The Medway Crossings of the Pilgrims' Way

THE MEDWAY CROSSINGS OF THE PILGRIMS' WAY By P.A.TRIOK THORNHILL, B.A. SOME years ago, Margary conoluded an artiole, 'The North Downs Traokways'l with the words, 'Thus it is not, I think, practicable or desirable to select one point as 'the crossing' point of the Medway for the old tra.okway, for there were probably several such orossings.' This was well said, at the time, but there have since been some investigations not directly concerned with the crossings, as well as some new borings at Snodland and Burham, which throw fresh light on the matter. If we may, for the purposes of this discussion, emancipate the 'Pilgrims' Way' from its Gothic type and jettison its medieval pilgrims, we can then define the Way as a band of ancient tracks that run for the most part along the summit and the south-facing escarpment of the North Downs. In the main, three tracks have survived: a summit track, a spring-line track near the foot of the escarpment and, in between, the track which by its obvious continuity on maps has usually attracted the gothic type. We need not suppose that the routes in this band were necessarily all in use at the same period, and there may well have been some separation of function. Sixty years ago a stretch of the Way between Puttenham and Compton in Surrey was known locally as the Sheep W allr. The writer also remembers seeing, about 1906, what seemed like a great river of sheep flowing slowly along the Horsham-Warnham road in a cloud of dust, followed by a train of waggons carrying the bleating casualties of a migration from the South Downs to the arable land around Dorking and Guildford, where the sheep would winter on root crops. This-though not on the Way-was a belated survival of 'transhumance', a Neolithic way of life which, greatly diminished, persists in Spain along well-defined tracks from north to south, and in the Balkans, where the migratory herdsmen may be foreigners in the lands they traverse. We may seek the origins of our own Way in prehistoric transhumance rather than in medieval piety. Moreover, it would be natural for cattle drovers to take the spring-line Way for their thirsty herds, leaving the high pastures of the summit Way to the shepherds. This is but speculation, hut it may be relevant to the question of the Medway crossing, since cattle could use a ford that would drown sheep. 1 I D. Margary, 'The North Downs Main Traokways', Arch. Oant., btiv (1961), 20-3. 91 PATRICK THORNHILL Within the area shown in Fig. I, the North Downs escarpment is breached by the Medway in a vale that narrows northward towards Rochester. The 'official' Way takes us down this vale through Upper Halling and Cuxton to the Rochester crossing, where the river escapes towards the sea between Strood Hill and Castle Hill. From the latter, the Way turns back southward through Borstal and Burham before resuming its normal course along the Downs escarpment east of Blue Bell 􀃉II. It seems that only the presence of a bridge or shallow ford at Rochester, and the absence of any other crossing suitable for sheep, could justify such a detour. There is no clear evidence that there was such a bridge or ford at Rochester before the Roman occupation, but it was a town of sufficient importance to possess a mint and it certainly guarded a crossing of some kind. An alternative to making this northward detour would be to tum southward at Coldrum and strike along the Lower Green.sand country to cross the Medway at Aylesford and thence take the line of Wood Road ba.ck to the Way near Kit's Coty. The presence of megaliths near either end of this route may be significant, but the Greensand country must have been more densely wooded than the open escarpment of the Downs, and the going less easy; moreover, on this route one had to cross and re-cross the band of heavy Gault Clay that separates the Chalk from the Greensand. Somewhere between Aylesford and Rochester there must have been a more direct crossing, but when we look for it we find that the Medway meanders from side to side of a valley floor covered with alluvial mud which in its natUI·al state-and even today, in part-forms marshes that are a more formidable obstacle than the river itself. Speculation about the crossing has always assumed the presence of this marshy barrier, but studies by Evans2 and KirbyS of the alluvial deposits in the Medway estuary, combined with the evidence from recent borings at Snodland and Burham, show that there has been a striking change in the state of the valley floor since prehistoric times. To simplify a complex story: In the final stages of the Pleistocene glaciation or soon after it (say, 12,000 to 10,000 years ago) the 'land surface stood about 200 ft. higher than it does now, and the proto• M;edway, wide as the modern Thames at-Woolwich and ·much swifter; scoured out a channel down the vale and spilled gravel, sand and silt into the proto0Thames and over the lowland that is now the North Sea. Then, through the Mesolithic (say 10,000 to 5,000 years ago) the sea-level rose as the northern ice-sheets melted. The river was shrinking, • 2 J. H. Evans, 'Archaeologioa.1 Horizons in the North Kent Marshes', Arch. Oan,., lxvi (1953), 103-46. • R. Kirby, 'Sedimentary Environments in the lower Medway Estuary''. (Unpublished Ph.D. thesis, London, 1969; copy in Rochester Reference Library). 92 · P.T. ISOO YA "05 1000 500 MILl 0 KfLOMlTAl 0 900 IOOO ANC/£NT WAYS: Vi1ibl• o, fin• on ground or •op - Visibl• o, 'hollow.wily' Proj•cttd or infftrr•d Mod•rll rood• -Fig. lo - Pori •h bou11dory, wh•r• r•l•11ont ....... . Li•it of Goult Cloy --_ ,,, Allur,u• ;}CK􀃋 Churc h M,golith or Otllfr prt-Ro•on •it, + Ro•on ,;,, R Spring S H•ig/lt1J11 fnt ,,15 Cho//t. or cloy.pit• c ,.. .."'l ----- 1500 MlT"lS Fm. l. (Inset) Fro. la. The Medway Croeaings of the Pilgrims' Way. Baaed upon 1M O.S. Map with IM Sanction of the Controller of H.M. Stalionery Ojfiee Crown Copyright ruerved. JJOXL! [/au p. 92 THE MEDWAY CROSSINGS OF THE PILGRIMS' WAY slowing down and dropping its gravel in its own channel, for it could now carry only its finer sediments down to the advancing sea. This gravel-filled channel, five or six times as wide as the modern Medway at Snodland, wa,s eventually to become the 'buried channel'. The Thames and the Somme have similar, though larger, buried channels. The Geological Survey's memoir4 says, 'North of Halling 15 ft. of gravel floors the Buried Channel, and at Snodland, boreholes for the paper mill show 8 ft. of gravel beneath a.bout 22 ft. of alluvium, while a further borehole there shows 32 ft. of gravel beneath 6 ft. of made ground.' We cannot be sure that any of these borings show the full depth of the bed of the channel, but our main concern is with its upper surface and with the alluvium that thinly covers it. Its precise course beneath the alluvium has not been reliably mapped; Kirby (op. cit.) has attempted such a map, which is probably reliable for the Medway estuary but further upstream is not confirmed by the latest borings. By the early Mesolithic, the North Sea was advancing into the Thames estuary and the tides were beginning to be felt in the lower Medway. The effect of salt is to flocculate the clay suspended in fresh water and make it settle faster. While the flood tide spreads it over the riverside flats the ebb is normally not strong enough to dislodge it, and far upstream from the salt water the tide is felt in a ponding back of the freshwater current, which results in a similar, though slower, deposition of silt. The greatest proven thickness of alluvium in the estuary is 98 ft. at Grain, and at Chatham dockyard it is between 50 and 30 ft., but at Snodland it is generally less than 15 ft. thick. While the general tendency since Mesolithic times has been for the sea to submerge the land and deposit alluvium, there have been spells of tidal retreat when the surface of the mud flats has been exposed above the water and become clothed in vegetation which, with the resumption of subsidence, has been covered by more alluvium and eventually converted into a peat bed. In the alluvium of Chatham dockyard Kirby identifies three peat horizons. The lowest corresponds to a. Mesolithic land surface now at least 35 ft. below O.D.; the next, Neolithic, is at 20 ft. below 0.D.; the uppermost, at l ft. above O.D., is the Roman occupation level, which was a.bout 15 ft. above the then sea-level and therefore beyond the reach of the tides, as is shown by the presence in the peat of hazel, birch, elm and oak By A.D. 800 the sea was again invading the land, and at the Grain salt-marshes the upper peat bed now lies beneath about 9 ft. of post-Roman silt . .A strip of alluvium, half a mile or more wide but only a few feet thick, now extends up the floor of the Medway vale and even widens 'Memob:s of the Geologicnl Survey: Geology of the Country around Chatham, H.M.S.O., London, 1964. 93 PATRICK THORNHILL out on the Gault, as in other Wealden river-valleys. It completely covers the course of the buried channel, and the records of the older borings in the files of the Geological Survey unfortunately lump together the alluvium and the underlying gravel as 'Drift', so they tell us nothing about the depth of the inter-face. However, in 1973 some new borings were made by Messrs. Arnold and Nathan for a riverside extension of Snodland Paper Mills, and across the river borings were also put down for a new station of the Medway Water Board. The writer was kindly allowed to see the records of both; the terms used in them are lithological rather than stratigraphical but it has not been difficult to reconcile them and construct a section, Fig. 3. The section was constructed on an east-west line through the Rochester Stone in the grounds of the M.W.B. station, so the Snodland borings are north of the line and those of the M. W .B. south of it, at the distances shown. The surface of Snodland Common is at 9 ft. O.D. and the bed of the river is, as nearly as one can judge, at or near O.D. at this point. The peat bed shown must be that of the Roman level. Incidentally, much of Snodland Common has recently been scooped out to form a lake, and peat fragments can be picked up on its slopes. Beneath the peat the Snodland borings show only 5 or 6 ft. of alluvium above the buried channel gravels. Even if one grants that its deposition was much slower than in the salt-marshes of Chatham and Grain, there are no signs of a Neolithic peat layer at its base and it is reasonable to conclude that the surface of the buried channel at Snodland was not covered with alluvium . until well after Neolithic times. So, if the Way be older than the megaliths that cluster around it, which can hardly be denied, it must have been well established before the alluvium was deposited and the marshes created. The river itself was there, of course, though it was no longer the great post-glacial waterway that had scoured out and filled in the yet-to-be-buried channel; it may even have carried less water than it does today, for the Medway is the most aggressive of Wealden rivers and has long been enlarging its basin at the expense of its neighbours. Moreover, it was not yet tidal and it flowed among the gravel and sand banks of the channel, which probably split it into a number of interlacing streams that could be forded without difficulty. Granted, then, that the Medway valley was much less of an obstacle than muddy marshland was later to make it, where may we look for the earliest crossings1 Above Trottiscliffe the 'official' Way runs eastward at about 400 ft. above O.D. between a, summit Way about 250 ft. higher and a springline Way some 100 ft. lower. The last, following the line of springs that break out at the junction of the Chalk with the Gault Olay, passes the megalithic burial chamber of Ooldrum to reach a strong spring at 94 THE MEDWAY CROSSINGS OF THE PILGRIMS' WAY Birling place. Above it, Crookhorn Hill projects eastward from the escarpment and here the summit Way comes down in a hollow-way to join the 'official' Way which swings northward through Upper Halling and Cuxton towards the Rochester crossing. But the spring-line Way, through Birling Place, continues by Paddlesworth to Holborough Hill, or what is now left of it-for monstrous excavators are here removing the chalk landscape at a terrifying speed. A length of this route has now disappeared but one hopes that the rest of it may survive, for it is here that Jessup in 19545 completed the excavation of Holborough Knob, a hill-top Roman barrow close to a ploughed-out Bronze-Age circle and an Anglo-Saxon cemetery-ample evidence of the longevity and importance of this route. (Margary, in the article quoted at the beginning of this paper, wrote of the junction of the two upper trackways leading down past this tumulus; if it did, it probably headed straight for the Knob, but it is not now visible on the ground). The lower Way passes close by the site of the Knob and converges on a lane (not shown in Fig. 1) which meets the A228 on the south side of the Ho I borough spring; but a map of Halling and Snodland in 1634, to which the Clerk to the Halling Parish Council kindly drew the writer's attention, shows that the lane at that time passed to the north of the spring and its stream (Figs. la and 2). It met the Rochester road about a hundred yards north of the stream and here, in 1634, it ended. Across the road lay a fringe of squattings and some fields that ran down to the alluvial marshes. H the ancient Way continues in a straight line beneath the alluvium here, it should reach the riverside at the northern end of the existing wharf belonging to the cement works. Until recently the Rochester road was crossed, a further 300 yds. to the north, by the tram rails of a large chalk-pit up the hillside to the west, and the standard-gauge rails are still to be seen on the westward track, which is straight, as one would expect a way-leave tramway to be. But one finds that this track forms part of the Halling-Snodland parish boundary (Holborough is in Snodland parish), and the map of 1634 clearly shows it crossing the Rochester road and almost reaching the river very close to the future site of the cement wharf. So here we have the convergence of a prehistoric and an apparent Roman route on the left bank of the river, both within 150 yds. of the northern end of the cement-works wharf. The chalk-pit, fortunately, was abandoned, with its industrial victoriana, before it had eaten away the whole face of South Hill. The pariah boundary persists through the pit and then continues in a straight line up and over the crest of the hill, where one finds that it runs in a hollow-way with traces of flint metalling, between yew trees. Its farther 6 R. F. JeBllUp, N. C. Cook and J. M. C. Toynbee, 'Excavation of a Roman Barrow at :S:olborough, Snoclle.nd', Arch. Oant., ttvili (1954), 1-61. 95 P.T. Fm. 2. Holborough in 1634. THE MEDWAY CROSSINGS OF THE PILGRIMS' WAY course north-westward is irrelevant to the subject of this paper. (It would be less than surprising if it were found to lead towards Henley Street, Gold Street, Sole Street, Round Street, Nash Street-and Watling Street.) Our present concern is that it comes over the Downs from the west and from the top of South Hill the line of this straight road could be sighted directly on the Medway crossing, two miles away. If we now move over to the eastern limit of Fig. 1, we find the 'official' Way coming in westward through Boxley at about 350 ft., and apparently converging on the spring-line Way at that chalky melange of megaliths and motor-roads that once was Blue Bell Hill. This united Way passes between Kit's Coty and the Countless Stones (Lower Kit's Coty) and becomes a, dual road for most of the mile to Little Ouland. It seems possible, though, that the two Ways remained separate here, for the writer came upon an a.pparently unrecorded group6 of at least a dozen large sarsens at Great Culand (N.G.R. TQ 73526143) precisely where the middle Way would have crossed the track that runs down from the Upper Bell to meet Bull La.ne, Eccles. A large oblong chalk-pit has obliterated the Way to the east of the Upper Bell tra,ck, but on the west side a path continues the line from the sa.rsens to join the Rochester road, which is here officially the 'Pilgrims' Way'. It seems likely, too, that from Great Culand there also sprang a westward branch that becomes Court Road, Burham. The land about here has been much disturbed and built upon, but whether we follow Court Road or project the line of the 'official' way westward from Little Ouland, we pass close by the site of a Roman hypocausted building, examined in part by Payne7 in 1896 and still visible as a stony scatter on a ploughed field at N .G.R. TQ 726617, and we reach the river just opposite the Holborough cement-works wharf. Here we a.re close to a riverside Roman building, described by Payne as 'a supposed Mithraic temple' but later secularized by Jessup8 as 'a, cellar for storing wine and oil imported by its owner, and perhaps corn for export', which suggests bulk transport by road, and a navigable river. It is here that the Medway, after meandering across the low-lying Gault Clay, rubs its right bank against the Chalk and straightens out for a while along the eastern edge of its alluvial strip. For a good ¼ of a mile downstream from the Roman cellar the right bank is firm and high. This has attracted modern industry and resulted in a jumble of • It may be one of those referred in Beale Poste's notebooks as being 'near Kit's Coty House'; see J. H. Evans, 'A Disoiple of the Druids', Arch. Gant., !xii (1949), 130-0. 7 G. Payne, 'Romll,Il Disooveries', Arch. Gant,, xxiii (1898), 1-23. • R •. F. Jessup, 'The "Temple of Mitlttas" at Burham', Aroh. Oant., lxx ( 1956), 168-71. . · . 97 1 :PATRIOK THORNHILL occupied and derelict sites, of crumbling concrete walls and rusty iron, while low tide reveals a foreshore of broken bricks and other debris, covered with grey-green slime. The site of the earliest crossing must surely have been here, but if we turn our backs on the river and look eastward for converging approach roads we find we are beset with chalk-pits. Yet one lane, Hall Road, starts directly on the line from South Hill. It runs past the site of Wouldham Hall and continues for nearly ¼ of a mile. Then a. chalk-put cuts it off, but the reprint of the 1819 edition of the One-inch O.S. map9 shows it as continuing up to the Rochester road at Scarborough House. Just to the north of that house the 1819 map also shows a track that winds upward through the woods of Wouldham Common to the top of the Downs escarpment. This track is not scored into the hillside and shows no signs of great antiquity; indeed, its careful gradient and the presence of a concrete dew-pond half-way up seem to indicate its use by modern breeds of sheep rather than by their skinny ancestors. To the right of this track the writer was surprised to discover in the undergrowth a deeply scored hollow-way running straight down the escarpment. It is not shown on any O.S. map, which suggests that it had gone out of use by 1819, being too steep for the fatted progeny of Bakewell's rams. It aims straight at Scarborough House, thus completing a route from the top of the Downs at South Hill and across the Medway to the top of the Downs at Wouldham Common. We have still to consider the claims of Snodland village, through which the route of the Way would be slightly shorter than through Holborough. If we start again from the west and move eastward through Ooldrum and Birling Place, we find that at Paddlesworth a lane branches off the Way to the right and runs straight down through Snodland to reach the river close by the church. Again, just below the top of South Hill, above the chalk-pit, a track branches off from the Holborough route, crosses the middle Way at La.d's Farm and heads in the same direction. Both of these tracks branch off from the convergent Holborough routes to converge on Snodla,nd, and this fact points to the latter route being less ancient than the former. One thing is quite clear: nobody would have chosen the SnodlandBurha,m crossing if the river had been where it is today, coiling in a great S-sha.ped meander between Snodland church and Old Burham church. This meandering, now fixed by embankments and industrial dumping, would have been further upstream in Norman times, and earlier still, before the alluvium was laid down, the river, as 11,lready suggested, probably fiowed in a number of streams through the gravels of the yet-to-be-buried channel. • Ordnanoe Survey, One-inoh mo,p of Maidstone distriot, first edition of 1819, reprinted by David and Charles, 98 THE MEDWAY CROSSINGS OF THE PILGRIMS' WAY Fig. 3 shows that the surface of the buried channel must be close below the present river-bed about here, and the Geological Survey memoir4 tells us that gravels of the buried channel 'are sufficiently thick and extensive to be dredged south ofSnodland' and then: FT. 20 10 'Gravel •.. appears near the surface, 600 yards east of Snodland Church in the Medway channel, to form the "Snodland Rocks"-a bar of iron-cemented conglomerate, with large and small flints and pebbly material (including chert, ironstone and ragstone from the lower Greensand) visible at low tide.' SNOOLANO PAPEP. MILL MEDWAY WATER WORK -- --- Fro. 3. Seotion across the Medway Valley at Snodland. I. Made Ground; 2. Peat; 3. Alluvium; 4. Buried Ohannel Gravel; 5. Gault Clay. This interesting feature is by no means easy to find and investigate. The writer has seen it only twice, at low spring tides and after a long drought, when the deep alluvial mud of the river's banks will just bear one's weight. At low water one can see that the river here cuts through about 10 ft. of alluvium to expose the gravel of the buried channel in its bed. 'Gravel' is perhaps a misnomer, for there are not many pebbles in it and most of it is a rubble of fist-sized flints and pieces of chert, mostly sub-angular and showing little sign of being water-worn. It forms a alight bar that runs diagonally across the river, with a long 'tail' on its downstream aide. About 50 yds. upstream is a partial bar that contains a few large (nearly unliftable) stones. The latter is only slightly south of a straight line drawn between Snodland and Old Burham churches (Fig. la), and both appear to run parallel with such a line. There is no visible continuation across the other two curves of the river, and the tongue of land within the horse-shoe meander, which once contained a gravel-pit, is now deep in industrial refuse. The formation of a natural bar, athwart the south/north run of the buried channel, would be hard to explain, and the Geological Survey memoir does not attempt to explain it. We feel justified, therefore, in 99 PATRICK THORNHILL suggesting that the Snodland Rooks mark the course of the beaten track or causeway that carried the Way from Snodland to Burham. Other possible crossings deserve a mention, if no more. There may have been one between Halling and Wouldham, though this would have lengthened the detour for the Way. Southward, the important Roman villa at Eccles stands on several feet of Hill wash Head overlying the Gault Clay; traces of a Roman road are said to have been found10 on the sides of the large clay-pit close to the villa. If the continuation of Bell Lane, Burham, marks its course it passed by the villa to arrive at the later site of the New Hythe ferry. It would be beyond the scope of this paper to trace the changes in the valley-floor of post-Roman times-the upstream encroachment of the tides (beyond Maidstone until Allington lock stopped them), the spread of alluvial marshes, the embanking and consequent confining and deepening of the river, the change from fords to ferries (Aylesford is the lowest named ford), and eventually industrialization and the decline of the ferries. The following conclusions appear to be consistent with the evidence: (1) There was a prehistoric crossing of the Medway at Rochester, which may then have been at or near the limit of tides. (2) A prehistoric Way ran by Coldrum, Biding Place, Paddlesworth, Holborough Knob and Holborough spring, and crossed the gravel and sands of the yet-unburied Medway channel close to the northern end of the present Holborough cement-works wharf (N.G.R. TQ 71036291 ), whence it continued eastward by way of Blue Bell Hill. (3) In Roman times a road was made, or improved, from the escarpment summit at South Hill straight to the Holborough crossing and then via Hall Road and Scarborough to the summit above W ouldham Common. (From South Hill a mile-long short cut was also made to the Way leading to Rochester). (4) Later, a branch from (2) was made at Paddlesworth, and a branch from (3) at South Hill, which converged on a crossing at Snodland that apparently went by way of Snodland Rocks to Burham and the Blue Bell Hill route. This suggests, though it does not prove, that the advance of the tides up the valley was by later Romano-British times forcing wayfarers to abandon the old Holborough crossing in favour of the new Snodland route. 10 V.O.H. Ke11t, ill (1982), 145, 153, too