and Cockering Road, some httle way outside the South Gate of Canterbury, showing that that road, an ancient trackway traceable continuously from Wye Downs, must have existed earher. CANTERBURY—UP STREET This road is typicaUy Roman in appearance, being very straight and raised almost throughout. The alignments on each side of Sturry appear to have been planned together. It is notable that the Reculver branch leaves this road at a point sufficiently far beyond Sturry to show that the Up Street road was aheady in existence. The slight bend to the southward near Up Street avoids low ground to the north. It is very hkely, of course, that there was a ferry across to Thanet from Up Street before the Sarre WaU existed. STURRY—RECULVER This road appears to have been laid upon a major alignment from the fork at Sturry to Ford, with short ahgnments on to Hfilborough and Reculver. The bends at Hoades Court and BuckweU probably represent lost portions of the road, indeed the west edge of a wood and then a hedgerow on to Rushbourne appear to represent the course at BuckweU upon the ahgnment. North of Ford the road was apparently placed a httle to the west of the main alignment in order to avoid a rather awkward hoUow there, and also a vaUey to the east at Hawthorn Corner. It is interesting to see how, on approaching Reculver, the road keeps west before turning direct to the fort, so as to avoid the low-lying ground, now marshland but then an estuary. This is the case with both the present Reculver Lane and with the older road from Bishopstone, now lost in the sea, which approached the fort directly from the west.2 It is possible that both were in Roman use. 1 It has the flattened top characteristic of Roman burial mounds. See R. F. Jessup, Antiquity, XII (1938), 103. 2 R. F. Jessup. Antiquity, Vol. X (1936), 189. 127 NOTES ON ROMAN ROADS IN EAST KENT VIINGHAM AREA ROM AH CtMtrentes DEERSQN SCALE OF Aftt.es WART'S DFUSNT (CKHAM GROUND \ rneeouRHE BMMUNG WINGHAM UWtF/t WlH&HAM HEAVY DOWNS DIRECT AIIGNMEUT FROM CAHTERauttV FIG. 1. ROADS IN WINGHAM AREA. ROMAN CAUSEWAY PROBABLE: COURSE _ OF ROMAN ROAD ^ES FM, «_ RtctiBonouGH !*..*.-V*.; ASH - VJOODNESBORO 'ISLAND REA COOPER'S STREET MUCH FLINT tiSTALUt/G fturupfc '?;f0ftT~ LOWER WADMCtOti S TIPPLEO F«. , . ross/ffi HAVEN SHATTERIIUG /ami ««*» ROMAN HOAO EPHHAIHV HAHSHBOROUCH WiZh»A» *">*« WOODtlES BOROUGH t^Dmccr AUCHMENT FROM DOVER FIG. 2. ROADS IN ASH-WOODNESBOROUGH AREA. (Scale as in Fig. 1.) 128 NOTES ON ROMAN ROADS IN EAST KENT DOVER—WOODNESBOROUGH After ascending the Downs out of Dover by a convenient combe the road foUows a single ahgnment very closely, except on hiUs. It approaches Pineham, near Whitfield, as a narrow field lane of doublelynchet type, about 12 feet wide. North of the Whitfield—West Langdon Lane its exact course for haU-a-mUe appears to be lost but then, to the west of West Langdon Church, it is very plain as a derelict hoUow way across the fields, 12 feet wide and up to 6 feet deep. Next it remains in use as a narrow by-road past Ashley and East Studdal. Shght bends on the top of the hiU by Great Studdal Farm may be original. North of the Mongeham lane it is derelict through Betteshanger Park but is remarkably weU preserved as a visible metalled agger, usuaUy along the west side of the present track, past Telegraph Farm to the Park. At a point 200 yards north of the farm a war-time trench disclosed a perfect section of the metalhng, just a thin layer of flints, 2 to 3 inches thick, upon an agger 15 feet wide. Where it leaves the Park, at the north, the agger is again very clear in a plantation, lying to the east of the present track, about 25 feet wide. Thereafter it is still in use as a road through Eastry to Woodnesborough. CANTERBURY—ASH, WITH CONNECTIONS TO WOODNESBOROUGH, SANDWICH AND RICHBOROUGH The main ahgnment of this road, almost due east from Canterbury, save for a curious divergence near Littlebourne and Wingham, is a striking feature upon the map, and must inevitably be considered as Roman, for it is obvious that such a route giving direct access from the Sandwich—Richborough area to Canterbury would have been much needed. Although some scepticism has been shown regarding the bends which this route appears to have followed, it can, I think, be shown that they were dictated by the obstacles encountered and that, in fact, the route f ohows the normal Roman custom of making the best use of the ground. The first 2 | miles from Canterbury was clearly laid out upon the main ahgnment, sighted on high ground, to the west of Littlebourne, and the three miles from Wingham to Ash foUows approximately the same fine, modified shghtly to keep clear of the eastern branch of the Wingham River which runs paraUel to it in a somewhat steep-sided vaUey. ActuaUy, the line through Shatterling, where a shght bend occurs, is cleverly sited to avoid the bends in this vaUey which practicaUy dictate the position of the road. The position at Wingham (see Fig. 1) is comphcated by two factors which affected the route taken; (a) an aheady existing pre-Roman ridgeway running N.N.E. from Barham, over Cooting Downs and Neavy Downs, through Wingham to Ehnstone, where it may have NOTES ON ROMAN ROADS IN EAST KENT ended at some smaU haven, and (6) the low ground in the vaUey of the Wingham River which widens out west of the vfilage at the junction of two southerly tributaries, the Port RiU and the stream from Wingham WeU. This low ground would certainly have made a crossing upon the general ahgnment a long and awkward one. It is fairly obvious that the existing easier crossing was utilized. At Littlebourne, too, where the Little Stour has to be crossed, the vaUey appears to offer an easier crossing at the site of the vfilage than farther north, near Ickham, where the general ahgnment would he. It seems quite evident, therefore, that a modified fine was foUowed here through Littlebourne and Branding as the road now runs. It does, however, seem very doubtful whether the diagonal connection from the Volunteer Inn at Branding to Wingham was, in fact, part of the Roman road, for the original road appears to go on fairly direct through Wingham WeU to Neavy Downs where it connects with the earher ridgeway. The diagonal road between the Volunteer and Wingham Green looks suspiciously like a coaching short-cut, and it involves, too, an extra river crossing at Wingham Bridge (on the Wingham WeU stream west of the village) which the other route avoids altogether. Thus the course of the road through Wingham can be readily explained by the topography, and only one other smaU difficulty here merits attention, the haU-mile deviation round the southern edge of Pine Wood, west of Littlebourne. There is no trace whatever of the course of the road within the wood, but, on the other hand, the present road is highly embanked in its western part and is accompanied almost throughout by such wide and deep ditches, meriting the term fosse, that it seems fairly certain this must be the original road. BRANCH ROADS FROM ASH (See Fig. 2) Another straight road, probably a pre-Roman ridgeway, joining this route at the west end of Ash village deserves to be mentioned. It runs shghtly north of west from Ash, foUowing the top of the ridge above Shatterling and carrying for some distance the Preston—Staple— Wingham parish boundaries; it crosses the Wingham-Elmstone ridgeway at Four Turnings, and is clearly traceable along the north side of Wenderton Heath, near Deerson Farm, right down to tidal water on the Little Stour, and near the Roman cemeteries which Ue along the high ground overlooking the river there, southwest of Preston Church, certain indications of a riverside settlement nearby. We come now to the problems associated with the Roman coastal connections near Richborough and Sandwich. It has been generaUy accepted that the Roman road from Dover to Woodnesborough branched directly thence to Ash and to Sandwich, and perhaps also continued northwards to Each End and Richborough. The first mile of the road 130 NOTES ON ROMAN ROADS IN EAST KENT eastwards from Ash to Each End seems a natural continuation of the main westward ahgnment, though we must remember that this could not then have run on to Sandwich for the marshes were at that time part of the estuary. The Ash-Woodnesborough-Sandwich routes are, I think, reasonably certain, for their alignments can stiU be traced by existing roads and paths. It does seem, however, that the term " Roman road " is wrongly placed upon the present Ash-Woodnesborough road through New Street (suspicious name) and Marshborough, whereas the original and straighter route appears to he just to the west of this, through Mount Ephraim, Coombe and Beacon HiU, by an obviously old road, in part dereUct, that foUows the top of the httle escarpment there. Similarly, on the Woodnesborough-Sandwich road, which skfifuUy avoids the marsh level until reaching its terminus, it is evident that the footpath near The Grange and, again, behind the Poulders Gardens Estate forms part of the original aUgnment. Both these ahgnments meet the Dover road just by Woodnesborough Church. Of the connection with Richborough, it is definitely known that a road left the West Gate of the fort, and this has been traced across the island to a causeway, stiU visible, which crossed to the mainland at the narrowest part of the estuary, near Fleet Farm.1 Traces of the metalling recognized locaUy as part of the Roman road, are also visible at Cooper's Street, just along the north-west side of the present droveroad where it crosses the marsh there. These traces can be foUowed back near the track leading thence north-eastwards to some cottages beside a smaU disused chapel, and seem to indicate that after gaining the mainland near Fleet Farm the road turned south-west through Cooper's Street towards Ash. If this course were continued, it would be represented now by the footpath, quite a featureless track, past Lower Waddington Farm and so to Ash. It is a fact, as Winbolt pointed out, that, just before Ash is reached, a large raised strip appears in a grass field adjoining this path on the south-east. The strip is not metaUed but it looks very like part of an earthen agger and it is exactly where it should be. It lends considerable support, taken with the undoubted metalling at Cooper's Street, to a direct Ash-Richborough connection which should thus be considered probable. Of a direct Woodnesborough-Each End-Cooper's Street route there seems to be little sign, save perhaps for some scattered stone to the south of Each End and the existence of lanes thence through East Street and Brook Street. If the Ash-Richborough connection existed, there would be httle to be gained by going north to Each End and East Street, only to have to foUow a long deviation round the estuary to 1 C. Knox, Arch. Oant., LTV (1941), 35 ; S. E. Winbolt, Roman Folkestone (1925), 149 ff. 131 NOTES ON ROMAN ROADS IN EAST KENT reach Cooper's Street; the Woodnesborough-Ash-Cooper's Street route would serve as weU. On the other hand, a direct ferry-crossing to Richborough Island from a point near Little East Street Farm, might have had advantages and it is noteworthy that an inlet stiU exists at this place, ending in a rectangular basin stiU holding water, situated just at the end of the ahgnment of the Dover road, which would have made such a use possible. A SUGGESTED RIOHBOROUGH-UP STREET ROUTE It seems necessary to add some comment here upon the alternative route from Richborough to Canterbury, by way of Westmarsh, Grove Ferry and Up Street, suggested by 0. Knox.1 This line is based upon the noteworthy coincidence of alignment of a short piece of road at Grove Ferry, a straight haU-mile length of watercourse at Westmarsh, a quarter-mile length of lane north-west of Fleet Farm, and finaUy the admitted Roman causeway to Richborough Island. The author suggests that the tactical need for rapid communication between Richborough and Reculver would have caUed for such a route. While the deshabihty of a direct link may be admitted, although (as in recent times) it may weU have been the practice to hold the frontal defences lightly and to maintain a central reserve for such emergencies—obviously at Canterbury in this case, yet, after examining the ground over which the suggested route would have passed, it does seem that the evidence is very strongly against its existence. From Westmarsh to East Stourmouth and from West Stourmouth to Grove Ferry, stretches of 1| and 1 mile respectively, the route crosses aUuvial marshland none too firm to-day and almost certainly forming part of the estuarine area at that time. Road construction on such land is at best a formidable undertaking, and it would have been so easy to avoid it here by foUowing an aUgnment shghtly more to the south at Westmarsh that it seems certain no Roman engineer would have chosen it. Moreover, apart from the coincidence of line, there seems to be no sign of any road metal or agger, particularly near the straight watercourse at Westmarsh, and the weight of evidence seems altogether against the probabihty of such a road. 1 Arch. Oant., LIV (1941), 38. 132