Roman Roads in West Kent

&XowM$jj^i^& ^n M>e CAtrrmBURY\ . MAIDSTONE V -O ^ , ^SOTTOH VALENCE *t£o .PLUCKLEY \STAf>lEHl/RSr ^-^ .jAsf/ronO JSissmaiuiar HIGH HALM*,—^Kmuim ,6/sswe/ttwsr •T^>^ SMDOXHUSS fi°"l Aumorw ,„( x- Imntc] 7iMTERaEII BmcNDeu <& HiMtore CDLESC0HBE~ —JZJZ&&$$!i$\'&* ^.DATTLB WesrriELO ROMAH ROADS RIOOEUAYS ROMAN IRONWORKS X ESTUARIES •^HASTINGS SCALE or Allies PLATE I. ROMAN ROADS IN WEST KENT. ( 29 ) ROMAN ROADS IN WEST KENT. BY IVAN D. MARGARY, F.S.A. THE Wealden area of West Kent is intersected by a group of Roman roads which, although well known in certain parts and even marked upon the Ordnance Survey maps for many years, have, nevertheless, received very little attention from archaeologists. Scarcely any reference to them can be found in the hterature, save for a recent article in the Roman section of the Victoria County History of Kent, and no attempt at a detailed survey of them seems ever to have been made. There are three main routes to be considered : I. Rochester—Maidstone—Hemsted—Bodiam—Hastings. II. Maidstone—Kingsnorth—Lympne. III. Hemsted—Tenterden—Ashford—Canterbury. Of these, the first was laid out on a course nearly due north—south, with the obvious intention of bridging the Wealden area in a manner exactly similar to the London—Lewes, London—Brighton and London —Chichester roads farther to the west. In this case, however, being much farther to the east, the road connected with WatUng Street at Rochester for its northern terminus, whilst its southern destination was clearly to be the iron-working region that lay inland from Hastings. Moreover, this road was near enough to the important settled area of East Kent for the need of connecting branches to it to be felt, and it was this which led to the provision of the other two roads, one from Maidstone to the south-east, the other from Hemsted (the most southerly point that was conveniently accessible from the east) to the north-east and Canterbury. Thus the three roads formed a triangular network giving inter-communication between the Rochester—Maidstone area in the north, East Kent, and the Wealden iron district in the south. THE ALIGNMENTS. Considerable lengths of aU these roads were evidently laid out upon properly surveyed ahgnments, selected with the usual care and eye for country that is such a striking feature of Roman road construction, but the ground in the Weald is often difficult, necessitating much local modification of the main ahgnments, whUst in some parts the definite remains show quite plainly that rigid straightness was not insisted upon. It may weU be, too, that the date of these roads is relatively late and their purpose entirely commercial, circumstances that would also lead to a less rigid appUcation of constructional rules. 30 ROMAN ROADS IN WEST KENT. I. ROCHESTER—MAIDSTONE—HASTINGS. The following aUgnments are clearly shown : (1) Rochester—Horsted. (2) Horsted—Bridgewood Gate. (3) Bridgewood Gate—North Downs escarpment. (4) Downs foot—Maidstone North. (5) Maidstone North—Mangravet Wood. (6) Mangravet Wood—Amber Green. (7) Amber Green—Staplehurst (Iden Bridge). (8) Staplehurst (Iden Bridge)—Hemsted Park. (9) Numerous short aUgnments onwards to suit the ground. Of these, (1) and (2) form very nearly a continuous Une, only turning very sUghtly at Horsted to foUow the ridge there, and (4) and (5) are much the same, (3) being a short Unking aUgnment along the edge of the steep Downs escarpment to the chosen point of descent. Thus far the intention is clearly to provide a direct north—south route. Next, with (6) we have a distinct turn to the south-east, and, since the point where Route I descends the greensand escarpment towards Staplehurst is not more convenient than a similar point farther west, near Boughton Monohelsea, would have been, it seems likely that (6) was definitely planned that way to lead to a fork for Routes I and II at Amber Green. With (7), the longest and most obviously Roman of aU the aUgnments on this route, we resume the southward direction, pointing straight at the area of the Sussex ironworks. This is the last aUgnment that the road closely foUows, for, although (8) is indicated by the general layout of the remains, it was greatly modified to suit the ground which becomes increasingly difficult beyond Sissinghurst. Indeed, it would be almost impossible to lay a straight route through this country of deep gills and steep-sided ridges, and short straights along such ridges as were available was the obvious method to be followed. II. MAIDSTONE—KINGSNORTH—LYMPNE. From the fork at Amber Green, one main aUgnment seems to have been laid, from the greensand escarpment at Sutton Valence to the beginning of the coastal range of hiUs at Aldington, south-east of Ashford, though it was locaUy modified in some hilly areas, as near Pluckley. On reaching the coastal hills the road was laid out in short alignments to follow the ridge. III. HEMSTED—TENTERDEN—ASHFORD—CANTERBURY. Two major alignments are traceable here : (1) Tenterden, St. Michael's Church—Ashford, Stanhope School. (2) Ashford—Godmersham Downs. ROMAN ROADS IN WEST KENT (Ii\feetitw „e-..«**» orated Bmattp A h "jA-fl Mffro$unWet S Ground *•«»•*» '» Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 32 ROMAN ROADS IN WEST KENT. SSL***©/ > ' / S i W » a-LU^/h^Si>.'> mood wr *> / ) . * * f »W $rteiraff? OAD IMrACT Sprint ClMDtA SECTION HOUOWJ Cfitrm «««<* A » . /ve"r< &mem ho^£«5-o. ROM> 8&%2;i Stream Y*i>>® Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 42 ROMAN ROADS IN WEST KENT. fV^ [envy a?? £ ?»1 •" «.•; S ^ i^ *•»*•<©« i*j«S3b Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office ROMAN ROADS IN WEST KENT. 43 WfPlacei,, Million B.M 27 e Bodiam Mill re 4 UohenXiwi Mtoat S/^iinonot fious< (SiU of) 7petet*8 Green 1 7x-v •'•*<• ~iSMALLfiWmi TsxKAce/ ^ x 5C /fociou. TERRACE'' Church Court lodge W4V > ~$X09KOUT tk.i3)\» *Va*#\ ^ > J8#J> 'SL^--'**™ \ .Ifw/mi&nuwi S^^ YV *: '.^^^Uiill/' v—" )='':• ,%««..«- A « Vis*=*r*5^ V • * « • . "/boMaft-Castlcaa « . \ \ . V i - i ^ ^ v S L * K . J . ' . •'« • U, Bodiam-Bridge Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 44 ROMAN ROADS IN WEST KENT. glaring example of this occurs to the south of Sandhurst rectory, where they extended the Une of the derehct lane as a straight Une pointing towards Bodiam ; this Une actuaUy runs across a quite impossible slope upon a steep hiUside, and, moreover, completely ignores a visible terraceway curving round the hiUside which is the obvious continuation of the road. Again, north of Sandhurst, they extended the alignment through Sponden too far, thus ignoring the obvious continuation along Sponden Road which fits in weU with the known length towards Sandhurst, whereas the two aUgnments as marked upon Kent Sheet 78 SE. are obviously incompatible with one another. But their worst and most misleading effort was at Kingsnorth on Route III, where they extended the aUgnment for 1,500 yards east of Stubb Cross, a purely imaginary length which has misled everyone as to the real continuation of the road towards Canterbury. In this case, too, there was, as we shaU see, abundant evidence of the metaUing along an obvious Une of hedgerows a Uttle to the west of their hne, the road is known to local farmers, and, most curious of aU, it is the only section of these roads for which there was literary reference available, Hasted giving a very inteUigible description of the route in his History of Kent.1 Again, it is evident that no one reaUy conversant with the course of Route III on the ground, through Tenterden, St. Michael's and Parkgate, especiaUy in Dawbourne Wood, near Breeches Pond and across the deep giU at Bishopsdale, could regard the Survey's Une as anything but imaginary or ideahzed. From all this it is clear that we shaU be justified in considering the Ordnance Survey's routes as no more than an early and honest, if clumsy, attempt at recording local information of very unequal value. II. MAIDSTONE—KINGSNORTH—LYMPNE. From the junction point in the orchard at Amber Green a shght agger can be seen running east from the parish boundary stone to the garden of the cottage at the corner of the Four Wents—Amber Green road. Just beyond Amberfield a hedgerow foUows the Une for 240 yards to Almery Cottages, and again for 380 yards farther to the Chart Sutton —Norton Forstal road, the ridge showing faintly in the field to the south of the hedgerow as it nears the road, for it is diverging from the hedge there. AU this land is under intensive orchard cultivation, and i f is not surprising that few traces remain. We now approach the steep escarpment of the greensand ridge at Sutton Valence, and the last portion of a lane which joins the main road north of the church may represent part of the course, for there are indications just east of the road there, opposite the new Council School, of an old terrace leading on eastwards behind the houses. Another Roman waUed cemetery containing about a hundred burials was found Second edition (1798), VH, 584. ROMAN ROADS IN WEST KENT. 45 only 170 yards north of the course there,1 in the grounds of the new College buUdings, just above the first descent on the escarpment, a suitably prominent spot for such a site. From the centre of the viUage the course next foUows the lower road past Sutton Vicarage, Sutton Castle and Sutton Place to Boyton Court. This is an underhill terrace of ancient origin, foUowing the main aUgnment closely, though, naturaUy, it winds somewhat along the hiUside. The Castle keep, a square stone tower, stands conspicuously on a spur immediately above and adjacent to the road which, obviously, provides the reason for its presence at this point in Sutton, for it is not a position that would otherwise be of any particular importance there. At Boyton Court there are no further visible traces, but the rest of the descent along the general aUgnment was easUy practicable, and upon approaching the lane that comes down from East Sutton, caUed Friday Street, traces of a hoUow, and then, across the lane, a length of hedgerow with a hoUow beside it mark the course. It is again marked by hedgerows 130 yards south of the Prince of Wales Inn farther east, but after this there is nothing for some 2\ miles, although the south side of Kingsnorth Wood Ues upon the alignment, and some hedgerows west and east of Woodsden, near Southernden, probably mark it. East of Southpark Wood, Southernden, there are traces of a shght ridge, with local stone, although we have now got on to a stoneless clay country, leading down to a stream beyond which it is continued by a line of hedgerows across Bedlam Lane and through Wanden until, near The WUderness, it meets a straight track leading east through Frith Wood. The track is quite a featureless woodland ride now heavUy covered with wartime rubble, but it leads into a road running past Frith Farm towards Pluckley which, in turn, connects with a long line of hedgerows that does appear to have some significance. At the east end of the Frith Farm road, where it forks at The Pinnock, a hoUow can be seen north of the present road leading more directly towards the commencement of the hedge Une at The Pinnock, and this is then continuous for a mile to a point near some cottages with the curious name of Stone Abbage. The Une curves sUghtly, apparently to suit the ground, but there is a definite appearance of a shght agger along it, with traces of stone, although the soU there is notably stoneless. About 440 yards east of Stone Abbage the lane to Dowle Street carries on the Une for a short distance and then leaves it again, but a hedgerow with considerable stone in it continues in the same direction, passing 190 yards to the south of the old farmhouse at Dowle Street. Beyond the lane leading south to SnoadhUl a hedgerow with traces of an agger and scattered flint leads directly to March Wood and, beyond, to the beginning of the large Hoad's Wood. This wood and Etchden Wood cover the next -f mUe over which there appears to be no trace, 1 Arch. Cant., X, 166 and XV, 88. 46 ROMAN ROADS IN WEST KENT. \t> -IB HOLLOtf fctSERW SLIGHT X • l > i . r.^m&&- Liask *ta£ Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. ROMAN ROADS IN WEST KENT. 47 .;, ^ Sr ° Sr ;*;w. % STONE I a HEDOeRa * HEDGE.-! F ^ RQW\ / c_ IfZ—' < fr / \ /'.'A \ i. • / \dJ v ^r __ .^ * " — ^ M ^ .— IT \ L -A rT\^ ; / u K i e t — ^ i \ ] / ' \ rf : i \1 T^5-JL \ I \ ^S-J-/ ^^^^rfr^ff . w1 rls^^JN S !"l \ f1" %o. A/ ^*^* s • ^ A§ ^ H i «SWf; "•""*•-. . . -' ->j^ > c> 1 Y^r ^v„ $ — : ' 0 0 ~/ "v ^^^^ 0\ *> SLrfl^ • * > T / 1 ** l ^>n 1 j . ' * i ^ ^ ^ I / J '»! 1 ^"V > - T ^ I C « = ^ ^ : \ _j£^5tr§ff*'L---—^ f p - ^ j ^ w ? * -"T ' " "IN. I — V f^ *-) / f i t 1 1 >i»•" \ L\ j-^vy- if——on / rf Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 48 ROMAN ROADS IN WEST KENT. '.'HepceRow °s1 'I fri. c\d & *",.» ».\ A *!/ QU0W % IY««?«? \ \ %?*** "a .«>ra ^ Based upon the Ordnance Survey Map, by permission of H.M. Statioriery Offioe. ROMAN ROADS IN WEST KENT. HEDGEROW ON iwe\ Jr* ** n On 1. oV *•• »_ . <' TP n . ROAD ° ON s*S». K»fe.^i»7»::£?bJrl»,!| 3 v* *«-*.*»» &!&*&£ HmERou M Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 50 ROMAN ROADS IN WEST KENT. for the visibihty in Hoad's Wood is fairly good. At the far side of Etchden Wood, however, a large bank, 27 ft. wide with a broad hoUow along its north side, extends towards Upper Coldharbour Earm for 170 yards. As it now stands, this bank may weU be an artificial rabbitwarren of medieval origin, but it must be admitted that, if so, it is a remarkable coincidence that it should happen to come exactly upon the true aUgnment of the road. Moreover, it is continued much more faintly into Etchden Wood, and it seems on the whole more probable that the original agger remained sufficiently distinct at the edge of Etchden Wood to have suggested its conversion to a rabbit-warren by widening it with additional material from the hoUow on the north side. Nothing is to be seen past Upper Coldharbour until a short piece of lane comes upon the line for 230 yards, and a very sUght ridge can be seen across the fields direct from Little Singleton to Moat Farm, Great Chart. Beyond this there is nothing for a mUe, until at Cuckoo Wood traces of an old hedgeline and then a line of hedgerows from the north side of Joy Wood to Stanhope School, near Ashford, seem to mark the course. This brings the road along the south side of Stanhope School gardens, where much stone can be seen, to the cross roads near Effingham and Westhawk Farms which is the point of junction with Route I I I from Hemsted and Tenterden. At EUingham Farm there is a shght curve at a crossing of a branch of the East Stour, and then the commencement of the drive to Park Farm, Kingsnorth, represents the course, although, farther on, this passes to the north of both the present farm and its adjacent earUer moated site. Then there is no trace for 1,100 yards across some low-lying ground to Bilham Farm, where the well-known portion of the route begins, first as the farm road and then as the high road to Aldington, Lympne and Hythe. It seems very probable that it was this wet area near Bilham Farm which caused the road to go out of use west of this point. The evidence here adduced for the " lost " portion of Route I I west of Kingsnorth is admittedly sUght, though it is just what one may expect in such cases. The strongest points in favour of the existence of the road are : (1) the coincidence of the aUgnments from Maidstone southeast and from Kingsnorth north-west both meeting with the Staplehurst aUgnment at Amber Green, (2) the position of the Roman cemetery at Sutton Valence, just where the route reaches the crest of the ridge, (3) the position of Sutton Castle beside the road as it climbs the escarpment, and (4) the traces of stone and flint near Dowle Street and elsewhere, at places where these are not normaUy found. III. HEMSTED—TENTERDEN—ASHEORD—CANTERBURY. At first glance it now seems strange that the south-easterly road from Kent to the Sussex ironworking district should have been directed ROMAN ROADS IN WEST KENT. 51 so far inland as to join Route I at Hemsted, 15 miles to the north of Hastings. Actually, however, there was a very sound reason for this, as it was at that time the most southerly route available that would not have involved ferrying across tidal estuaries. To what extent the area of Romney Marsh was by then emerging from the sea is a matter of controversy, but it does at least seem clear that the Rother, under its old name of Limen, then flowed eastwards along the old coastline as a large tidal estuary to an outlet near Lympne, and that until a much later date there were long arms of tidal water running in behind the Isle of Oxney on each side of Tenterden. It was to avoid these that the road was sited so far inland. The road leaves Route I just by the mansion of Hemsted, proceeding through the park close to the east drive ; it crosses the Sissinghurst— Benenden road 80 yards south of the turning to Goddard's Green, and is visible as a shght agger under the trees just before leaving the park. There also seems to be a trace of it, as a hollow, just east of the new school buildings on the east drive. It appears as an agger with traces of slag metalling along the south side of the narrow shaw beside the Goddard's Green road, and comes up to the Green through an orchard, where it shows as a distinct hoUow. The modern road then seems to Ue upon it past Eaton and Redhouse Farms, for the highway is distinctly raised there and, moreover, the next clear traces, through Uppergate Wood, where it is a large agger, 36 ft. wide, in places heavily metalled with slag, lead directly from the modern road at Redhouse. The course of the road is here directed in a series of short straights along an east—west ridge. This involves a shghtly more southerly trend past Cleveland Farm, which Ues some 320 yards north of the road, the course being quite definitely marked by a hedgerow with remains of the agger and slag metalhng. Next, the Ordnance Survey shows it turning north-east straight across a httle vaUey and up to Bexhill Farm. Traces of the agger are visible near the stream crossing, and near Bexhill Farm, where it again turns almost due east, a Une of hedgerows marks its course to Bishopsdale. These turns might be due to a desire to keep along the ridges as much as possible, nevertheless the southward bulge here from the general Une seems unnecessarily pronounced, and a possible explanation is provided by a Une of hedgerows traceable south-eastwards for about a mile towards Tenterden. Traces of an earth agger and numerous pits appear along this line and it seems quite hkely that this may have been a branch road leading to the head of what would then have been an estuary running inland to a point just north of Rolvenden Station. The bends near BexhiU Farm would then be due to the road having forked there. East of Bishopsdale the slag metalhng is traceable, about 60 yards north of the Ordnance Survey Une which here, as elsewhere, is evidently ROMAN ROADS IN WEST KENT •^% 8. Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. ROMAN ROADS IN WEST KENT. 53 WOA0-JJ * STONE C/ioe. CwDER, •r /ka;& * CINDER S<5*n*a> *. 5 ' Jro/vr -Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 7 54 ROMAN ROADS IN WEST KENT. !o^T= &Sgj 5 5i5 ivV- -• <£"** r»* IA «/fc*l V*|*"i I^Sfu ,>*«» u NT Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. ROMAN ROADS IN WEST KENT. 55 >?*&V t r*g| fcss'Hfi ii Mo&e> *L \ £& asm n DCHO OF STR£AH mow cum ROAD Jt AGGER *, STONE: $ w Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 56 ROMAN ROADS IN WEST KENT. somewhat ideaUzed. A steep giU in FUght Wood must have been directly crossed by a bridge, east of which the slag metalling in the wood is weU preserved. Beyond, at Parkgate, the lane leading past Breeches Pond and Brown's Corner to St. Michael's, Tenterden, clearly represents its course; the lane runs in very straight short lengths designed to fit the somewhat difficult ground near the Pond, which the map Une fails to do, especiaUy near Reighton Wood. It is very probable that the bay of the Pond is based upon the hne of the old agger, as so often happens. The zig-zag ascent at its eastern end is most hkely the original route, too. At Brown's Corner it is possible that the straight lane leading south to ChenneU Park, and showing as an old hoUow way in the park there, was another branch connection to the estuary, giving access from the north-east. From Brown's Corner eastwards the modern road runs very straight along a ridge and there seems no reason to doubt that it represents the course, the Ordnance Survey line to the north being unsupported, and at St. Michael's, where the roads diverge north and south of it, there is a distinct hoUow leading on a Uttle way towards the church, which stands on the Une. Through Dawbourne Wood, to the east, the course is plainly traceable, first as a hoUow in the grounds of " Little Dawbourne," then as the remains of an agger heavily metaUed with iron slag in places. Beyond the wood it runs very close to a hedgerow past Three Vents Farm and into Lancefield Wood, where some traces of the agger remain. A lane then foUows the line for 500 yards, and upon leaving the woods it is marked by a hollow up to the farm of Tiffenden, and then by a hedgerow with a sUght agger along it to Trottingale Wood, and similarly on approaching Brook Wood, through which the agger is also faintly visible. Beyond, at Plurenden, the modern road comes close beside the Une for 400 yards, and the agger is faintly visible in the fields both to the south-west of Plurenden and opposite Plurenden Cottages; also the north edge of Nine Acre Wood, near Great Engeham, Ues upon the line. The agger can again be faintly seen, together with scattered metalling, in the field north-east of the crossroads there, which he exactly upon it. From Brown's Corner, through St. Michael's, to this point the road foUows very closely a single aUgnment, but it now turns very sUghtly more to the east, at a point 520 yards beyond Great Engeham crossroads, in order to avoid a stream near Whitepost Wood. The point provides an interesting confirmation of the ancient origin of the road, because, though the intention of the engineers to avoid the stream is obvious, the hne of the road does in fact cut aoross the most extreme bend there by about 60 yards. Since the construction of the road the stream has evidently shifted its course to this extent, and as it is a small and very sluggish one we may assume that such a change would take a ROMAN ROADS IN WEST KENT. 57 long time. A similar case, upon a larger scale, was noted upon the London—Lewes Roman road at WelUngham, north of Lewes, where a bend of the River Ouse cuts it.1 Beyond Great Engeham the road is visible as a shght hoUow in the field between Mayshaves Farm and Little Ruck Wood, then it is marked by the scattered flint and slag metalhng aU along the hne past Harlakenden, in Whitepost Wood (near the north-east side) and up to the lane north of Little Criol Farm. East of Criol Farm a Une of hedgerows with traces of an agger marks its course for 860 yards past Snailwood, and the agger is very clear, with traces of pebble metaUmg, across the last field to Stubcross Wood, where for 150 yards the modern road takes up the Une. A farmer here has found a foundation of large stone slabs underlying the road south-west of the Wood. At Stubb Cross the road turns more to the north again, making direct for Ashford; its course is plainly marked by a long line of hedgerows with abundant traces of the metalling along its western side to Court Lodge Farm, an old moated site, and again through Westhawk Farm beyond, along or near a hedgerow marking the Kingsnorth—WUlesborough parish boundary. The last part of this, forming the approach road to Westhawk from the north, Ues upon a distinct agger, running up to the crossroads at Stanhope School where it meets Route II. The part of the road just described is clearly mentioned in Hasted's History of Kent2, where under Kingsnorth he writes : " . . . close to the western boundary of the parish is the manorhouse of West Hailes, which has been a large antient building, most probably of some consequence in former times, as there appears to have been a causeway once from it, wide enough for a carriage, which led through the court-lodge farm towards Shadoxhurst, Woodchurch, and so on to Halden, remains of which are often turned up in ploughing the grounds." In view of this evidence and the fact that the road is well known to local farmers, it is surprising that the Ordnance Survey should have been baffled as to its correct course. They continued the ahgnment from SnaUwood straight on beyond Stubb Cross, pointing towards Dennard's Farm at Kingsnorth Pound and ending " in the air " some 1,130 yards short of Route II. This hne is featureless, whereas the remains of metalhng are so striking along the obvious hne of hedgerows a little to the west. The reason for the northward bend at Stubb Cross was actually to regain the main Ashford—Tenterden ahgnment that had been left at Great Engeham to avoid the stream near Whitepost Wood, and thus 1 S.A.C., LXXIV, 28. 2 Second edition (1798), VII, 584. 58 . ROMAN ROADS IN WEST KENT. ^!.-AH£R CihDeR &SroNc ? ClNPER Continued, on map, p. 62. Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. ROMAN ROADS IN WEST KENT. 59 S2*FS Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. 60 ROMAN ROADS IN WEST KENT. OLD i& va-*o?;i-* MINIM Based upon the Ordnance Survey Map, by permission of H.M. Stationery Office. ROMAN ROADS IN WEST KENT. 61 6 ».<38 o S.:«A* !n a-:v|*K:a I'J 4 ...«i"° * : 'W! " » a a " 1 II ?;1 15 ft & " dQa V «» I aa *»,i«» ^ / / aaift / »aai» #*r« * » 44 1 r * ^ ' • im » j « a O^^VJ0 . , ' , ; ' « DO s^ vs fPCA^^s'iIS.^Ax t r#?i & i * - - ' 6 £r" M?T &<* *• we *n \ r> *** u u \ \ »*«»'.?* M,?i a »N n i \ &cB \ >"«•V-. Hfo%fct4f -^ '<&K;^*mo r9»r° \ *sJaJ^B Vj*! \ 5fc& ! a "l S* *, / " flj *» o-i.? ff *-.«> tf a a « « •« iO° »J» ••» -^ %t * V % v> il V »• *" a 1' (k fcU»«.?i *v^ i": »> II « »> '•• & *S IP-J \ i . \