The Reclamation of Romney Marsh: Some Aspects reconsidered

THE RECLAMATION OF ROMNEY MARSH: SOME ASPECTS RE-CONSIDERED JILL EDDISON, M.A. The last major upset and upheaval in the fluctuating land-sea dialogue in the large area of marshland generally known as Romney Marsh took place in the thirteenth century. That is, recently enough to have had profound effects on settlement sites and on the use of land, but too early for contemporary records to be readily available. At that time the port of Old Winchelsea was lost, New Romney apparently overwhelmed' and most of Walland Marsh inundated. Then during the following three centuries so much estuarine ground silted up and was reclaimed that the first local maps, made about 1590, are very similar to that of today. That would have been far from the case had there been a map of 1290. Over the centuries a corpus of 'information' about the sequence and timing of marsh reclamation and of suggested early courses of the Rother has grown up. Each group of scholars and local enthusiasts - Renaissance, Victorian and those of the 1930s - added their own information and interpretations. Not until 1968 was a detailed scientific assessment of the complexities of the marsh sediments published - by the Soil Survey of Great Britain.2 The purpose of this paper is to outline the sequence of marsh reclamation, using that detailed scientific background coupled with additional evidence, and to call in question certain aspects of marsh 'history'. In particular the idea that the Romans built both the Dymchurch Wall and the Rhee Wall and thus reclaimed the whole of Romney Marsh proper (i.e. the area north of the Rhee Wall) artificially, in one move. This was reiterated most recently by 1 E.W. Parkin. The Ancient Buildings of New Romney, Arch. Cam., lxxxviii (1973), 117. 2 R.D. Green, Soils of Romney Marsh (1968), Soil Survey of Great Britain, Bullletin No. 4. 47 Shingle [[[[l Upland Posi Uo, of gaps jn the shingle beaches Oy1:1ehurch Wall Rhee Wall cordon of aedi eeval sea walls Deposits related to gaps in lh!' beaches: A AS B H H 1 Hythe creek ridges (Nole, H 1 probably also had an oullel al New Roov,ey) H 2 Hythe later phase N8 Ne" Romney R/W Rye/Winchel sea Appledore Islesbrigge Appledore Station lydd Sroo1dand N NC\o'!Church fairfield NR 􀀞 Romney llylhe OR Old Romney Fig. 1. Romney Marsh: The Surface Deposits related to the three Gaps in the Shingle Barrier Beaches (mainly after R.D. Green). THE RECLAMATION OF ROMNEY MARSH Williamson in 1959.·' Also that four twelfth and thirteenth century archbishops inned from the sea large blocks of land between Brookland and Old Romney (an idea stemming from Elliott, a nineteenthcentury engineer of the marsh),' which 'innings' appear on most present-day maps claiming to illustrate marsh reclamation. The various and controversial dates which have been given for both the construction and the silting of the Rhee Wall - in reality built as a canal - are also considered. THE GEOLOGICAL BACKGROUND For the past few thousand years a system of shifting shingle barrier beaches has lain across the Romney embayment, that is, between Fairlight and Hythe.' Behind this sediments accumulated in varying environments, including marine and fresh-water, keeping roughly pace with sea level as it rose after the lee Age. Of particular significance both to the later development of the micro-topography of the marsh landscape and to land-drainage problems is a widespread peat horizon, which appears to have been forming about 1,000 B.C. (although our evidence here is very limited). At different times in the past two thousand years there have been three major gaps in the shingle barrier - near Hythe, New Romney and Rye (Fig. 1). Each gap led to the development of an inlet behind the line of the shingle barrier and became a focus for human activity. Each inlet in turn eventually silted up, causing the eclipse of the respective ports. The Rother waters may have made their way to the open sea by each inlet in turn, but the history of the various courses of the river has been the subject of a great deal of conjecture and is by no means established yet. The Hythe inlet was the earliest and the longest-lived of the three. R.D. Green of the Soil Survey" outlined two distinct series of sediments which filled it: an older, now decalcified, group followed after a pause by a younger calciferous series.' The older group (H, on 'J.A. Williamson, The English Channel (1959), 45-6. ' James Elliott reported in T. Lewin, The Invasion of Britain by Julius Caesar with Replies to the Astronomer-Royal and of the late Camden Professor of Ancient History at Oxford (1862), cxi and cxiv, map 'shewing what lands had been inned previous to the 14th century'. s Jill Eddison, 'The Evolution of the Barrier Beaches between Fairlight and Hythe', Geographical Journal, 149 (1983), 39-53. 6 R.D. Green, op. cit. in note 2, 20. 7 The two groups are distinguished by their calcium carbonate content. Because the calcium carbonate in new sediments is gradually leached out, the calcium carbonate 49 J. EDDISON Fig. 1) consists of an intricate dendritic system of channels (now standing out as creek-ridges), the channels being mainly sandy, the pool areas between them more clayey. This pattern is assumed to represent a salt-marsh environment, which directly succeeded the peat. It was followed by laminated clays which represent a final, quiet lagoonal phase of sedimentation in an area centred on Newchurch (H, on Fig. 1). Most of the trunk creek-ridges appear to be orientated towards the gap in the shingle south-west of Hythe, though one or two may have debouched at New Romney. The later lagoon also seems to have had an outlet at Hythe - but not, apparently, at New Romney. Unfortunately, there is very little detailed evidence upon which to date the development and subsequent silting of the Hythe inlet. One radio-carbon date" shows that peat was still forming in the marsh below Warehorne around 1070 B.C. Green found archaeological evidence that 'the infilling of the marsh creeks, and perhaps their formation, had not occurred in early Roman times' .9 A single radio-carbon date obtained on shells (not growing in situ) in a minor creek near Wheelsgate, Old Romney, shows that that creek was not fully silted up by about A.D. 400. "' It is clear that the mouth of this inlet must have been open around A.D. 300, since the Saxon Shore fort at Stuttfall below Lympne Castle provided a base for a sea-borne force from c. 275 to c. 350. 11 Further, the grant of a saltern at Sandtun, identified by Ward12 as Sampton in West Hythe (N.G.R. TR 122238) (BSC 148) indicates that this was still a salt-water inlet in 732. Salterns are also mentioned at Eastbridge in the Domesday Book. Further inland, however, the Newchurch Hundred had been established with its meeting place at Newchurch itself by mid-tenth century, which indicates that the centre of the lagoon had dried out by that date at the latest. The lack of reclamation embankments is evidence that this half of the marshes silted up and dried out by mainly natural processes. B y the end of the Saxon period settlement was widespread - see content of the surface layers diminishes with time. There is, unfortunately, no easy guide line as to the length of time necessary for the decalcification to be complete. x 3020 ± 94 B.P. W.J. Callow et al., Radiocarbon, 6 (1964), 26. • R.D. Green, op. cit., in note 2, 27. "'1550 ± 120 B.P. W.J. Callow et al., op. cit. in note 8, 26. 11 B. Cunliffe, 'The Roman Fort at Lympne, Kent, 1976--8', Britannia, xi (1980), 227-88. Cunliffe also mapped the sites of Roman finds in relation to the creeks in 'The Evolution of Romney Marsh: A Preliminary Statement' in Archaeology and Coastal Change, (Ed.) F.H. Thompson, Society of Antiquaries (1980), 37-55. 12 G. Ward, 'Sand Tunes Boe', Arch. Cant., xliii (1931), 39-47. 50 THE RECLAMATION OF ROMNEY MARSH Appendix for lists of churches mentioned in Domesday Book and Domesday Monachorum. The Rhee and a parallel sandy channel, both related to the New Romney gap in the shingle, cut across and are therefore later than the creek-ridges of the Hythe inlet. The New Romney embayment east of llesbrigge was reclaimed progressively by a combination of natural processes and artificial walls from about 1300 onwards.'' The marine inundation of the Rye/Winchelsea inlet was retained south of a cordon of sea walls which extend from Appledore Station. then skirt south of Fairfield church, the Woolpack at Brookland and Little Cheyne Court before turning north-east to Midley House and then joining the shingle south-west of Lydd (Fig. 1). The Rye/ Winchelsea sediments truncate and abut the patterns of not only the sediments of the Hythe inlet but also the later man-made drainage scheme (see below, page 00). The inlet which gave rise to these sediments is therefore considerably later than the Hythe inlet. Old Winchelsea, presumed to have been sited on a shingle bank across the mouth of this inlet, was progressively lost during the thirteenth century.'" This loss, coupled with repeated Commissions of Wallis et Fossatis in this area between 1303 and 1368,'·' suggests that the thirteenth and fourteenth centuries were the time of maximum extent of this inlet. After that what was to become Walland Marsh was re-claimed in a sequence of man-made innings. THE DYMCHURCH WALL From Dugdale's conclusion (based, he said, on circumstantial '·' Cal. Pat. Rolls Ed III vol I. 1327-1330, 14, February 23, 1327. 'Licence w prior and convent of Bilsington to drain, enclose, bring inlo cultivation and hold in mortmain a salt marsh in Lyde'. (This salt marsh was in the Belgar area). "The Calendar of the Patent Rolls refers to the necessity for sea-defence measures at Winchelsea on: 8 June, 1244, lO June, 1245, 31 March, 1249, 26 April. 1250. 8 March, 1251, 4 April, 1252, 12 July, 1262, 20 July, 1269 and in 1276. This led up to the Commission on 11 November, 1280 'to buy or obtain certain lands suitable for the new town of Winchelsea, which is to be built upon a hill called Yhamme. the old town being for the most part submerged by the sea·. Matthew Paris, Chronica Majora, v. 176. recounts great destruction at Old Winchelsea by a tidal surge in October 1250, and further damage (p. 272) on 13 January. I 252. "Cal. Pat. Rolls, 1303, 1309, 1316, 1329, 1354, 1360, 1363 (X 2). 1368. M. Teichman Derville, The Level and Liberty of Romney Marsh, (1936), 13. quotes the text of the Charter of Romney Marsh, p. 47, which refers to the lack of provision of adequate sea-defences in the Marsh districts beyond the water course from Snargate to Romney, and thus differentiates these districts from Romney Marsh proper. A better system of protection for those districts, based on that already functioning in Romney Marsh proper, was set up early in the fourteenth century. 51 J. EDDISON evidence)'" that the Romans actively reclaimed Romney Marsh from the sea, it has come to be assumed that they built the original Dymchurch Wall. However, both the pattern of the Romney Marsh sediments revealed by Green" and the geomorphological evidence of the shingle'R indicate that in former times a shingle bank lay parallel with and somewhat outside the line of the present Dymchurch Wall. The sites of Roman finds at Dymchurch - a coin hoard, a cremation site and pottery possibly connected with salt works - would have been protected by this rampart of shingle. There is no indication that this natural sea defence began to give way until the later Middle Ages. The Patent Rolls, which feature sea-defence problems in Walland Marsh and up the Rother valley almost annually from 1289 onwards, make no specific mention of the Dymchurch coast at all until in 1375 there was a Commission de Wallis et Fossatis 'from Hethe to Romenal, thence to Promhill (Broomhill, east of Camber), thence to Apuldre ... '. This was swiftly followed in 1378 by another Commission for 'the marsh of Romeneye, viz. from the town of Hythe along the sea coast to the town of Appledre'. Even these Commissions may simply be outlining an administrative area of the marsh, and it is surely significant that Dymchurch itself, which would have been immediately in the front line had the shingle been breached, is not mentioned in this connection in the Patent Rolls up to 1485. However, some of the shingle of this useful defensive bank was moving away north towards Hythe and a smaller quantity was moving south to New Romney and the bank was becoming inexorably weaker. The critical point seems to have been reached in the sixteenth century. A very early map of this coast'9 shows that by c. 1592 the Dymchurch Wall was well-established. The inscription reads: This Territory or Levell of Romney Marsh is deffended from the inundation of the Salt Sea by a Wall ...c ommonly known by the name of Dymchurch wall beginning at Everdens groyne, and continuing by sundry and diverse denominations unto the '0 W. Dugdale, The History of Imbanking and Drayning of divers Fens and Marshes, ( 1662), 16. " R.D. Green, op. cit. in note 2. "Jill Eddison, op. cit. in note 5. 1• 'The plat of Romney Marsh, describing as well the common watercourses with their heads, arms, pinocks, bridges and principaU gutts, also the high waies and lanes within the same: shewing likewise thee true places of the parish churches, dwellinghouses and cottages within the saide levell'. Drawn temp. Elizabeth. B .M.C ott.A ug. I i 24. Copy S/Rm PS in K.A.O .. The original is said to be in All Souls, dated 1592. 52 THE RECLAMATION OF ROMNEY MARSH end of High knock Wall, containing in the whole 1060 Rodds in Length: which wall is armed and fenced against the wash and rage of the Seas with certain courses of needle work and groynes or knocks continually maintained at the Cost and Charge of the whole Levell.' What had started as piecemeal patches with various names had become a sophisticated structure intended both to keep out the sea and control the Iongshore movement of the remaining shingle. This map shows three sluices through the wall - Willop, Marshland and Clobsden gutts, with just a few groynes on the Willop Wall. Matthew Poker (1617)2<' shows nineteen groynes, some bifurcating at their lower end. After many years of problems, the Dymchurch Wall was completely reconstructed by James Elliott between 1840 and 1847, and raised a further metre in 1971. THE RHEE WALL The Rhee Wall consists of two parallel walls some 50 m. apart east of Snargate, widening irregularly to about 100 m. apart to the west of Snargate. The tract of land between the walls is raised well above the level of the adjacent marshland. It now supports the road from Appledore to Ilesbrigge and continues to within half a mile of New Romney. It has a complex history and is likely to be found to date from several different periods. Dugdale2 ' implied that the Rhee was a Roman sea-bank built to exclude the sea (on the south) from Romney Marsh proper (on the north). This idea was first questioned by Ward. 22 It has recently been suggested23 that it was constructed as late as mid-thirteenth century, the date of the first known written evidence (see below). The truth must lie somewhere between the two. The Rhee was not built as a sea-bank. It cuts straight across the complex of creeks of the Hythe inlet, whose deposits (except near Appledore) are continuous and unchanged from one side of the Rhee to the other. If it had been a sea-bank some difference in the sediments from one side to the other would be obvious. Also, if it post-dates the complete silting-up of the creeks it can hardly be 20 Matthew Poker, 'The Description of Romney Marsh, Walland Marsh, Denge Marsh and Guildeford Marsh .. .' (1617), Maidstone Museum. 21 W. Dugdale, op. cit. in note 16, 16. 22 G. Ward, in discussion, 'Past sea levels at Dungeness', W.V. Lewis and W.G. V. Balchin, Georgr. Joum., 96 (1940), 281-2. 23 N.P. Brooks, 'Romney Marsh in the Early Middle Ages', in The Evolwion of Marshland Landscapes, Oxford University Dept. for External Studies (1981 ), 78. 53 J. EDDISON Roman - although, as already shown, that silting cannot be dated precisely on present evidence. Only between Appledore and a point some 500 m. east of Appledore Station does the Rhee now mark the division between younger sediments on the south and much older deposits on the north. This dates from a relatively late stage when, centuries after silting had ceased in Romney Marsh proper, this short section of the Rhee did indeed act as a sea-bank, keeping the waters of the Rye/Winchelsea estuary out of the Appledore Dowels. The Rhee was in fact built as a canal, and the written evidence is found in the Patent Roll of 28th June 1258, which reads: · As the king has understood that the port of Rumenal is perishing, to the detriment of the town of Rumenal, unless the course of the river of Newenden, upon which the said port was founded, and which has been diverted by an inundation of the sea, be brought back to the said port. and now hears by inquisition that the river cannot be brought back or the port saved unless the obstructions in the old course be removed, and a new course made through the lands of certain men of those parts ... so that a sluice be made below Appledore to receive the salt water entering the river by inundation of the sea from the parts of Winchelsea, and retain it in the ebb {1f the sea, that such water with the water of the river may come together by the ancient course to the new course, and so by that course fall directly into the said port; and that a second sluice be made at Sneregate and a third by the port where the said water can fall into the sea to retain merely the water of the seas inundation on that side that it enter not into the said course; ...· The sluices can only have been on the canalized channel, which had therefore been built some time before 1258. The king's letter makes it clear that the new course, which had been already plotted out, lay in the vicinity of Romney - that is, at the seaward end of the Rhee. Thus the obstructions blocking the river channel probably consisted of influxes of marine sediments, cast up in the repeated storms of that period. It is also important that it was salt water which was entering the river by inundation from the parts of Winchelsea. In other words, it is certain that by 1258 at the latest the tidal flow i n the Rye/Winchelsea inlet was running up at least as far as Appledore. Various later 'trenches' are mentioned in the Patent Rolls. It has yet to be proved where, and indeed whether, they were constructed. When, by whom and for what purpose the Rhee canal was originally constructed is a subject for further research. 24 It has been 24 An early enquiry into the Rhee may be found in W.A. Scott-Robertson, 'The Cinque Port Liberty of Romney', Arch. Cant., xiii (1880), 261-79. Although several of his interpretations must now be ruled out in the light of modern knowledge, the article sets out the fruits of wide-ranging and thorough research. His map of the boundaries of the Liberty showing the link between Romney and Appledore is particularly significant. 54 -t P:AIRflELO , , ..... - __ , , / , BONIFACE INNINGS Mil•• + BROOKLANO BALDWIN$ INNINGS + 8REHZETT ST. THOMAS' INNINGS , , Evidence from the Soil Sux-vey 111ap ,/"--- Recl""1at1on vallo / 􀀙 - 􀀚 Possibl6 walls no evid􀀎nce ot valle ..,-- - - Ml OLEY + Fig. 2. Elliott's (1862) supposed Archbishops Innings. J. EDDISON suggested that it replaced a natural river course - the sandy channel running from Snargate to New Romney. This, however. seems unlikely because the Snargate end of that channel is extraordinarily narrow for a marshland river. Theoretically, such a river would have had a wide, salt-marsh fringed estuary with one or more navigable channels within it, as was the case when the Rother flowed from Appledore to Rye around 1600. If we are indeed looking for a river course which the Rhee replaced, we probably have to look elsewhere. The canal could have been a transport route linking the various monastic and archiepiscopal manors of the marsh, or the Cinque Port of New Romney with its Wealden hinterland, though even so it is remarkably wide for this purpose. Eventually, it was used to channel water to scour out and keep open the New Romney harbour. But was this its primary purpose? THE ARCHBISHOPS' INNINGS The so-called Archbishops' Innings subtend from the south side of the Rhee in a large block between Brookland and Old Romney (Fig. 2). Elliott/-' using his knowledge of local names, suggested that these blocks of land had been inned from the sea by St. Thomas (c. 1162), Baldwin (c. 1184), Peckham (c. 1229) and Boniface (c. 1250). It is, however. clear on the ground that none of these 'innings' is surrounded by a reclamation wall. They are simply abutted on the south by the inner walls of the Rye/Winchelsea estuary, which were constructed in defence of settled land rather than as walls of new innings from the salt-marshes. This is an important distinction. The sedimentary evidence shows that these parcels of land include some of the upper courses of the creeks of the Hythe inlet, and were in fact one of the earliest areas of the marsh to have dried out, some hundreds of years before the archbishops' time. What, then, is the connection with the archbishops? The area in question is distinguished by an orderly rectangular pattern of ditches (now followed by the Brookland roadways). This represents a very large-scale re-organization of previously winding natural watercourses, which could only have been undertaken by a major and influential land-owner. The finger points at the Church, the major land-owner of the marsh. The re-organization presumably dates from the conversion of the original pasture land to arable, and may possibly be attributable to these archbishops. 1􀀂 J. Elliott. op. cit. in note 4, cxi. 56 THE RECLAMATION OF ROMNEY MARSH There is, however, an additional uncertainty. Brooks􀂁" relates Baldwin's Wall and Baldwin's Sewer not to the archbishop of that name but to an earlier charter of c. 1150 by which Prior Wibert of Christ Church, Canterbury, gave Baldwin Scudaway 'as much land about Misleham as he could inne at his own expense against the sea'. It is also important to note that the characteristic rectangular drainage pattern of these 'innings' can be seen in ghost form on aerial photographs continuing on the south side of the inner walls of Walland Marsh, which shows that the area of organized drainage originally extended further south, but was land subsequently lost in• the marine inundation of the Rye/Winchelsea inlet in the thirteenth and fourteenth centuries. CONCLUSIONS I. The Romans did not build a precursor of the Dymchurch Wall. At that time there was no need for a wall because that coast was protected by a natural shingle embankment. It seems likely that the shingle barrier held firm until the Tudor century. 2. The Rhee Wall was not originally built as a reclamation wall, though late in its career it did become a line of sea defence between Appledore and Appledore Station. It was a canal, whose origin and purpose need further investigation. The canal is quite unlikely to have been a Roman construction, since it cuts across creeks apparently still open in late Roman times. Nor is there any place-name evidence to suggest a Roman origin. It had suffered critical obstructions by 1258. The absence of any known written evidence on the construction of such a major feature, however, inevitably suggests an early origin, possibly pre-Conquest. 3. In the absence of surviving reclamation embankments it seems that the Hythe inlet (which includes all of Romney Marsh proper and the area of the so-called Archbishops' Innings) dried out mainly by natural processes and was never subjected to large-scale embanking schemes of reclamation. At the end of the Roman era much of it was still a marine inlet: by mid-tenth century settlement was widespread. 4. In view of the continuity of deposits immediately north and south of the Rhee Wall, and the lack of geological distinction between the two sides, the early history of the Brookland area cannot be dissociated from that of Romney Marsh proper. 5. The Archbishops' Innings are not innings of initial reclamation. 26 N.P. Brooks, op. cit. in note 23, 90. 57 J. EDDISON They cover one of the earliest parts of the marsh to have dried out - centuries before the time of the archbishops whose names they bear. The drainage ditches of this area had already been re-organized some time before the inner sea-defence walls of Walland Marsh were built, and some of the re-organized land was subsequently lost to the medieval marine inundation of the Rye/Winchelsea inlet. 6. Any future work on the reclamation and settlement of the Romney Marsh area must aim to include the evidence available from all possible sources - archaeological, historical and scientific. The conclusions drawn must be compatible with the reliable evidence from all the relevant disciplines. My thanks are due to Miss Eleanor Vollans of Bedford College, London University, for discussion on the Archbishops' Innings, and to the Kent Archaeological Society for a grant towards the expenses of my research. APPENDIX27 Churches mentioned in Domesday Book (1086): Brenzett Newchurch East bridge Burmarsh Blackmanstone Romney Midley Lydd Churches in Domesday Monachorum (c. 1100) but not in Domesday Book: Ivychurch Orgarswick Dymchurch 27 J.K. Wallenberg, Kentish Place Names, Uppsala (1931) and The Place Names of Kent, Uppsala (1934). 58