Medieval Tonbridge reviewed in the light of recent excavations at Lyons, East Street

THE DEVELOMENT OF MEDIEVA L TONBRIDGE REVIEWED IN THE LIGHT OF RECENT EXC AVATIONS AT LYONS, E A ST STREET ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM The redevelopment of land at Lyons, East Street, Tonbridge (TQ 5920 4660) led to an archaeological evaluation of the site (Fig. 1). The evaluation trenches revealed features dating from the late eleventh/thirteenth century towards the East Street frontage. An open area excavation ensued covering some 340m2, comprising three trenches on the north, east and west sides of the study site (Fig. 2). The excavation was carried out in June 2001 and January 2002 by Pre-Construct Archaeology Ltd. The site lies south of East Street on a south-facing slope grading down towards the Medway, some 1 00m south of the Parish Church of St Peter and St Paul and 1 00m east of Tonbridge castle. The eastern boundary of the site is formed by a section of the medieval town rampart (Fig. 2). The geology of the site is shown by the British Geological Survey to be Tunbridge Wells Sand. Archaeological background In 1972 and 1976 excavations were carried out to the NNW of the study area in the vicinity of the north-western town wall. The results of these excavations have been reported elsewhere but can be briefly summarised here (Streeten 1976). Three sections were excavated through part of the northern ramparts which showed that much of the standing rampart is a recent deposit which contained a residual fourteenth-century French jetton. Beneath this, the medieval embankment comprised a series of earthen deposits containing sherds of medieval pottery and a residual sherd of samian. There was no sign of revetment for the bank although there was a slight benn between ditch and bank. The construction techniques do not appear to have been uniform as one section revealed a turfline and the others did not, although it may be that part of the original topography had to be levelled up. No evidence was found to suggest either a timber palisade or stone wall on top of the defences. Three small areas were opened to the west of the High Street behind 119 J· ELIOTT WRAGG, CHRJS JARRETT A D JEREMY HASLAM • Tonbridge Bngnton 0 N t 2km -c:::::=--==:::a--=-c:::::=--===i• Fig. l Site Location plan. the defences. A drainage ditch was located parallel to the High Street containing medieval pottery. No further features were recorded predating the eighteenth century. The lack of SMR evidence for prehistoric material locally and the recovery of only three residual struck flints from the Lyons site suggests limited prehistoric activity in the area. Roman finds are confined to one Roman coin from the Medieval Castle mound and one fragment 120 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE West trench - Fig. 2 Plan of Trenches. of pottery. No further evidence of Roman activity in the town has been recorded. The Lyons site, however, produced five residual Roman sherds, thus considerably increasing the assemblage for the town. Historical Background Although the name Tonbridge suggests a pre-conquest origin for the town, there is no documentary or archaeological data to support this thesis. The town of Tonbridge is first recorded in the Anglo-Saxon Charter c. l 087 as Tonebridge and Tonebrieg. It is referred to as Thunnebrigge in 1230. It is situated at a crossing point of the Medway on the London-Hastings road. The town grew up to the east of the castle which is first mentioned in 1088. Henry l1l granted a license to enclose the town with a crenellated wall in 1259, although it is not certain whether the town's defences date to this time or, indeed, whether they were surmounted by a stone wall. Cartographic evidence suggests that although East Street is a medieval 121 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM feature, in the area of the current development it does not appear to have been associated with dwellings. The economy of the town was linked to the iron industry of the Weald, and while the economic and demographic picture is unclear for the medieval period, the population rose by 50 per cent between 1550 and 1640. In the period 1587-1609 a 'dozen houses' were situated on East Street, although the location of these is unknown, unless the Port Reeves House at 25 East Street was among them; East Street was previously known as East or Mill Lane and Swan Lane (Ford 2000). There is no evidence for the re-use of the fortifications in the civil war of 1642-45, possibly because they had become unsuitable. An eighteenth-century map reproduced in Hasted's History and Topography of the County of Kent (1797-1801) shows the castle, church and town to the north of the Medway. The High Street is depicted as having buildings but East Street, whilst shown, is not represented as having any. The Tithe map of 1841-2 shows that a property known as Lyons had been constructed on East Street and a number of other properties towards the western end. Much of the frontage appears unoccupied and only the Port Reeves House is shown to the east. The town defences are shown on the Ordnance Survey map of 1867 as a wide north-south running ditch forming part of the boundary of the study site and extending south towards a ditch or stream which, in turn, joins the Medway. The Ordnance Survey map of 1897 shows that Lyons Crescent had been built to the south of the study site, itself remaining substantially the same. Subsequent maps show development in the environs of the site but not on it. T he Archaeological Sequence at the Lyons site The natural deposits on the site consisted of a moderately compacted light brownish yellow, slightly clayey sand with occasional areas of iron panning and occasional small sub-angular and sub-rounded gravels. These deposits were encountered between 27.79 and 27.06m oo sloping down from north to south following the general topography of the area. Three residual struck flints were recovered; a microlith, a blade and a blade-like flake. The microlith indicates a Mesolithic date, being diagnostic, while the other two pieces are compatible with this period. This small assemblage is indicative of Mesolithic activity in the vicinity. The presence of five Roman pot sherds from residual contexts indicates the possibility of Roman activity on or near to the site. The tail of the town rampart was found, although not fully excavated, in the East Trench, comprising a number of layers of sandy clay and clay sand, but no turf line, to a height of 1.50m above ground level. No evidence either of a timber palisade or a stone wall was found which confirms the generally accepted hypothesis, discussed below, that the 122 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE defences of Tonbridge consisted merely of an earthen rampart and ditch. No artefactual evidence was found precluding precise dating. As discussed below, however, they appear likely to belong to the period 1260-1300 if, as Haslam argues below, they predate the building identified in the excavation. It has been suggested further that the town defences were constructed around the same time as the castle gatehouse in 1265, see below, which is consistent with the inferences drawn from the excavation. Two pits were located in the East Trench within the rampart, one of which contained pieces of undiagnostic metalworking slag, and one in the West Trench similarly included small quantities of slag. These contained sherds of pottery which suggest a thirteenth-century date (see Appendix). They probably represent rubbish disposal in an area of open ground. An east-west running ditch was located in the West Trench. This probably constitutes a field boundary, and the pottery recovered from it indicates a thirteenth/fourteenth-century date. It also held a piece of undiagnostic metalworking slag. In the North Trench four pits contained a number of sherds of pottery ranging in date from at least the eleventh or twelfth century to the thirteenth century. Three were probably for rubbish disposal, two contained evidence of metal (though not iron) working including the base of a possible mould and possible fused sand, while one appears to have been used, at least initially, as a cesspit and included amongst its finds a fallow deer antler, undiagnostic metalworking slag and a fragment of a vitrified hearth. To the west of these pits was a north-south running linear feature. It measured 0.60m wide by more than 3.00m in length, extending into the northern limit of excavation, and had shallow sloping sides and a flat base with a depth of 0. 12m. It was interpreted as a beam slot and probably represents the eastern extreme of a building fronting onto East Street. It is likely that the pits discussed above were linked to this structure. The spatial relationships of the features suggests that the area to the east of the building fronting on East Street was open during this period and was probably used for rubbish disposal as were the areas investigated by the East and West Trenches. This is compatible with the argument put forward below that this area of the town was originally an open space behind the zones of domestic occupation. The pottery assemblage for this phase suggests domestic activity while the slag indicates metalworking in the vicinity, although smelting did not take place on the site (Keys 2002). The lack of features for the fourteenth century indicates that this sector of the town at least had contracted in line with the general decline affecting the surrounding area and England as a whole at this time as a result of climate change, soil-exhaustion, and civic, financial and political instability (Platt 1978, 91-125). 123 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM Various features dated to the sixteenth century suggest that activity was once more taking place after the fourteenth century decline and that metalworking may have recommenced in the vicinity. The North Trench revealed three postholes within 3m of one another in the south of the trench which could indicate some form of structure, two more postholes were uncovered which appear to belong to this period and were probably also part of a structure. Two pits were located in the North Trench which may have had a rubbish disposal function and been associated with the features discussed above. One pit was located in the East Trench which contained a piece of undiagnostic metalworking slag and may have been for rubbish disposal. Six pits were recorded across the site, one of which contained a boot spur, dating to the seventeenth/eighteenth century, the period of the construction of the house standing to the immediate west. It seems likely that these features represented rubbish pits or garden features associated with the house. A north-south running gulley was recorded in the North Trench which may have had a drainage function or may represent the remains of a heavily truncated boundary ditch. At the east of the North Trench, two postholes were recorded indicating the possibility of a building extending eastwards. A north-south running ditch was uncovered in the West Trench which widened considerably towards the south and contained two pieces of undiagnostic metalworking slag and one fragment of possible furnace lining. This was interpreted as a probable boundary ditch, the pieces of slag indicating industrial activity in the vicinity. A series of pits, postholes and stakeholes, which may represent garden features, were recorded across the site dating to the nineteenth century. One of these contained a piece of possibly residual undiagnostic metalworking slag and one contained fragments of kiln furniture comprising a ring stilt. In a section of the rampart the highest construction layer appears to have been disturbed in this period, representing more intense landscaping and gardening. Most of the above features were sealed by a layer of clayey silty sand which appears to have been a ploughsoil, suggesting a period of agricultural activity or a radical change of gardening practices. Cut into this layer in the North Trench were an ovoid soakaway and two postholes which may have supported an outhouse structure. The soakaway contained one complete and two fragmentary ring stilts. The north of the site was further sealed by three layers of dumped sandy silt indicating a substantial degree of landscaping. The presence of the kiln furniture from this period would seem to indicate pottery working in the vicinity, There is, however, no documentary evidence for pottery production in Tonbridge in the nineteenth century and it must be assumed, therefore, that these pieces must have been transported in from elsewhere. 124 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE THE ORIGINS AND DEVELOPMENT OF MEDIEVAL TONBRIDGE by Jeremy Haslam There are three sources of evidence which are important in any analysis of the origins and early development of Tonbridge and its defences: - the town plan, and the spatial relationships between its various elements - archaeological findings - documentary references The usefulness of these classes of evidence in developing models describing the early development of the town and its defences is enhanced by considering them as being essentially complementary. Furthermore, a study of the defences can only be set in a meaningful context by considering them in their spatial and functional relationship to the town and castle which they enclose. These three aspects of the available evidence together combine to make possible a workable hypothesis about the origin of the town and its defences [which is explanatory on a number of different levels]. The town plan The earliest and best evidence for the spatial relationship between the various elements of the town plan is the Tithe Award map of 183 8, at approximately 1: 1250 scale. This is followed by the ! 81 edition OS map, at both 1 ;500 and 1 :2500 scales (1867 & 1866-9). A ll earlier depictions of Tonbridge - e.g. Hasted's small map, and the slightly larger Bull Inn map of 1699 (Tonbridge Local History Collection A6 ) - are on too small a scale to be really helpful. The Tithe Award maps and the OS plans show the extent of the town defences and the castle, and their relationship to other elements of the town plan particularly clearly. Each of these sources supplies topographical information not given by the other. The town plan, as depicted in these maps, can be divided for the purposes of the present analysis into several discrete areas, which are marked on Figs 3-6. The boundaries recorded on this plan are selected to represent the longest continuous lines.1 These, therefore, reflect (but do not necessarily represent in their entirety) the pattern of the layout of the property boundaries which there is every reason to believe have their origins (with some exceptions) at an early stage of the evolution of the town itself. The divisions of the plan are as follows: 1 the main road (the High Street) running approximately N-s through the town 2 the bridge over the Medway, joining this road and the southern 'suburbs' 3 the castle, consisting of a motte, the inner bailey to the SE of the motte, and the outer bailey to its north-west 125 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM 4 the market area, to the north-east of the castle 5 blocks of burgage plots facing onto the High Street and the market area, numbering 6 in all ( a - f) 6 the area occupied by the bank and ditch defences, which form an envelope around the town 7 the church and its curtilage 8 several discrete areas inside the defences, which are not occupied by burgage plots, lying between the backs of the burgage plots and the line of the inner edge of the bank, numbering five in all (i - v) (including the church and its curtilage) 9 East Street, which enters the defences on the east side and provides a connection between the market area and the Town Mill 10 the town mill There are grounds for suggesting, from evidence and inferences detailed below, that all these elements - apart from a few areas of later infilling - made up the extent of the medieval town in, say, the late thirteenth century. Furthermore, it is possible, by combining observations about the relationship between these elements and a few key documentary references, to put forward a model for the spatial development of the town through some two and a half centuries from the Norman Conquest, and to make some tentative inferences about the significance of the site in the pre-Conquest period. The principal element in the town plan is the Norman castle, sited next to 'one of the few ancient crossing places of the valley' on a marked spur of land of hard Grinstead Clay on the north side of the Medway (Simmons 1996, 101). Since this lies comparatively close to dry land on the south side, it must have been recognised as being readily defensible, commanding a good crossing-place of the braided river and the marshes which no doubt filled the river valley. It would be reasonable to infer that the castle was built to guard or command a crossing over the Medway which was already in existence (Fig. 3), and which therefore had some importance in the pre-Conquest period. The fact that the road approaching the town and castle from the north, through the north gate, makes a noticeable bend to the east as it passes through the town and around the east side of the castle, could be taken to indicate that its straight alignment towards an original crossing was deflected by the building of the castle (Fig. 4). It is more likely, however, that the alignment of the length of the High Street nearest to the bridge was deflected to the east by the building of houses on the edge of the ditch of the inner castle bailey to front onto the High Street (shown clearly on the Tithe Award map), by a process akin to the infilling of properties in the Market Place, alluded to below. These were late medieval timber-framed houses which were only 126 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE Fig. 3 Tonbridge: town development phase 1. t Phase 2 Fig. 4 Tonbridge: town development phase 2. 127 Phase 1 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM recently demolished (in an act of civic vandalism unmatched before or since). If this is so, the position of the bridge has shifted eastwards to its present site. This arrangement implies that the castle bailey and its ditch were built right up to the line of the existing - and, by inference, preConquest - routeway to the river crossing (Fig. 4). The second important feature of the town plan is the central market area. The original extent of this can be reconstructed by removing the two blocks of properties between Back Lane on the west and the High Street on the east. These can be interpreted as classical 'infill' properties which developed within the space of the original market place probably from the later medieval period onwards. There has also possibly been some encroachment on the original market area from the side of the castle, to effectively extend the area of the castle to form a barbican in front of the thirteenth century gateway, though the exact extent of this is unclear. This problem is allied with the uncertainty about the exact course of the ditch of the outer (north-west) bailey at this point, and whether the bank and ditch defences extended around the north side of this barbican.2 A further development of small properties has also taken place to the east of the outer castle bailey over what can be inferred was the latter's outer ditch. The original market area which can be reconstructed by removing these later features thus shows a close organic relationship with the castle itself in being placed immediately outside its principal entrance. It also incorporates access from the inferred pre-existing routeway from the north and from the bridge over the Medway to the south (Fig. 5). Such a spatial relationship implies both a temporal and therefore a functional relationship of the market to the castle and its main gate, the implications of which are explored below. The third important element in the town plan is the area of burgage plots which front onto the market area on its east and west sides. Insofar as the original property boundaries can be reconstructed from the Tithe Award map and the early OS maps these form six distinct blocks or areas: block a - those on the north-west side of the present Back Lane blocks band c - those on the inside of the north gate, on either side of the High Street block d - an area between the High Street to its west and the church to its east block e - an area of longer plots between the market area to its west, a back lane to its east, and East Street to its south blockf- an area between High Street to its west and East Street to its north Between them the edges of these blocks of burgage plots define the extent of the market area and access to it from both its ends. It is therefore surmised that the laying out of the burgage plots and the creation of the market place were contemporary. The properties in the two areas on 128 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE t Phase 3 Fig. 5 Tonbridge: town development phase 3. either side of the High Street and inside the north gate (blocks b and c) do not show the development oflong burgage plots as in the other blocks, and so may well be later additions to spaces which were not developed in the initial phase represented by the burgage plots described above. An important inference which can be made from these observations is that the market area, the areas of burgage plots (not including plots band c - see below) and the access points from the north and south, together form the basic components of the town plan which had their origins as a contemporary planned arrangement. The close spatial relationship of these elements to the main castle entrance shows that in both its layout and its f1.mction it was dependent on the castle, as well as the existence of the river crossing (Fig. 5). The relationship of the four other elements of the town plan to this basic core is of some significance. The church, with its cartilage, is set some way from the castle itself, behind a block of burgages (block cl) which front onto the High Street, to which it is connected by a lane. This suggests that the foundation of the church was not made for the benefit of the castle's inhabitants but for those of the town. It must therefore be seen 129 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM as contemporary with the primary layout of the market and its burgages. The subject of the relationship of the parish of this church to the parishes surrounding it would be a fruitful avenue for further research. The layout of the defences in relation to the town plan elements discussed above is also significant. On both the west and east sides of the town the defences (the line of which is shown quite clearly as a 'Fosse' on the 1 st edition OS maps, but less clearly on the Tithe Award map) are set out some distance away from the backs of the central burgage plots, leaving a zone two to three times the width of the bank and ditch defences as presumably open space inside the defended perimeter. On the east side there are therefore two such blocks behind burgage-groups e and f The space behind the burgage group dis occupied by the church. It seems probable that burgage-groups band c, on either side of the High Street inside the north gate, may well have been open spaces at the time of the original foundation. The space between burgage group a/band the western defences is a similar open area which in the nineteenth century was occupied mainly by a cattle market. From a diagnostic point of view perhaps the most interesting of these intra-mural spaces is that left between the north-west ( outer) bailey of the castle and the defences in this corner. From the point of view of the construction of the defences it would have been more logical to have incorporated these town defences with those of the castle bailey itself, thus shortening their line and taking advantage of the slight break of slope represented by the 25m contour. That this was not done has two main implications for the development of the town. Firstly, it implies that this particular open space was seen as being an important element of the layout of the town which had some significant function at the time of the construction of the defences. The same inference can therefore be drawn in relation to the other intramural spaces. Secondly, it can be concluded that the construction of the defences was not integral with the layout of either the castle with its baileys or the market with its burgage plots, and that the defences were therefore built at a date which was later than the fullest development of the latter (Fig. 6). This topographical hypothesis is strengthened by the comparison of the town plan of To nbridge with those of other castle towns, discussed below. There are also indications from which it can be deduced that East Street/Swan Lane was an original component of the town plan. Its rather sinuous course within the town suggests that it represented a realignment within the town of a routeway from Hadlow to the north-east which originally headed straight for the river crossing, but which was diverted towards the area of the market nearest to the castle gate along a line which was made conformable with the property boundaries of the adjacent burgage plots. The early date of this routeway is also implied by the fact that it gives 130 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE Phase 4 Fig. 6 Tonbridge: town development phase 4. the only direct access from the centre of the town to the Town mill(s), which, being used by the owners of the burgage plots, would have been an original component (with the leat feeding it) of the town as a functioning entity. These ideas are strengthened by the documentary and archaeological evidence, discussed below. The construction of the leat (Figs 5 and 6) accounts, at least in part, for the artificiality of some of the courses of the river, which was probably canalised in various different ways throughout the last nine centuries or more. The original course is probably represented by the more sinuous channel, which is at a lower level to the more northerly course nearest the town, and which feeds the mill leat. It is indeed possible that the mill leat was at one time fed entirely from the flow of the Hilden Brook flowing into the Medway just to the west of the castle, which would have formed part of the southern water defences of the castle, and that the main stream of the Medway has become canalised at a later date to bring its wharves closer to the town. However, it would be difficult to reconstruct the early courses of the river with any degree of certainty, or to be able to ascertain how or at what period they would have been changed from their early 131 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM medieval, or even pre-Conquest, alignments. Further observations on the subject are made by Simmons (1996). The topographical observations have led to certain hypotheses about the development of the town, which can be summarised thus: - the main north-south street (High Street), as well as East Street, represent routeways which may well have been in existence before the formation of the castle and town. As will be argued below, this implies the existence of a significant crossing of the river at this point before the Conquest. - the market place with its burgage plots together represent a new urban foundation which was planned and laid out at the same time outside the principal gate of the castle, and therefore after the latter's construction. - the church was also a new foundation at the time, built to serve the needs of the population of the newly created town. A similar inference can be made about the Town Mill and its leat. - the original bank and ditch defences were created at a date which was later than the laying out of the burgage plots and market, and were constructed in such a way as to leave open spaces inside the defended area. The documentary and archaeological evidence It is necessary to operate within this framework in any examination of the rather sparse evidence from documentary and archaeological sources relating to the town and its defences, and to compare it to other town plans of similar type. What material there is, however, does set the development outlined above more closely in its temporal and therefore functional context. In Domesday Book, Tonbridge itself is not named as a place or as a market, although the castle there had clearly been established around this time as the caput of the extensive estates of Richard de Tonbridge. In her study of Tonbridge, Ward (1980, 119) argues that the grant of Tonbridge was made to Richard, son of count Gilbert ofBrionne, soon after the coronation of William I in 1067, as part of the king's requirement to secure Kent and Sussex after the Conquest. The building of the motte and bailey castle at Tonbridge therefore belongs to this time, and mirrors the granting of estates in Sussex at Arundel, Lewes, Hastings and Pevensey to loyal followers of the king. T he old rapes in Sussex became castelries on the continental model, superseding former administrative units, with much consequent disruption. A similar reorganisation was planned for Kent, with large parts of the county being given to bishop Odo of Bayeux, and to Hugh Montfort (Williams 1995, 18) who had a castle at Saltwood overlooking Romney Marsh. The site at Tonbridge must have been seen as strategically important in securing the 132 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE important crossing place of the Medway on the routeway between Rye and Hastings on the south coast and London. This has certain implications for any assessment of the pre-Conquest significance of the place, although there are rather contrary indications. While this must be somewhat conjectural, it could be inferred that Tonbridge was not only a settlement before the Conquest, but also the site of a late Saxon fortress or burh guarding and commanding a bridge, possibly at a royal centre - a situation reflecting the importance of fortified bridges in the defence of southern England against the Vikings (Brooks 1971 ). This would account for the early existence and importance of the crossing over the river, argued above as probably predating the establishment of the castle. It might also explain the somewhat unusual association of the tun and the bridge elements in the early forms of the place-name (Nicholaisen et al. 1986; Gelling and Cole 2000, 70), and the assumption by a Norman of an Anglo-Saxon appellative, in which the combination of the tun and bridge name-elements seem likely to have had pre-Conquest origins. The existence of such a fortress, combined with a bridge, would therefore explain both the choice of site for Richard's castle in c.1067-8 at a strategically important place, which conjunction conforms to a pattern of the foundation of other early Norman castles on sites guarding important routeways (Drage 1987, 117), and the currency early in the Norman period of a Saxon name which itself indicates this strategic significance. This interpretation is also implied in the sense of the place-name given by Currie (1994, 369), as 'the original sense of a bridge connecting two parts of an early estate here that was separated by the river'. On the other hand - and seemingly in direct contradiction to the last point - work on the manorial and parish boundaries has suggested to Dumbreck (1958, 147) that 'there was no unified area around the site of the castle in Saxon times' - implying the absence of an established preConquest estate centre or settlement at Tonbridge. Dumbreck goes on to conclude that 'the primary purpose of the Lowy was the maintenance and defence of the castle', in origin an artificial post-Conquest administrative area. The hypothesis of the existence of a pre-Conquest settlement at Tonbridge, while a possibility, is therefore not supported by the available evidence. The contrary indications are at least consistent with the suggestion that in the period immediately after the Conquest the castlebuilders chose the site for the castle because of its commanding position overlooking an important and customary crossing place of the river, and that with the construction of the first castle they also built a bridge, or rebuilt an old one, the more easily to control and regulate the crossing. It is clear that the castle at Tonbridge was centrally important for Richard of Tonbridge from early in the Norman period. As well as being the caput of his extensive estates in both Kent and Surrey it was also the centre of his 133 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM court (Ward 1980, 120). This 'honorial court' allowed the lord to hold a court independent of the shire and hundred, as opposed to merely taking the profits of justice, and was 'perhaps the most significant innovation in post-Conquest England' (Williams 1995, 194-5). These factors have certain implications for the assessment of the course of development of the town itself. During the years following its foundation, the castle would have become the focus of traders, merchants, factors, agents, suppliers, lawyers and craftspeople who serviced the needs of the castle and the performance of its jurisdictional and other roles in the management of Richard's estates. That the place Tonbridge was not mentioned in Domesday Book carries the implication that the town and market had not yet been formed in 1086. The castle itself was besieged in 1087 by the forces of W illiam II, but the inhabitants submitted and the castle remained in the hands of the de Clare family ( alias de Tone bridge or FitzGilbert) (EHD II, 174). It seems likely, however, that the market and burgage plots would have been laid out sometime after this episode - possibly towards the end of the eleventh century. It is clear from the discussion of the town plan that these elements of the town were laid out at the same time as a single planned entity immediately outside the castle gates, at a position which also had good access from north and south, as well as from the north-east via East Street. In view of the importance of the castle from the Conquest it seems likely that this market place and its associated burgage plots would have been the formalisation of a more organically arranged collection of traders and others around the castle gate, which had perhaps become established in the 1070s and 1080s. This process has been discussed by Beresford (1967, 55-97), who aptly describes this as the 'magnetic force of attraction' exercised by a castle. The Town Mill must also be recognised as being an essential part of this creation, if it did not already exist before the Conquest. This development forms part of a wider pattern in which castles were often 'powerful stimulants to urban growth' which is reflected in the fact that 80 per cent of all new towns in Wales, and 75 per cent of new towns in England, before 1150 grew up or were established next to a castle (Schofield and Vince 1994, 42). The development of Tonbridge is also mirrored in the fact that while 80 per cent of these boroughs were attached to royal castles, 'in the early twelfth century ... the proportion of royal to seigneurial foundations changed in favour of the latter' (Rowley 1983, 89). A reference of 113 6 to a bequest by Richard de Clare of 20s. per annum to the monks of Lewes from the tolls of goods passing through Tonbridge suggests the town was on a viable commercial route by this period (Wadmore 1885, 19; Salzman 1932, 156, Ward 1980, 124). Norman work is in evidence in the church of St Peter and St Paul, in the form of three rounded windows in the chancel whose simple style would be consistent 134 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE with a late eleventh-century date. Since the church was arguably built for the inhabitants of the new town (above), it can be argued that the town was a thriving community by the turn of the eleventh and twelfth centuries. This arrangement invites comparisons with other planted towns of the period, which will be discussed below. The evidence thus combines to suggest that by c.1100 the market and burgage plots had been laid out in a way which formed a tight-knit but undefended urban community which had been created as a seigneurial plantation by the lords of Tonbridge. Its focus was the market place outside the castle gates, but its traders (if indeed it was more than merely a local market for local produce) would have had access over an established crossing of the river and via well-used routeways to London and Rochester in one direction and towns on the south coast in the other. It is not improbable that the development of the town at this time encouraged the greater exploitation and utilisation of the resources of the Weald, such as iron-making and tool production, as well as timber - a general pattern in new towns noted by Beresford (1967, 125-33) and Rowley (1983, 111). This in turn could have led to the growth of river transport of these goods to Rochester, which is such a feature of its development in the post-medieval period (Chalklin 1961 ). The defences As discussed, the defences of Tonbridge are a major component of the town plan. The defences enclosed a space in which the market was more or less central, which included the church, and which also included several areas behind the blocks of burgage plots which at the time of the construction of the defences were probably open spaces. The bank must have incorporated a gate at the north end of the High Street, as well as on the line of East Street, on the line of which a row of houses is shown on the Bull Inn map of 1699. The constriction of the width of the High Street at the North Gate which is so evident from the later plan is but one indication that the defences were built at a later date than the development of the market. It is however difficult to establish with any certainty the date at which these defences were built (Streeten 1976, 105-6). The historical background is however clarified by comments by Tom Hollobone (pers comm.), who points out that 'a licence to crenellate was granted by Henry III to Richard de Clare in December 1259 (CPR, 44 Henry III, 108). Relations between the Clares and the Crown were complex, particularly during the Montfortian crisis in the 1250s and 1260s. It can be argued that the licence to crenellate was not needed for Tonbridge castle, as one had existed on the site since soon after 1066. The licence could have been for the town, but the more likely reason was to confirm the Clares' 135 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM legal right to title of To nbridge. There had been a long running major disagreement between the Clares and the archbishops of Canterbury over the ownership of the Lowy of To nbridge and various franchises which the Clares claimed for it. It is probably no co-incidence that the licence to crenellate (which the Crown would only have granted if it was satisfied that the Clares had good title to the place) was granted in the same year that Richard de Clare made a historic agreement with the archbishop of Canterbury concerning the nature and extent of the Lowy. Another but probably subsidiary reason was that Henry III needed to keep Richard de Clare sweet, at a time when there was increasing opposition to the King from his son the Prince Edward, Simon de Montfort and his supporters. The licence to crenellate the town has given rise to speculation as to whether this was a licence to add a stone wall to already existing timber and earth defences (e.g. Wadmore 1885, 26). But this would also have applied to the castle itself, and coincides with a period of strengthening of the castle gatehouse from 1265 in response to the uncertainties of the times (Renn 2001, 246). There was no evidence of the existence of a stone wall in excavations on the north and north-east sides of the defences (Streeten 1976), and there is no surviving masonry shown along the line of the bank on the 1 st edition OS maps. It must be concluded, as has Kenyon (1990, 185), that the town defences never incorporated a stone wall, either as an original component or as a secondary addition. This is perhaps a reflection of the fact that the building of a wall around the length of the defences, as opposed to the walling of the castle bailey and the provision of a shell keep in the early thirteenth century (Platt 1982, 20), would have involved a massive effort in quarrying and carting the stone, as well as constructing the wall, which in the scale of resources required would have dwarfed the relatively simple job of digging a bank and ditch. There is therefore some uncertainty as to whether the defences were constructed in or soon after 1259, or whether they were already in existence at this date. The question is only likely to be resolved through finding good dateable material from sealed contexts in further excavations. From a historical point of view the defences could have been added at any time between the mid twelfth century and the 1260s. Whether or not defences were added to towns and castles, or whether either was strengthened at any time, depended very much on local conditions and the perceived threats from outside, as well as such factors as whether the town defences would bring enhanced status for the community. From the point of view of the local seigneur, who in the case of Tonbridge would have born the cost of the works, the addition of defences to his borough would have increased his control of both the market and the bridge, giving him fixed access points at which to collect tolls, as well as providing more intra-mural space in which further development could be encouraged. 136 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE This intention seemed to have been realised to a limited extent by the excavated evidence of medieval property boundaries established in the formerly open area at the Lyons site. All these factors might lead to expectations that the defences were in place in the twelfth century rather than the thirteenth. However, Renn (2001, 246) has argued that the gatehouse of the castle was built in c.1265 by Gilbert de Clare, before he was engaged in the Welsh Marches from 1268, and before his power was reduced by King Edward I in the 1280s and 1290s. As Kenyon (1990, 185) has also argued, this would certainly have been a suitable context for the addition of defences to the town. The most plausible hypothesis must be that the town defences were constructed at the same time as the reconstruction of the castle and its impressive gatehouse in the 1260s in order not only to defend the town but also to augment the landward defences of the castle. Comparative town plans The plan and early development of Tonbridge has been shown above to have developed in close association with the castle of c.1067-8. A large number of such castle towns developed or were founded in England and Wales in the first hundred years after the Conquest. The variety of plan types is large, showing variations influenced by such factors as the siting and location of the town or castle as well as both the military and economic aspirations of their builders. These have been discussed by a number of authors (Beresford; Platt; Rowley; Drage; Kenyon; Schofield and Vince; Creighton), variously on broad and narrow canvasses. It is not therefore the intention to set the early development of Tonbridge in its national context here. However, one of the more interesting features of the growth of the town plan of Tonbridge is the development of the defensive circuit to enclose an area which is rather greater than that of the burgage plots and market. This has led to the idea (above) that its defences are a later addition to the development of the town. In this aspect it can be contrasted with several other castle towns of the late eleventh and early twelfth centuries in which the burgage plots within are closely confined within its defences, and where these town defences form part of the outworks of the castle. In these the formation of the town is a planned unit, created by the lord of the castle, which is therefore contemporary with the layout of that part of the castle. A recent suggestion made by Creighton (2002, 161 ), which is diametrically opposed to the conclusion made above, that Tonbridge is an 'example of a semi-circular borough with a similarly decayed street plan' [to Trowbridge, Wilts] cannot be sustained. The plan analysis of Tonbridge shows, firstly, that the development of the town is not semicircular (i.e. it is not constrained within a semi-circular bailey as 137 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM with the case of Devizes, below), and secondly, that it has not contracted from a larger medieval extent. Perhaps the most interesting new town of the period whose foundation is closely associated with a castle is Devizes, Wiltshire, which was laid out within the outer bailey of a castle built in c.1120 by Roger, Bishop of Salisbury, on the borders between two of his manors (Beresford 1967, 504; Haslam 1980; Rowley 1983, 91), but close to a pre-existing village. The burgage plots of the first phase, together with a new market place and church, were laid out around a curved street inside, and defined by, the outer bailey of the castle. The success of the town and its consequent expansion through the twelfth century has led to a development of a secondary market place within the inner bailey of the castle, which shared the church of the castle as its parish church, and whose shape reflected the curved shape of the inner bailey. These separate plan elements were however defined by an envelope of bank and ditch defences, within which there was no open space apart from the centrally-placed market. Other towns of the period with similar plans, whose plan-elements are constrained by being fitted into the defensive enceinte of a castle, can be recognised at, for instance, Richmond, Plessey, Launceston, Wisbech, Tutbury and Eye, and are discussed by Rowley (1983, 92-8). The development of this type of castle-town at the same time as the establishment of Tonbridge as a defended settlement as well as a castle serves to emphasise how it expanded in several stages. MEDIEVAL TONBRIDGE IN CONTEXT The tail of the town rampart was found in the East Trench of the excavations although in order to ensure preservation in situ it remained largely unexcavated. It comprised a number of layers of sandy clay and clay sand to a height of 1.50m above the surrounding area. No turf line was observed, although this may have been present below the layers exposed. Furthermore, no evidence of either a timber palisade or stone walling was found and no artefactual evidence was recovered. Two pits were recorded to the west of the rampart in the East Trench, and one in the West Trench, suggesting an open area behind the ramparts, possibly for refuse deposition. In the North Trench, a beam slot was recorded suggesting some form of structure fronting onto East Street. Four pits immediately to the east of this structure indicate an open area used for the deposition of refuse probably associated with the structure. An east-west running ditch was located in the West Trench which was interpreted as a boundary. The animal bone assemblage from this phase included examples with evidence of butchery and suggests domestic activity. The pottery assemblage similarly implies domestic activity and includes sherds of 138 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE Kingston, London and Saintonge wares as well as the more predominant local Limpsfield and North or West Kent wares, suggesting fairly widespread contacts. Pieces of metalworking slag were recovered from five contemporary features indicating that metalworking was taking place in the vicinity. The lack of evidence for the fourteenth century suggests a contraction of activity in this sector of the town in common with the demographic and economic downturn elsewhere in England at this time. The area of the rampart exposed backed onto open area v (see Haslam's study above). The presence of three pits, probably for refuse disposal, would seem to confirm that at least part of this area was open land behind burgage plot f To the north of the area, however, the presence of the beam slot suggests that occupation was increasing along East Street from burgage plot f, although the four rubbish pits located to the east indicate that this may have been the greatest extent of building during this period. The construction of the rampart was similar to that found in the excavations undertaken in 1972 and 197 6 (Streeten 197 6), on the northern defences. No evidence of a stone wall or wooden palisade was found there, although a buried turf line was observed in one of the sections. This may imply either that the rampart was constructed over a sizeable period of time using slightly different construction techniques, or that irregularities in the immediate topography necessitated either building up or levelling down of the ground. One sherd of pottery dated to 1270- 1320 was recovered from the upper layer of one of the sections. Only one other feature of medieval date was recorded during these excavations, which was a drainage ditch running parallel to the High Street which was thought to be somewhat later in date than the rampart. There are a number of examples of town ramparts constructed of earth without evidence of stone walling, such as Taunton (first recorded in 1158 and improved 1215-16) and Richard's Castle (built c.1200). Other earthen defences were improvements to earlier Saxon ramparts such as at Hereford (improved in 1190) and Ipswich (early thirteenth century) (Kenyon 1990). From the mid thirteenth century into the fourteenth century, however, both new and rebuilt town defences seem to be predominantly constructed of stone such as at Conwy (built 1283-1287), Southampton (built in earth and timber c.1202, rebuilt in stone 1360- 1372) (Platt and Coleman-Smith 1975), Canterbury (built late fourteenth century) and Newcastle (built mid-late thirteenth century), although those at Rhuddlan (built in 1277) were still constructed of an earth bank and wooden palisade (Kenyon 1990). The date of the building of the town ramparts at Tonbridge has long been a matter of some conjecture. In the excavations of 1972 and 1976, one sherd of pottery dated to 1270-1320 was recovered from the upper layer 139 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM of one of the sections which may indicate the period of construction. It may, however, possibly relate to repair work carried out after the receipt in 1318 of a murage grant for three y ears (Ibid). A licence to crenellate was granted by Henry III to Richard de Clare in December 1259 which could have given the go ahead to the building of the ramparts, or could have been for political reasons as discussed above. Indeed it may have been granted retrospectively, the ramparts having been built earlier in the thirteenth or twelfth century in common with most others of this type (although one should bear in mind the example of Rhuddlan showing that the old style had not entirely died out). The expenditure involved in constructing ramparts suggests that the town would have been going through a period, of prosperity and, presumably, also expansion. The pottery assemblage from the North Trench of our excavation suggests a probable date of 1250-1300 for the building and associated rubbish pits. While Haslam suggests that area v was open when the ramparts were built it seems probable that construction would have taken some time and, therefore, expansion could have been taking place along East Street immediately after or, indeed, concurrently with the rampart construction. If the extent of Tonbridge was as shown in Figs 5, 6 at the conception of the ramparts, then the most logical place for expansion within them was along East Street as the second most important thoroughfare of the town. Bearing in mind that a licence to crenellate was issued in December 1259 and assuming that the pottery found in the earlier excavations relates to repair work associated with the murage grant of 1318 this would suggest a possible date of 1260-13 00 for their construction. The evidence that the power of Gilbert de Clare was curtailed by Edward I in the 1280s and 90s suggests that this activity is most likely to have occurred c.1260-1280 while simultaneously, or very shortly after, occupation was extending along East Street. ACKNOWLEDGEMENTS Pre-Construct Archaeology Ltd would like to thank Fairclough Homes for funding the project through CgMs Consulting (Archaeological Consultant, Paul Chadwick). The authors would like to thank Gary Brown, the project manager, David Divers and Frank Meddens for all their advice and assistance. Thanks are also due to Simon Deeves who supervised the first phase of excavation and to the excavation staff: Ken Bazeley, Strephon Duckering, Ruth Duckworth, Colin Forcey, Fiona Keith-Lucas, Dan Slater and Will Valentine. Thanks are also due to Giles Hammond, the surveyor, and Richard Young, the photographer. Finally, thanks to Barry Bishop, Nick Branch, John Brown, Jeremy Haslam, Chris Jarrett, Lynne Keys and Beverley Meddens for their specialist contributions, to Tom Hollobone 140 THE DEVELOPMENT OF MEDI):IVAL TONBRIDGE for the fruits of his research and to Josephine Brown and Michael Miles for the illustrations. Chris Jarrett would like to thank John Cotter of Canterbury Archaeological Trust for his assistance with the identification of the pottery. ENDNOTES 1 I.e. the straightest boundaries, uninterrupted by later modifications altering their linearity. 2 In calling this forecourt a 'barbican', which implies that it was a fortified extension of the castle, it is not intended to prejudge the issue of whether this area was or was not in fact defended. An observation made by King (1782, 279-80) suggests that there was a ditch immediately in front of the gatehouse which was crossed by a wooden bridge, which reference is in the writer's view misinterpreted by Streeten (107) as referring to a north gate of the town defences, an arrangement which would fit the local topography at this point. Anthony Streeten has shown that the area inside the north gate was not developed until the early eighteenth century. APPENDIX POTTERY ANALYSIS by Chris Jarrett There are 364 sherds of pottery dating mostly to the thirteenth century and early post-medieval period through to the nineteenth century. There are also five residual sherds of Roman pottery. The pottery is classified according to the Canterbury Archaeological Trusts's coding system. Thirteenth century The main pottery-type (58.3% by sherd count) comprises North or West Kent sandy and shell-tempered ware (fabric EM36) (Figs 7, 8.1- 11 ), occurring mostly as jar-shaped vessels; some vessels show evidence that they are of a composite manufacturing technique where the bodies were hand built and the rims formed on a turntable. This is a technological feature seen in other local industries, such as at Titsey in Surrey (Jones, 1997, 68). Although a number of production centres could have produced this ware, it seems likely that these came from the Limpsfield area. Other products from the Limpsfield kilns, some 13 miles north-east of Tonbridge, were found and include a grey sandy ware (fabric M44) (Figs 8, 9.13-15), of which there are five sherds (4.6%), dated between 1150-1300 (Prendergast 1974; Ketteringham 1989). This fabric is not dissimilar to the grey ware of South Hertfordshire, and both wares are part of a tradition for grey ware pottery in the London area and South East England between the late twelfth and mid fourteenth century. Coarse Limpsfield-type ware (Fig. 9.16-17) is present and represented by 17 sherds (15.7%). It is distinguished by sparse bivalve shell fragments and coarser sand tempering (fabric M44A). It is present mostly as jar-shaped vessels with squared rims (Fig. 9.17) and convex bases. North or West Kent sandy ware (fabric M38A), often with sparse coarse chalk !nclusions, accounts for 11 sherds (12%) of the pottery and is present as jar and Jug sherds. The ware was probably produced at several places in east Surrey and north-west Kent, and a possible production site was at Crockhurst Street just 141 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM ,1 t .. ,, ... ,.,􀀞}'+4')w:·{•H\􀀟-􀀠.:••, 􀀡,,􀀢. , Fig. 7 Pottery: items 1-7. 142 ,. '/II,_, 0 10cm lw----1 4 ~--~-- -~~ .~. 60 ···~ --~~- '., ' ,, ., 1; , ' ~'. ',_ ;~·.,. 70 I . - ' I ~ = = = = = Ii THE DEVELOPMENT OF MEDIEVAL TONBRIDGE 1􀀄1· 12,r- ,, 14,,;,. I I r 11.􀀓;;;; ;;;,: ;;:;;; ,.,;;;,"" 'L 19cm MM MM Fig. 8 Pottery: items 8-14. 143 .1.~ ... . · , ' i ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM 15 •- . . I I 16 J􀀄.· ,. 􀀄 ,. ,, 1·.s• :-·:, ::+. 􀀊 . ., •· . ." •',. . .. ', •, ', ·􀀅 I 18 19 20 0 10cm -----1 Fig. 9 Pottery: items 15-20 144 -----~""':-"~"' - -·,.,.. ~·""" = ::: = = = THE DEVELOPMENT OF MEDIEVAL TONBRIDGE to the east of Tonbridge (pers. comm., J. Cotter). Three sherds have similarities between this sandy ware and the shelly ware (EM36/M38A) and it comprises jar-shaped vessels (Fig. 8.12). Glazed wares are present in low quantities and are either of a London, Kingston, Surrey or local source, with small sherds of a greenglazed, fine whiteware that is probably from Saintonge, in south-west France. There were a number of pits, [14], [28], [48], [120], [136], [138] which produced medieval pottery. Pit [14] produced a single sherd of a Limpsfield-type greyware jar with external sooting. Limpsfield-type wares are dated to between 1150-1300. The pottery present in pit [ 48] consisted of three sherds, all from jar-shaped vessels and representing Limpsfield-type shell-tempered ware (EM36) with an internal carbonised deposit, a wheel thrown flat base in North or West Kent sandy ware (fabric M38) and a sherd of shell and sand tempered ware (fabric EM36/M38). This would indicate a late twelfth and thirteenth century deposition date. A total of 64 sherds representing approximately 32 vessels were recovered from pit [120]. There are two sherds which are early medieval in date and fall within the North or West Kent sand and shell ware tradition fabric (EM36). The first vessel is in the form of a handmade bowl or dish with a slightly thickened, but :flattened, knife-trimmed top (Fig. 7.1). It is probably eleventh or twelfth century in date but could be Late Saxon (pers. comm., J. Cotter). The second vessel is the rim of a jar dating to the twelfth or early thirteenth century. The main pottery types in this pit are Lympsfield-type wares that consist of two sherds of the finer ware (M44). Limpsfield-type coarse ware (fabric M44A) is present in a medium sized rounded jar with a squared rim and knife stabbing decoration on the top (Fig. 9.16). Shell-tempered Limpsfield-type ware (EM36) accounted for 36 sherds of pottery in this pit and is present as fragments of jars (Fig. 8.8-10). There is also one with an applied, pinched vertical strip (Fig. 8.9). Open forms in this ware were well represented and include rims of medium rounded bowls (Fig. 7.2-3). There is also a socketed example (Fig. 7.5), the socket being roughly handmade and a dish. Sherds of North or West Kent sandy ware (M38) occurred in four sherds of thin walled vessels, probably from unglazed jugs, but one sherd has an external green glaze, while a jar in a sparsely, fine shell-tempered version of the fabric had a hooked rim (Fig. 8.10). Other glazed wares occur as single, small sherds from jugs. Pit [ 119] produced the twisted handle of a Kingston ware (M7) jug, dated 1230/40-1400 and a small green-glazed sherd of Saintonge ware (M22G), dated 1250-1650. London-type ware (fabric MS) was present as three sherds that include the base of a baluster jug, another sherd with an external white-slip and coarse clear-glaze, while a body sherd had a combed wavy line. London-type ware was produced between 1080-1350. There is also a single intrusive jug sherd in Medway hard silty-sandy ware with chalk (LM34B), dated c.1450-1525. Present in a possible glauconitic sandstone is the base of a possible mould (pers. comm., J. Cotter). The exterior surface was burnt but there were no other residues associated with metallurgical processing present on the mould. There is other metal working evidence in pit [120] represented by (?) fused sand in Lyn Keys report. Pit [136] contained two sherds of shell-tempered ware (EM36) as a socketed bowl (Fig. 7.6) and two jar sherds. Pit [138] produced a total of 32 sherds representing approximately 23 vessels. Limpsfield-type ware (fabric M44) was present in the form of a strap handle of a jug with point stabbing and a rim sherd 145 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM from a possible jug which was decorated with knife slashing on the top and notches on the side of the rim while the neck had diagonal knife cuts (Fig. 9.15). Coarse Limpsfield-type ware (fabric M44A) comprised a jar sherd, while the shell-tempered wares included the flat-topped rim of a rounded bowl (Fig. 7.4) and sherds of jar-shaped vessels such as a rim with point stabbing on the top (Fig. 8.11 ). The latter comprised wheel-thrown and handmade examples. Point stabbing is a particular identifying technique used in the Limpsfield industry (Jones 1998, 218). There was also a sherd of a North and West Kent sandy ware jar. Pit [28] contained a single sherd of medieval Limpsfield coarse ware and a sherd of (?)Wealden/Harlepan hard fine sandy ware (LMI 7), which appeared to be intrusive. In pit [140] the rim of a North and West Kent sandy ware jug was found with fine surface rilling typical of the industry (pers. comm. J. Cotter) and the rim of ajar (Fig. 8.12) in a shelly version of(EM36/M38). Beam slot [172] produced a single sherd of a (?)Wealden/Harlepan hard fine sandy ware (LMI 7) with horizontal incised lines which appears to be intrusive and a sherd of medieval sand and shell/ North and West Kent sandy ware (EM36/M38). The forms present imply domestic activity, with a predominance of jar-shapes, some of which were used as cooking pots as indicated by the presence of external sooting and internal residues, but some jars were probably used for storage and other kitchen activities. There were at least three socketed bowls (Fig. 7.5-7), one being residual, bowls and dish fragments are relatively common in the assemblage. Medieval glazed wares are sparse and occur as single sherds. Excavations previously in the town at Lansdowne Road also produced medieval coarse pottery of a similar date, with shell-tempered wares that were believed to have been of a Wealden origin (Streeten 197 6, 112-15). However, the study of pottery assemblages in this area of West Kent is fairly limited; the pottery types found here, especially the glazed wares are found on other sites in the area such as at Eynsford Castle (pers. comm., J. Cotter). There is a fragment of a possible metal-working mould that indicates limited industrial activity in the vicinity. A small number of eleventh- or early twelfth-century residual sherds may indicate activity nearby that is associated with the earlier history of the town. The jar forms here are of a late twelfth- and thirteenth-century date. Pit [120] included Kingston ware, dated 1230-1400 and a sherd of Saintonge ware, dated 1250-1650 which would indicate a deposition date in the late thirteenth century. The pottery groups recovered from features in Phase 2 almost entirely contained shell-tempered wares that stopped being produced by c. 1300 or soon afterwards. It seems likely that activity on the site stopped sometime in the early fourteenth century as pottery types from this and the succeeding century are largely absent. Boundary ditch [5], apart from two residual sherds of Roman pottery (including part of a Samian bowl), comprised medieval wares. These were sherds of Limpsfield greyware, parts of a bowl and jug, Limpsfield coarseware, while the sandy shell-tempered ware included the rim of a bowl or dish. Glazed wares included a small sherd of Saintonge green glazed ware, a sherd of a jug in a white-slipped Surrey or Kingston fabric (KING SL), dated c.1230-1400, and a small sherd of Coarse Limpsfield/Earlswood-type ware, dated to the thirteenth or fourteenth century, with an internal splashed green glaze. 146 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE Sixteenth century Pit [26] produced three sherds of Limpsfield-type ware but it also included a sherd of Calcareous 'Peppered' smooth ware (PM64), dated 1550- 1650. The pottery in pit [107] was less fragmentary and produced a ?Wealden/ Harlepan hard fine sandy ware (LMl 7) flared bowl with a white slip wash around the rim (Fig. 9.18) and the base ofa flared bowl in Wealden buff fine sandy ware with haematite (PM2), dated 1525/50-1625/50. Pit [114] produced a single sherd of (?)Wealden/Harlepan hard fine sandy ware (LMl 7), dated 1525-1600. Pit [125] produced a single sherd of a Wealden buff fine sandy ware with haematite (PM2), dated 1525/50-1625/50, comprising a possible jug with light incised lines. Seventeenth/eighteenth century There were four sherds of pottery recovered from pit [97] which included a sherd of (?)Wealden/Harlepan hard fine sandy ware (LMl 7) with an internal coarse glaze and a small sherd of (?)Wealden light orange sandy ware (PM2.7), dated 1650-1750. The finewares included the base sherd of a Chinese porcelain blue and white (PM40A) medium rounded bowl with 'an hua' (incised) decoration in the form of a scroll and the base of a Tin-glazed earthenware (PM9) charger with a blue pinwheel flower on white. Both the Chinese porcelain and the Tin-glazed earthenware charger are of a mid seventeenth-century date. Pit [134] produced a single sherd of a Calcareous 'Peppered' smooth ware.(PM64), dated 1550-1650 jug and two sherds of (?)Wealden/Red Border ware, dated c.1675-1760 that includes a strap handle. Ditch [22] produced mostly thirteenth-century Limpsfield-type wares but it also included two sherds of (?)Wealden fine buff sandy ware (LM15), dated c.1450-1550, and Medway hard silty sandy ware with chalk (LM34B), dated 1450-1525/50. There is also a sherd of Tudor green ware (LM15) and a sherd of a brown-glazed Wealden buff fine sandy ware with haematite (PM2), dated 1525/ 50-1625/50 and a possibly residual sherd of Red earthenware with iron-streaked glaze (?High Balden), dated 1800-1900. Pit [44] produced a sherd of Frechen stoneware (PMS) and the base sherds of a London stoneware (PM25) tankard and a gorge (rounded mug). Probably contemporary with the London stoneware was a blue on white tin-glazed plate with a Chinese style jasmine design, dating to the start of the eighteenth century. There are also three sherds ofa ?Wealden light orange fine sandy/(?)Wealden fine pink-buff with marl inclusions (PM2.7/PM2.6) firepot, with a collared rim and a cutout in the body. The term fire pot comes from Flemish and Dutch paintings where they are illustrated as containing embers and placed on the floor under the skirts of lace-makers. In Colchester there was a community of Dutch immigrants largely in control of a flourishing textile industry, to which a number of local redware firepots have been linked (Cotter 2001, 215-16). However, although a Low Countries form in origin, it could represent a general brazier (pers. comm., J. Cotter). Pit [68] contained in its fill [67] a sherd of Staffordshire white salt-glazed stoneware (fabric PM26A), dated 1720-1780 and a largely complete polychrome painted Pearl ware coffee pot, dated c.1790-1820. Local Wealden fine sandy redwares were also present and notably there was a fragment of kiln furniture in the form of a late post-medieval redware (LPMl) ring stilt (Fig. 9.20). Pit [100] produced a single sherd of a plain white tin-glazed earthenware (PM9W) bowl, 147 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM dated 1630-1800. Post-hole (103] contained in its fill [102] a fragment of an eighteenth-century London stoneware tankard. It had incised on it 'Da .. B.' which may possibly refer to the owner, or possibly a marriage or a local public house. Pit [104] produced a single sherd of Medway hard silty-sandy ware (M34A), dated 1450-1525/50. Post-hole [116] contained as its latest dated pottery types the base of a Staffordshire white salt-glazed stoneware (PM26A) jar, a Developed Creamware plate with a queens edge rim, dating to the late eighteenth century. Also present are sherds of (?)Wealden light orange sandy ware (PM2.7), dated 1650-1750 and brown-glazed Wealden/Red Border ware that included sherds of a jug. Late eighteenth to nineteenth century Backfill [63] of soak away cut [61] produced Post-medieval red wares (PMZ and PM64) dating to between 1550- 1650. The infilling (fill [64]) of the masonry soak away [62] indicates an end of the eighteenth century or start of the nineteenth century deposition date by the presence of Developed Creamware, painted English Hard paste porcelain and a sherd of Transfer-printed ware. Of interest are the presence of kiln furniture comprising one complete and fragments of two ring stilts (Fig. 9.19-20). The plough soil [69] produced mostly local red earthenwares but also a sherd of Developed Creamware, dated 1775-1880. Post-hole [86] contained in its fill [85] post-medieval redwares, a sherd of Frechen stoneware and a residual sherd of battered Tin-glazed ware of seventeenth- or eighteenth-century date. Nineteenth century The levelling layers [58], [59] and [60] all contain nineteenthcentury industrial finewares as the latest pottery and include Developed Creamware, Pearl ware, and 'Ironstone' type refined whitewares, often decorated with transfer printed designs. Kiln furniture and local post-medieval redwares Of interest for the late eighteenth to nineteenth-century features are the presence of kiln furniture in the form of ring stilts (Fig. 9.19-20). A complete example and parts of two other ring stilts were found in fill [64] of the soak away [62] and sherd links to a fragment of these items in fill [67] of pit [68] to produce another complete example. These ring stilts are in the Late Post-medieval red earthenware· fabric (LPMl), their dimensions range between I OO- I 20mm in diameter and they are wheel thrown. They have flat bases and a rounded top. However, one example had a fillet of clay added to the base to form an internal lip (Fig. 9.20). There are three regularly spaced upward pointing prongs on the top of the ring, formed by pinching the finger and thumb together. Examples have a red wash and one has a clear glaze. Ring stilts were used as spacers to stop glazed wares fusing to other vessels, but there is no evidence that these items were used in a kiln and additionally there was an absence of pottery wasters on the site. However, these ring stilts may be the first archaeological evidence for pottery production in or near Tonbridge, alternatively they may represent imported dumped material. Pottery production is known in the area during the nineteenth century, Wrotham, which is well known for its earlier seventeenth- and early eighteenthcentury dated slipwares products, while at Hadlow, some 4 miles to the north-east of Tonbridge there was a pottery operated in the 1850s by G. Richardson. W. Richardson had moved here to Shipborne Road in 1887 from Tunbridge Wells (Brears 1971). 148 THE DEVELOPMENT OF MEDIEVAL TONBRIDGE The post-medieval redwares found on the site mostly fall under the Wealden light orange fine sandy ware code (PM27), and are typically pasty in appearance, with fine sand and only sparse to moderate coarser quartz inclusions. T hey fire to a bright buff colour, pink-buff, or light orange buff colour. Vessels usually have a clear ( often bright orange-red) or brown glaze. This fabric could have been produced at a number of sites in the Weald, but is most likely to have been of a local origin and may include Wrotham products, but the well-known slipwares are absent from the site. Fabric PM27 is dated from 1525, becoming common after 1650 and by 1825 it merges with the Late post-medieval redware fabric (LPM I), c.177 5-1900 which subsumes all Kentish red wares. However, the LPM 1 fabrics on the site do have the appearance of being Wealden in origin and are probably of a local source; Wrotham, Tonbridge or further afield, from High Halden near Ashford as the wares from these potteries are difficult to separate from each other (pers. comm., J.Cotter). BIBLIOGRAPHY Beresford, M.W., New Towns of the Middle Ages, 1967. Brears, P.C.D., The English Country Pottery: its History and Techniques, 1971. Brooks, N.P., 'The Development of military obligations in eighth and ninth century England', in P. Clemoes and K. Hughes (eds), England before the Conquest (Cambridge), 69-84, 1971. Brown, R.A., 'Royal Castle-building in England, 1154-1216', in Castles, Conquest and Charters: Collected Papers (Woodbridge), 1989. Chadwick, P., 'Archaeological written scheme of investigation, Lyons, East Street, Tonbridge, Kent', CgMs unpublished report, 2001. Chalklin, C.W., 'A Kentish Wealden P arish (Tonbridge) 1550-1750', BLitt thesis, University of Oxford, 1960. Chalklin, C. W., 'A seventeenth Century Market Town: Tonbridge', Archaeologia Cantiana, LXXXVI, 152-62, 1961. Cotter, J. P., 'P ost-Roman pottery from excavations in Colchester, 1971-1985', Colchester Archaeological Report, 7, 2000. Creighton, O.H., Castles and Landscapes, London, 2002. Currie, A.M., Dictionary of British Place Names, London, 1994. Dumbreck, W.V., 'The Lowy of Tonbridge', Archaeologia Cantiana, Lxxm, 138- 147, 1958. Drage, C., 'Urban Castles' in Schofield and Leech 1987, 117-32. Douglas, D.C. and Greenaway, G.W. (eds), English Historical Documents II (1042-1189), 2nd edn, Oxford, 1987. Ford, S., 'Lyons, East Street, Tonbridge, Kent; an Archaeological Desk-Based Assessment', unpublished report, 2000. Gelling, M. and Cole, A., T he Landscape of Place-names, Stamford, 2000. Haslam, J., 'The Excavation of the Defences of Devizes, Wilts', Wiltshire Archaeological Magazine, 72173, 59-65, 1980. Hasted, E., The History and Topographical Survey of the County of Kent, v., 1798. Hayman, G., 'The excavation of two medieval pottery kiln sites and two sections through the London-Lewes Roman road at Clacket Lane, near Titsey, 1992', Surrey Archaeological Collections, Vol. 84, 32-76, 1-88, 1997. 149 ELIOTT WRAGG, CHRIS JARRETT AND JEREMY HASLAM Jones, P., 'The Pottery', in Hayman, G., 'The excavation of two medieval pottery kiln sites and two sections through the London-Lewes Roman road at Clacket Lane, near Titsey, 1992', Surrey Archaeological Collections, Vol. 84, 32-76, 1997. Kenyon, J.R., Medieval Fortifications, Leicester University Press, 1990. Ketteringham, L. L., 'Two medieval pottery kilns at Limpsfield Chart', Surrey Archaeological Collections, Vol. 79, 125-146, 1989. Keys, L., 'Metalworking slag from Lyon's East Street, Tonbridge, Kent', unpublished report, Pre-Construct Archaeology, August 2002. King, 'Mr King's Sequel to the Observations on Ancient Castles', Archaeologia, vi. 1782. Mortimer, R., Land and Service: the Tenants of the Honour of Clare, ANS 8, 180, 1986. Nicholaisen, W.F.H., Gelling, M. and Richards, M., The Names of Towns and Cities in Britain, London, 1986. Pine, J and Saunders, M., 'Lyons, East Street, Tonbridge, Kent, an Archaeological Evaluation', 2000. Platt, C and Coleman-Smith, R., Excavations in Medieval Southampton 1953- 1963. Vol 1., Leicester University Press, 1975. Platt, C., Medieval England, London, 1978. Platt, C., The Castle in Medieval England and Wales, 1982. Prendergast, M.D., 'Limpsfield coarseware; a descriptive analysis', Surrey Archaeological Collections, Vol. 70, 55-77, 1974. Renn, D., Norman Castles in Britain, 2nd ed., 1973. Renn, D., 'Attack and Defence at Old Soar Manor, Plaxtol', Archaeologia Cantiana, cxx,, 237-250, 2001. Rowley, T., The Norman Heritage, 1055-1200, London, 1983. Salzman, L.F. (ed.), 'The Cartulary of the Priory of St. Pancras, Lewes', Sussex Ree. Soc., XXXVIII , 1932. Simmons, S., 'Tonbridge Castle: further observations on an ancient castle', Archaeologia Cantiana, cxv1, 101-46, 1996. Schofield, J. and Leech, R. (eds), Urban Archaeology in Britain, CBA Research Report 61, 1987. Schofield, J. and V ince, A., Medieval Towns, Leicester University Press, 1994. Streeten, A.D.F., 'Excavations at Lansdowne Road, Tonbridge, 1972 and 1976', Archaeologia Cantiana, xcn, 105-118, 1976. Wadmore, A., 'Tonbridge Castle and its Lords', Archaeologia Cantiana, xv1, 12- 57, 1886. Ward, J.C., 'The Lowy of Tonbridge and the lands of the Clare F amily in Kent, 1066-1217',Archaeologia Cantiana, xcv1, 119-131, 1980. Williams, A., The English and the Norman Conquest, Woodbridge, 1995. 150