Palaeogeography of marine inlets in the Romney Marsh area

Introduction This paper reviews the evidence bearing on the evolution of the Rother estuary, and its relationship to the marine inlets which have provided anchorages in the area during and since the Romano-British period. First the settlement of the Old Marshland (Green 1968, Fig. 10) is briefly considered, then the origin of each of the inlets, and port development and reclamation within them, are placed in a chronological framework. Aspects of the growth of the shingle system are also considered. The aim is to trace successive courses of the Rother and to show that the changes that have taken place reflect the development of a geomorphological system in which estuarine processes have interacted both progressively and episodically with marine processes forming and degrading the coastal barrier. The river name Rother is used throughout this account to refer to the main discharge of river water entering the Romney lowland past the Isle of Oxney, either to the north of it, or to the south, or both. The Old Marshland The Old Marshland is divided here into four sub- regions. The Dymchurch Shore Archaeological findings (Dunkin 1844; Issacson 1846; Elliott 1847,1862) during the reconstruction of the Dymchurch Wall, and archaeological remains at Jesson Farm near St Mary's Bay (TR 082276) (Green 1968,27) show that the Old Marshland close to the modern coastline was occupied during, and possibly before, the Romano-British period. Saxon settlement immediately inland from Dymchurch is indicated in the document- ation of 8th- and 9th-century land ownership - at Sellinge Farm in AD 700 (Ward 1936) and at Burmarsh in AD c. 850 (Ward 1933b), while Dymchurch itself is recorded in Domesday Monachorum. It has generally been assumed, mainly for lack of documentary evidence to the contrary, that the Dymchurch shore from Roman times until the 14th or 15th century was protected from the sea by a natural barrier, presumed to be of shingle and destroyed by the action of the sea in the late medieval period (Eddison 1983a, 47). There is however good documentary evidence (Smith 1943) of repairs to sea defences at Dymchurch at least as early as the second half of the 13th century, which suggests that by that date coastal configuration there may not have been greatly different from that of the present. The Lydd Shingle Jones (1953) and Philp and Willson (1984) report evidence of Roman occupation at sites on shingle banks immediately to the west of Lydd, which indicates the existence of the coastal shingle barrier here in Romano- British times. More recently the discovery of a Bronze Age axe hoard (Needham 1988) in the shingle north of Lydd (TR 0472 19) suggests an even earlier date for some of the Lydd shingle. The shingle in the Broomhill area, which lies considerably lower and is overlain by a later generation of fine sediments, is presumed to be earlier still. Tooley and Switsur (1988) have obtained a radiocarbon date of 3410 f 60 BP from limus immediately overlying the flint cobbles. Saxon occupation of the Lydd area is first indicated in a charter of AD 774 (Ward 1931a; Brooks 1988). Romney Marsh Archaeological remains at Ruckinge (Bradshaw 1970) show that inland areas in the north-western part of the Old Marshland were probably occupied in the Iron Age, and certainly in the Romano-British period. Saxon occupation is indicated in the southern part of Ruckinge in AD 724 and AD 805 (Ward 1933a). Walland Marsh In the south-western part of the Old Marshland, evidence of Romano-British occupation has been found at Snargate (Kelly 1968), and there is documentation of Saxon occupation at Misleham at some date between AD 833 and AD 858 (Brooks 1988) and possibly at Agney in AD 785 (Brooks 1988). Even the brief review offered here of the early settlement history of the inland parts of the Old Marshland indicates the strong case that can be made to dispel any lingering belief that the Rhee Wall separates 168 Christopher Green landscapes with fundamentally different histories of early reclamation and settlement. The Marine Inlets The histories of the three marine inlets are considered in turn, and in relation to the evolution of the Rother estuary. The Hythe Inlet The Marine Inlet The Roman naval anchorage, Portus Lemanis, appears to have occupied an inlet which may have extended beyond Lympne to the west and beyond West Hythe to the east. It had a history that certainly included the 2nd century AD. The tract of New Marshland delineated here by Green (1968, Fig. 10), and extending south- westward towards Old Romney may indicate the former extent of marine influence in this area. In this situation, the Roman anchorage must have been sheltered by an off-shore coastal barrier. This is generally thought to have been the distal end of a shingle spit extending across the Romney embayment from Cliff End near Fairlight. At a later date a Saxon Shore Fort was constructed. Its remains are now visible at Stutfall Castle and its occupation spanned at least the interval AD 275 to AD 359 (Cunliffe 1980). Although material from the earlier naval station is found incorporated into the fabric of the Saxon Shore Fort, the latter appears to have been constructed on previously unoccupied ground, and the exact location of the on-shore facilities of Portus Lemanis remain unknown. Specific archaeological evidence that the Saxon Shore Fort was associated with a naval anchorage is lacking. However, it is likely (Cunliffe 1980) that the position of the fort was chosen to protect and control sea-borne trade, so a 3rd- or 4th-century naval and commercial anchorage seems likely. Thus the existence of a substantial marine inlet can at present be confirmed for the 2nd century and is probable at least into the 4th century. By the mid-10th century, the central part of the tract of New Marshland was certainly occupied at Newchurch (Green 1968, Fig. 9). However the persistence of marine influence associated with an inlet here is suggested by the record of salt works at Eastbridge in Domesday, although a major marine inlet is not necessarily indicated at that time. The Rother Estuary It is possible that the Hythe inlet was also the estuary of the River Rother. Several authors (e.g. Holloway 1849) have argued that the river originally flowed almost due east from the neighbourhood of Appledore, close to the northern edge of the Marsh, and Green (1968, 36) considers that the balance of historical and field evidence supports this view. Such a course would have brought the Rother into the Hythe inlet on its northern or north- western side. However, neither Elliott (1862, 1874) nor Topley ( 1875) could find any evidence of a river course across the northern part of Romney Marsh, and Elliott proposed a generally southward course from Appledore, turning eastward past Midley to New Romney. Elliott showed this course entering the sea via the Romney estuary. A similar course was suggested by Lewis (1932). Green (1968) regards this course as unlikely, perhaps chiefly because it is now occupied between Agney and Old Romney by decalcified soils which he evidently believes are considerably older than any of the identifiable riparian or estuarine tracts. He does not however present any conclusive evidence and recognises that a thick sandy substratum is present along the line of such a course in the vicinity of the Wainway channel and around Midley. In the present account, the case for an early estuary of the Rother following the course originally proposed by Elliott is further developed. It is suggested that prior to the establishment of the marine inlet at Romney, the main branch of the Rother flowed north between Old Romney and St Mary-in-the-Marsh to enter the Hythe inlet from the south. Along this alignment a sandy substratum is again present (Green 1968). This course is shown in Fig. 14.1. The grounds for and the implications of this proposal are more fully developed in subsequent paragraphs. The Romnty Inlet The Marine Inlet A charter of AD 741 (Brooks 1988) indicates the presence of a settlement on the site of New Romney, and suggests that the Rother (termed Limen in the charter) entered the sea close to this point. The use of the name Limen here for the river at New Romney suggests that this name was used for more than one river in the Romney lowland, as it is found on several occasions in the 8th and 9th centuries (Ward 1931b, 1933a, 193315) attached to water-courses in the vicinity of Warehorne, Kits Bridge, Ruckinge and West Hythe on the northern edge of the Marsh. Various aspects of the development of the Romney ports can be documented during the pre-conquest and medieval periods, with New Romney superseding Old Romney as the principal settlement and the site of the harbour facilities by the mid-12th century (Tatton- Brown 1988). The date at which the coastal barrier was breached at New Romney is completely unknown (except that the charter ofAD 741 appears to provide a terminus ante quem), and the course by which the Rother entered the marine inlet created by the breach has been a source of controversy. A broad lower estuary flanked by salt marsh, similar to the modern estuary at Rye prior to reclamation, might be expected. In Green's interpretation (1968) of the soils and sediments, the space for such an estuary cannot be readily identified. The scope for re-examination of Green's conclusions is outlined below. of Marine Inlets The Lower Rother It is suggested in the present account that the cause of the breach at New Romney was depletion of the shingle supply to the barrier itself, near New Romney, at some time prior to AD 741, possibly associated with encroachment of the estuarine Rother on the landward side of the barrier. Flux in the movement of shingle is amply reflected in the alternations of shingle and alluvium in the Dungeness foreland. Investigation of this flux has yet to be undertaken. At the time of the breach, the course of the Rother upstream from Romney to the vicinity of Appledore (Fig. 14.2) would have remained essentially as it had been in the Romano-British period. In this direction the space for a substantial tidal river and associated salt marsh can be found in the area now occupied by the Walland and Guldeford Marshes, and the changes required in Green's interpretation of the marsh soils and sediments, mainly between Agney and Old Romney, are relatively minor. The existence of a substantial channel in this particular area, between the Romney and Rye estuaries, giving access from the sea to the vicinity of Lydd is indicated in various documents from the 8th to the 16th centuries. At first in a charter ofAD 774 (Ward 193la), the sea may be indicated to the east and north of Lydd, in the direction of the Romney inlet; but finally (Gardiner 1988), vessels appear to have been approaching Lydd from the west. Early maps, of late 16th- and early 17th-century date (Symondson 1594, 1596; Poker 161 7), show the Wainway Channel as a broad inlet here, open to the sea in the direction of Rye. The only alternative course for the Rother between Appledore and New Romney that has been canvassed is the narrow meandering channel lying parallel to and north of the Rhee Wall and extending from Snargate to New Romney. As Eddison (1983b) points out, this forms a very narrow passage through the sediments of the Old Marshland, apparently much too narrow ever to have accommodated the estuarine Rother. 169 substratum. Further investigation may therefore reveal scope for revision of the boundaries between 'older' and 'newer' land surfaces in the marsh. The Diversion of the Rother There is some uncertainty regarding the exact date at which the Rother ceased to flow into the Romney inlet. Camden (1 789) refers to a catastrophic avulsion of the river as a consequence of a storm in 1287 which "removed the Rother ... out of its channel; stopping up its mouth and opening for it a nearer passage to the sea by Rhie". On the other hand, the frequently quoted Calendar of Patent Rolls for 1258 indicates "that the port of Rumenal is perishing, to the detriment of the town of Rumenal, unless the course of the river of Newenden (understood to be the Rother), 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". In either case, a major change of the estuarine course of the Rother seems to he indicated in the mid- to late 13th century. A contributary cause may again have been the encroachment of the estuarine Rother on the landward side of the coastal barrier, in this case near Rye, at a time when historical records indicate exceptional storminess in the English Channel, which is likely to have weakened the barrier itself. Medieval accounts of three of the great 13th-century storms refer specifically to events in the vicinity of Romney Marsh. October 1250 " ... when the moon was in its prime, on the first day of the month there appeared a tumid, ruddy new moon, a sign of storms to come ... In the first weeks of the moon's growth the air began daily to be violently unsettled ... And (what was more damaging) the sea was disturbed, and crossed its usual limits, twice flowing without an ebb. ... at one port alone, that is, Decalc$cation and soil development The conclusions reached in the previous section are incompatible with some of the interpretations of marshland sediments and soils proposed by Green ( 1968). It is implicit in Green's exposition that the areas mapped by him as decalcified marsh were already reclaimed, either naturally or artificially, when Anglo- Saxon settlement commenced and that areas mapped as calcareous marshland may have been reclaimed as long ago as, but not before, the period of Anglo-Saxon settlement. Thus decalcification, if correctly interpreted by Green, is a characteristic by which pre-Anglo-Saxon land surfaces can be separated from land surfaces originating in or after the Anglo-Saxon period. This view may prove to be an oversimplification. Rates of decalcification depend on a large number of variables, including for example substratum permeability. Thus sandy substrata promote and accelerate decalcification, and the resulting decalcified soils may be considerably younger than decalcified soils developed on a clay Hertburn, apart from small and medium sized boats, three noble ships were swallowed by swelling waves. At Winchelsea, a certain eastern port, besides salt- cotes and the retreats of fishermen and bridges and mills, more than three hundred buildings were destroyed in that same district by the violent surge of the sea, as well as several churches. Holland in England, and Holland in the parts beyond the sea, suffered irreparable losses. The rivers issuing into the sea, were so thrust back that they swelled, and destroyed meadows, bridges and crops which the barns had not received, and which were standing in the fields." Matthew Paris, Chronica Majora, ed. H. R. Luard (1880) v, 175-6. 13th January 1252 "...during the day and the night which followed, a raging east wind and an angry south-westerly wind occasioned much damage and ... drove back the waves of the sea from the shores, and unroofed or destroyed houses. The largest oaks were torn up by the roots ... : it tore lead off the churches, sunk in the 170 Christopher Green deep the largest and stoutest ships, and inflicted great and terrible damage ... If we are silent about other injuries and damage we will introduce an account of what we experienced. At the harbour of Winchelsea, a very necessary one to England and especially to London, the waves of the sea, as if indignant and furious at being driven back on the day before, covered places adjacent to its shores, and washed away and drowned many men." Matthew Paris, Chronica Majora, ed. H. R. Luard (1880) v, 272. 4th February 1288 'On the day before the nones of February the sea flooded so greatly in Thanet and that neighbour- hood, and in the marsh of Romenal and all the adjacent places, that all the walls were broken down and almost all the lands covered from the great wall of Appledore towards the south and west as far as Winchelsea." Gervase of Canterbury, Opera, ed. W. Stubbs (1880) ii, 293. The Rhee Wall The original function and the date, or dates, of construction of the Rhee Wall are problems closely related to an understanding of the course of the Rother towards the Romney inlet. Following the opinion of Wallenberg ( 1934) and Ward ( 1940), amply confirmed by Green (1968), there has been general agreement that the Rhee Wall was never a wall related, except incidentally, to land reclamation or sea defence, but an artificial waterway feeding water from the Rother, together with tidewater, into the marine inlet at Romney. The Patent Roll of 1258 can only be understood to indicate that an artificial waterway between Appledore and Old Romney was already in existence at that date. But for how long it had been in existence is uncertain. It seems possible that it had not long been completed (Tatton-Brown 1988). This interpretation is to some extent supported by the fact that this is the earliest known reference to an artificial channel between Appledore and Romney. The history of the Liberty of Romney, with a boundary intimately related to the Rhee system, can also be traced to an origin apparently in the 13th century (Teichman Derville 1936). Moreover, after 1258 there are many documents referring to maintenance of sluices and artificial channels between Appledore and Romney. On the other hand, it is difficult to interpret the 'old course' referred to in the 1258 Patent Roll, except as a pre- existing portion of the Rhee. The word 'old' can perhaps be taken to imply the passage of two or three generations since the date of construction. Tatton-Brown (1988) argues that the Rhee between Appledore and Old Romney is unlikely to have been constructed before about 1190. While it seems clear that the Rhee in its present configuration between Old Romney and New Romney was constructed after 1258, two alternative explanations of the age of the Rhee between Appledore and Old Romney seem possible. a. Work on the Rhee began before the middle of the 13th century in response to deterioration of the natural estuarine channel of the Rother between Appledore and New Romney. Such deterioration may have been a consequence of silting, perhaps associated with land-use change, and may not at first have involved diversion of the Rother into the Rye inlet. The balance of evidence appears to the present author to favour this interpretation, with construc- tion of the Rhee at sometime in the late 12th or early 13th century b. The whole Rhee was built after a 13th-century diversion of the Rother - an alternative considered by Tatton-Brown (1988). Whichever explanation is correct, it seems clear firstly, that the Rhee system is not aligned on or even close to a former course of the Rother, and secondly that from the mid-13th century, whether or not that was the date of construction, the Rhee system presented fairly severe problems of maintenance (Vollans 1988). In the 14th century the struggle to keep water flowing into the Romney inlet from the landward side was being lost. A document of 1337 indicates for example that an artificial channel near Appledore had already been blocked and disused for thirty years (Cal. Pat. Rolls 1334-38, 457). However, as late as 1487-8 the town of Romney was still investigating possible sources of water in Walland and Guldeford Marshes to provide a flow from landward into the harbour. The Rye Inlet (Fig. 14.3) The Marine Inlet prior to the 13th Century The histories ofwinchelsea and Rye as ports prior to the 13th century (Cooper 1850; Holloway 1847), and the Domesday record of numerous salt works in the area of Rye, indicate the proximity of these towns to a sheltered inlet from at least the 11 th century. If the Rother did not enter the sea at Rye until the 12th or 13th century, the nature of this inlet needs to be elucidated. The sites of Old and New Winchelsea, their configuration and links with neighbouring settlements give some clues as to the coastal topography prior to the 13th century. a. Early 13th-century documents (Historical Manu- scripts Commission 77, vol i, 72, 80) refer to a direct road link between Winchelsea and Broomhill. This argues against any broad marine inlet between the two settlements. b. The historical links of Rye and Winchelsea are with Hastings. This association separates them from the Romney-Lydd-Broomhill-Appledore-Small E-Tythe elements of the Cinque Port federation, which can realistically be seen as dependent on the Rother. This historical association seems to preclude the possibility, considered by Lewis (1932), that Rye and Winchelsea originally faced onto a Rother estuary which entered the sea at Ron~ney. In any Palaeogeography of Marine Inlets Fig. 14.1 Reconstruction of the course of the Rather into the Hythe estuary, showing the shingle spit enclosing the estuary on its south-east side. (The dotted line marks the modern coast) case, early records of the approaches to Winchelsea harbour suggest an inlet directly accessible from Rye Bay, e.g. accounts of the taking of Winchelsea by Louis of France in 121 6, his subsequent investment there and relief by a squadron of French warships (Turner 1904). 171 Fig. 14.2 Reconstruction of the course of the Rother into the Romney estuary. The Old Marshland (Green 1968) is indicated, but no attempt is made to represent the marine inlet presumed to exist from at least the 11th century onward in the Rye/ Winchelsea area. c. The establishment of New Winchelsea on the west of the Rother estuary, with a harbour in the Brede valley also suggests that the hinterland, and therefore the main overland links of Old Winchelsea lay to the west, and that the town itself, or at least a major part of it, lay to the west of the inlet. It seems most likely that Winchelsea and Rye stood originally on the estuary of the Kivers Tillingham and Brede. This hypothesis is consistent with the westward orientation of the historical links of the two towns, and their separation from a Rother group within the Confederation of the Cinque Ports. Moreover, without the discharge of the Tillingham and Brede, it is difficult to understand how a marine inlet open to the south towards Rye Bay in the Rye/Winchelsea area could remain free of shingle drifting eastward from Fairlight. On the other hand, if a marine inlet did exist, any reconstruction of the medieval coastline must explain what prevented marine erosion encroaching upon the marshland in what is now the Guldeford and Walland Marsh area (as it subsequently did, very actively, in the 13th and 14th centuries) and capturing the Rother. Perhaps the most plausible site for Old Winchelsea is on a broad coastal barrier, and mainly to the west of a relatively narrow channel through the barrier, with h'arbour facilities facing onto a broader stretch of water, the Camber, to the north of the coastal barrier. Given the prevailing eastward drift ofshingle around Rye Bay, Fig. 14.3 The diversion of the Rather in the 13th century, and the establishment of the Rye estuary. any outlet here through the barrier is likely to have had a broadly eastward or south-eastward orientation, so that an outer spit developing from the west would protect an inner harbour, and harbour entrance, as well as the inner shingle bank and adjoining low-lying alluvium, from direct marine attack. Thirteenth-Century Inundations The dramatic changes of the 13th century can be A.D. 1600. 1500. 1400. 1300- 1200- 1100. loo@ 900. 800. 700 600. 500- 400. 300. 200. 100. 0. OLD MARSH North Walland North East L. ~ 7 0 0 Sellinge West L' A720 Rucklnge r T JessonFm Ruckinge * I A785 A!~'D'Y - Snargate .L SHINGLE BANKS . Lydd ROTHER ESTUARY Dymchurch Hythe Newchurch A741 Lydd A741 New Romney - I- I ---? C 350 Rolnney I __---* Saxon Shore Fort *c W 275 Portus Lemanis .r G d * Dymchurch * A c.100 ANCIENT TOWNS Winchelsea [~hee] Rye r A h latest record = part only Rye approximate per~od of Romano-Br~tish occupation earliest charter evidence of occupation earlier occupation inferred 741 earliest record R. Rother TillinghamlBrede estuary 1 Fig. 14.4 A spatial and temporal framework for the evolution, settlement and reclamation of Rornnty Marsh. Historical events - population decline o Black Death Increased storminess population Increase o Norman Conquest I" L S 0 N Limen recorded Saltworks Old New major er~ods of mehva~ reclamation and land-use change Palaeogeography of Marine Inlets understood as reflecting the gradual erosion of the south side of the coastal barrier near Old Winchelsea, to a point when the sea had direct access to the low ground behind the barrier. This process was no doubt intensified by the increased storminess which characterised the 13th century. Marine erosion of the fine-grained sediments between the TillinghamIBrede estuary and the Rother estuary would quickly open a breach through which the sea could rapidly encroach on the low-lying Rother salt marsh and through which the Rother could enter Rye Bay. It may be significant that the Patent Roll of 1258, in connection with the sluice below Appledore, refers to salt water entering the waterway there by "inundation of the sea from the parts of Winchelsea", possibly suggesting that salt water had only recently reached the vicinity of Appledore from the direction of Winchelsea. Thus for a short while in the second half of the 13th century, Old Winchelsea may have stood at the mouth of the Rother, in a position which appears to have been described by Jeake (1 728) "washed by the British Ocean on the south and east, and the mouth of the Rother on the north". The Rother Estuary after the 13th-Century Inundations If the diversion of the Rother from Romney to Rye is correctly placed in the 13th century, the Rye estuary dates from that time. The documentary record of very active sea defence measures to the south-west of the Rhee Wall and so far north as Appledore in the 13th and 14th centuries suggests strongly that there was a period of rapid adjustment following a major change of coastal configuration. The boundary between the Old Marshland and the New Marshland is the cordon of sea walls created at this time. On aerial photographs, patterns of pre-inundation drainage works on the Old Marshland can be traced across these sea walls into the New Marshland giving some indication of the loss of land that occurred here. Since the 14th century, the history of the Rye inlet has been mainly one of reclamation (Eddison 1988; Gardiner 1988; Tatton- Brown 1988). An important factor in the process of reclamation has been the growth into the estuary mouth from both east and west of a complex of shingle systems (Eddison 1983a). This shingle protects the inner estuary from direct erosional attack by the sea, and probably has done so since as long ago as AD 1700. Prior to this, the low-lying shores of the open estuary - the Appledore Water - would have been very susceptible to marine erosion. References Calendar of Patent Rolls Bradshaw, J. 1970: Ruckinge. Arch. Cant. 85, 79. Brooks, N. P. 1988: Romney Marsh in the Early Middle Ages. In this volume, chapter 8. Camden, W. 1789 edn.: Britannia. (London). Cooper, W. D. 1850: The History of Winchelsea (London, John Russell Smith). Conclusion The specific aim of the present paper has been to trace successive courses of the River Rother. The interpret- ation which is offered and which is summarised in Figs. 14.1-14.3 depends very much on archival records and very little on geomorphological field evidence. From the archival record it may be possible to say where the Rother was at a particular time, but the nature and causes of change remain largely unknown. This illustrates usefully the present state of geomorphological research in Romney Marsh and points to the future potential. Was coastal barrier breaching encouraged by episodic depletion of the shingle supply, and if so, what caused the flux? Did land-use change influence patterns of sedimentation in Romney Marsh? Was proximity of the river to the landward side of the coastal barrier a factor in determining where barrier breaching occurred? These are some of the problems waiting to be investigated. The paper also illustrates the importance of interdisciplinary investigations. Landform development in Romney Marsh during the last 2,000 years has involved major changes in coastal configuration and in the position of the Rother estuary, but these changes are only known because geomorphological elements of the landscape are sometimes documented incidentally in the record of land ownership, or have come to light in archaeological investigations. In the interpretation of this essentially historical evidence, whatever t'he implications for reclamation, settlement or land-use change, the record of events must also be geomor- phologically credible, both in terms of spatial pattern at any one time and in terms of the sequence and timing of successive geomorphological events. The development of a plausible framework for geomorphological events will in turn assist the interpretation of reclamation, settlement and land-use. It is important therefore to place the strands of historical evidence in some sort of spatial context, so that the principal constraints in modelling 'past landscapes can be clearly recognised. Thus, Fig. 14.4 illustrates, albeit selectively, broad patterns of reclamation and settlement across the whole Marsh in relation to land types and to the changing course of the River Rother. At the same time, it indicates the possible inflhence of broader demographic events and trends. In such a juxtaposition of the findings of different disciplines lies the key to the Romney Marsh landscapes. Cunliffe, B. W. 1980: The Roman fort at Lympne, Kent, 1976-8. Britannia 1 1, 227-88. Dunkin, A. J. (editor) 1844: Proceedings of the British Archaeological Association Me~ting held at Canterbury, 1844. (London, John Russell Smith). Eddison. J. 1983a: The evolution of the barrier beaches between Fairlight and Hythe. Geog. Journ. 149, 39-53. 174 Christopher Green Eddison, J. 1983b: The reclamation of Romney Marsh: some aspects reconsidered. Arch. Cant. 99, 47-58. Eddison, J. 1988: 'Drowned Lands': Changes in the course of the Rother and in its estuary, and associated drainage problems. 1635-1735. In this volume, chapter 12. Elliott, J. 1847: Account of the Dymchurch Wall, which forms the sea defences of Romney Marsh. Min. Proc. Inst. Ciu. Eng. 6, 46&-86. Elliott, J. 1862: Maps, and views reported in Lewin, T., 7he inuasion of Britain by julius Caesar. (London). Elliott, J. 1874: Views reported in Furley, R., The History ofthe Wealdof Kent. (Ashford). Gardincr, M. F. 1988: Medieval settlement and society in the Broomhill area, and excavations at Broomhill church. In this volume, chapter 10. Grren, R. D. 1968: Soils of Romney Marsh. Soil Survey of Gt. Britain, Bull. 4 (Harpenden). Historical Manuscripts Commission 1925: Report on the manuscripts of Lord Ue L'Isle and Dudley preserved at Penshurst Place. (Publication No. 77, London, HMSO). Holloway, W. 1847: History and antiquities of the ancient town andport of Rye, in Sussex: compiled from the original documents. (London, John Russell Smith). Holloway, 1849: The History of Romney Marsh. (London, John Russell Smith). Issacson, S. 1846: Discovery ofRoman urns and other ancient remains at Dymchurch in Romney Marsh. Archaeologia 31, 487-8. Jeake, S. 1728: The Charters of the Cinque Ports. (London). Jones, I . 1953: Roman remains on Lydd Rype. Arch. Cant. 66, 160-1. Kelly, D. B. 1968: Snargate. Arch. Cant. 83, 2654. Lcwis, W. V. 1932: The formation of Dungeness Foreland. Geog. Journal. 80, 309-24. Luard, H. R. (editor) 1872-83: Matthew Paris, Chronica Majora. Rolls Series 57, i-vii. (London, Longman). Needham, S. P. 1988: A group ofEarly Bronze Age axes from Lydd. In this volume, chapter 5. Philp, B. and Willson, J. 1984: A Roman site at Scotney Court, Lydd. Kent Arch. Rev. 77, 156-61. Smith, R. A. L. 1943: Canterbury Cathedral Priory. (Cambridge). Stubbs, W. (editor) 1879-80: The historical works of Geruase of Canterbury. Rolls Series 73, i-ii. (London, Longman). Tatton-Brown, T. W. T. 1988: The topography of the Walland Marsh area between the eleventh and thirteenth centuries. In this volume, chapter 9. Teichman Derville, M. 1936: The Leuel and Liberty of Romney Marsh. (Ashford). Tooley, M. J. and Switsur, V. R. 1988: Water level changes and sedimentation during the Flandrian age in the Romney Marsh area. In this volume, chapter 3. Topley, W. 1875: The Geology of the Weald. Mem. geol. Suru. U.K. Turner, G. J. 1904: The Minority of Henry 111. Trans. Roy. Hist. Soc. N.S. 18, 245-95. Vollans, E. C. 1988: New Romney and the 'river of Newenden' in the later Middle Ages. In this volume, chapter I I. Ward, G. 1931a: Saxon Lydd. Arch. Cant. 43, 29-37. Ward, G. 1931b: Sand Tunes Boc. Arch. Cant. 43, 39-47. Ward, G. 1933a: The river Limen at Ruckinge. Arch. Cant. 45,129-32. Ward, G. 1933b: The Saxon charters of Burmarsh. Arch. Cant. 45, 15341. Ward, G. 1936: The Wilmington charter of AD 700. Arch. Cant. 48, 11-28. Ward, G. 1940: In Discussion in Lewis, W. V. and W. G. V. Balchin, Past sea levels at Dungeness. Geog. Journ. 96, 258-85. Wallenberg, J. K. 1934: The place-names of Kent. (Uppsala).