Archaeological Investigations along the Lamberhurst bypass: a medieval iron-working site at Spray Hill
With contributions from: Luke Barber, Rod Mackenzie, Rob Scaife, Eden Hutchins
The construction of the A21 Lamberhurst Bypass provided archaeologists the opportunity to sample a linear swathe through the landscape to the east of the village of Lamberhurst, including part of the Scotney Castle estate. During the course of the archaeological investigations evidence was recorded for a prehistoric palaeochannel in the valley of the River Teise, and for a previously unknown 13th-century iron working site on Spray Hill. Excavation was also undertaken on the line of the medieval park pale to the Scotney Castle estate, within post-medieval quarries at Old Ruffetts and Piercefield Shaw, and on the line of the Scotney Park drive .
The A21 Lamberhurst Bypass runs to the east of the village between junctions at (NGR 56784 13515) and (NGR 56825 13737) (Fig. 1). Advance investigation, commissioned by Kent County Council, identified six potential archaeological sites along the proposed 3.2 km route (Barber 1992). A subsequent heritage evaluation of the route, commissioned by the National Trust and including a review of the history and landscape development of the Scotney Castle estate, made recommendations for further archaeological evaluation (Bannister 2001). On the basis of these, the Highways Agency defined six areas of archaeological interest, designated as Cultural Heritage Areas, for further investigation. In addition, the contractors were required to maintain a watching brief on all topsoil stripping operations, in order to safeguard any other archaeological remains which might be uncovered. All Ordnance Survey mapping used in this report is published under (c) Crown copyright 2024 Ordnance Survey Licence no. AC0000819218.
[fg]jpg|Fig. 1 Site location. (Reproduced from the Ordnance Survey with the permission of the Controller of His Majesty’s Stationery Office, Licence no. AL100014861.)|Image[/fg]
The topography of the bypass route is typical of the High Weald of Kent and East Sussex, comprising rounded hills, cut through by river valleys. From the northern end of the road corridor the ground falls from 85m to 36m aod at the bottom of the Teise valley. The valley bottom is essentially level, and then the corridor rises in steps as it skirts the Scotney Castle estate to the summit of Spray Hill at 86m, from where it crosses the castle drive and falls to the southern junction at 70m.
The corridor crosses five different geological drift deposits, with the alluvium deposits of the valley floor surrounded by Ashdown Beds. Wadhurst Clay covers the northern end of the route and the area from Piercefield Shaw southwards as far as Spray Hill, which is dominated by Tunbridge Wells Sands giving way to clay.[pg132][pg133]
Archaeological and historical setting
Prehistoric: very little evidence for prehistoric activity or occupation has been identified along the line of the Lamberhurst Bypass. A single broken ovoid flint scraper was collected at the northern end of the route in the course of the field survey in the early 1990s (Barber, 1992). This suggests that any settlement of this period may have been sited on the better drained soils of the higher ground, overlooking the valleys of the rivers Teise and Bewl.
LIA/early Romano-British period: ironworking is known to have begun in the High Weald in this period. Most of the recorded sites are concentrated to the south and south-west of Lamberhurst, to the east of Roman Road 14 (Cleere & Crossley, 1995). No evidence for Roman activity has yet been identified within the bypass corridor.
Anglo-Saxon: research by Nicola Bannister on the Scotney Castle estate has indicated that this part of the Weald was the last to be occupied by Anglo-Saxon/ Jutish settlers (Bannister, 2001a, 16). They established four dens (woodland swine pastures) in the southern part of the parish of Lamberhurst, which included the land between the Bewl and Teise as well as that at Bewlbridge and Lamberhurst Down (ibid.). It is likely that small settlements developed in some of these swine pastures as the swineherds felled trees to construct huts (Muir 2000, 5-6). Place- name evidence, including the frequency of ‘hurst’ as a suffix, suggests that this area was relatively well settled by the late Saxon period.
The possible Saxon settlement site of Hansfleote (‘an enclosure in, or near, a stream’) has been identified on the southern bank of the river Teise, to the west of the bypass route (Moon 1992). This may represent the site of the earliest settlement in Lamberhurst, which grew up around a river crossing point. At some point it moved west to its present location, possibly as a result of frequent flooding on the original site. It has been postulated that Bishop Godwin built the first church in Lamberhurst in the late 10th or early 11th century, possibly as a means of re- establishing control in the area (ibid.).
The parish of Lamberhurst lay on the county boundary between Kent and Sussex from the 10th century. A Perambulation Charter of 1077 describes the line of the boundary, which is now marked by Parsonage Lane (Moon 1992).
Medieval (1066-c.1550): the southern end of the route of the A21 bypass crosses the former deer park to the west of Scotney Castle (Pittman 2012). The documentary evidence suggests that there has been a park at Scotney since the early 14th century, when John de Grofhurst was granted Right of Free Warren in his lands at Scotney Curthope. This was granted by the Crown and permitted the keeping or hunting of beasts and fowls of the warren (Bannister 2001). The grant may have also allowed the right to enclose an area of land to keep deer and other game. At present the date of the construction of the park boundary is unknown. It may date either from the grant of free warren in the early 14th century, or from the early period of Darrell ownership in the 15th century. It was certainly in existence by the early 17th century, as shown by a map of Scotney Park, dated 1619, which shows the enclosed area (ibid.).
[pg134]Post-medieval/modern (c.1550-present): during the 17th century the enclosed area at Scotney was turned into farmland, possibly forming part of an estate called ‘Tuttyshams’ (Bannister, 2001). This period of disemparkment coincides with the Civil War and may indicate that the Darrell family, as Catholics, had their park destroyed by the Parliamentarians (ibid.). Alternatively, the huge demand for timber by the iron industry may have caused the clearance of the park, and its subsequent conversion to farmland.
The park appears to have remained either as farm or common land until the 19th century. In the 1830s Edward Hussey set about restructuring the park and gardens at Scotney on the principles of the ‘Picturesque’ style. This may have been initiated by the re-alignment of the Hastings Turnpike road (the modern A21) between Lamberhurst village and Kilndown in 1831.
Industrial activities were also associated with the estate in the post-medieval period, including brick-yards and a tile kiln, stone quarries, marl diggings and hop-growing. In addition, the Darrell family owned and ran a number of forges and furnaces in the area during the 16th and 17th centuries.
Phase 1 evaluation works
Eight areas were investigated as part of the Phase 1 works (Figs 2 and 3) in order to clarify the archaeological potential: of a number of features identified in an earlier stage of fieldwork (Barber, 1992), and to evaluate the archaeological potential in the location of a number of key constructional elements. The study areas included: Earthworks in Plots 7/8 (NGR 568315 137155) Earthworks in Plots 11 (NGR 568394 136886) Earthworks in Plots 13 (NGR 568422 136674) Earthworks in Plots 15/17 (NGR 568410 135484) River Teise diversion in Plots 15/17 (NGR 568536 136421) Pierce Barn overbridge in Plot 20 (NGR 568369 136109) Soilmarks in Plot 36 (NGR 568096 135530) Ceramic findspot in Plot 40 (NGR 568016 135435) Results Of the eight areas investigated, two yielded significant information:
Five boreholes were sunk along the line of the proposed new river diversion channel in Plots 15/17, using a window corer/gouge (Fig. 4). These extended to depths in excess of 3m. Only Borehole 4 contained any organic sediments of note, including macroscopic wood and twig remains. Further samples were therefore taken using a sleeved corer. The samples demonstrated the presence of a palaeochannel, cut on the inside edge of a curve in the river valley.
[fg]jpg|Fig. 2 Northern field plots.|Image[/fg]
[fg]jpg|Fig. 3 Southern field plots.|Image[/fg]
[fg]jpg|Fig. 4 Plan of palaeochannel adjacent to Borehole 4. River diversion in plots 15/17.|Image[/fg]
The evidence recovered from the cores indicated a woodland environment, dominated by lime/linden, with oak and hazel in the lower levels. Few grasses and herbs were evident, suggesting little evidence of human intervention in the earlier prehistoric period. A change to increasing sedimentation and pollen degradation, with an expansion of herbs and [pg135][pg136][pg137][pg138]grasses suggest a possible opening up of the woodland on the interfluves in the Neolithic/Bronze Age period.
The evaluation of the findspot of a sherd of medieval pottery in Plot 40, exposed during topsoil stripping demonstrated the presence of a medieval iron- working site on Spray Hill. The evaluation prompted a full open-area excavation of the site (see below).
Phase 2 investigation works Five areas were investigated as part of the Phase 2 works (Fig. 5), all forming part of the Scotney Park landscape The study areas were:
Scotney Park Pale (Plot 40) (NGR 567927 135381): the remains of the southern line of the park pale were investigated in Plot 40, to the north of Ruffetts Old Quarry, with the aim of obtaining evidence concerning the construction, size, and age of the boundary, a record of the ditch fills, and evidence of any fence or palisade associated with it. Extant medieval park boundaries are comparatively rare, having been largely destroyed by changing land usage.
[fg]jpg|Fig. 5 Cultural heritage features in Scotney Park. Scale 1:5000.|Image[/fg]
Two trenches were excavated across the line of the pale, at a distance of approximately 90m apart (Fig. 6). The line of the ditch was observed in both trenches. In the western trench it measured 1.10m in width and 1.30m in depth, with shallow concave sides and a slightly rounded base. The ditch section recorded in the eastern trench was significantly wider but shallower, measuring 2.88m in width and 0.47m in depth, with shallow concave sides and a rounded base. Ploughmarks were evident in the western trench, indicating that the area had been subject to deep ploughing in the past. A single fill was recorded in both excavated sections, each being of a similar nature, comprising a loose light brown sandy silt. No datable artefacts were recovered from either fill.
[fg]jpg|Fig. 6 Northern Park pale (EFF56) and Southern Park pale (EFF55).|Image[/fg]
A single posthole was also recorded in the western trench, adjacent to the northern edge of the ditch. This feature was irregular in shape, measuring in excess of 0.70m in length, 0.70m in width and 0.60m in depth, with steep, almost vertical, sides and a slightly rounded base. It contained a single fill and a clear postpipe. Brick and tile fragments were collected from the posthole fill. It is likely to form part of a modern fence following the line of the ditch after it had gone out of use. Although no evidence for a bank was observed within the confines of the trenches, a low bank was noted running east to west, along the southern edge of the boundary between Plots 40 and 43. This bank can be seen intermittently to the north of Ruffetts Old Quarry and survives to approximately 0.5m in height and 2m in width. It is likely that this length of the park pale was maintained as a field boundary long after the park was abandoned. In general, evidence for the pale survives well in areas of woodland, but elsewhere, as in Plot 40, cultivation has eroded the surface evidence.
Scotney Park Pale (Plots 27/28) (NGR 568320 135945): the remains of the northern line of the park pale were investigated in Plots 27/28, along the northern edge of Piercefield Shaw Quarry (Fig. 6), with the same aims as those for the southern line. Two trenches were excavated across the line of the boundary, at a distance of approximately 13m apart. No evidence for a bank or ditch was observed in [pg139][pg140][pg141]the eastern trench, suggesting that the boundary had been destroyed by the later quarrying. However, it survived in the western trench as a ditch, running on an east to west alignment. The feature measured 2.20m in width and 1.20m in depth at this point, with shallow, slightly concave, sides and a rounded base. It contained two fills, the upper fill consisting of an eroded topsoil, which had apparently been washed into the ditch, the lower consisting of a dark greyish blue clay. No datable artefacts were recovered from either fill.
A low bank, approximately 0.30m in height and 1.40m in width, ran immediately adjacent to, and parallel with, the ditch on its southern side. The bank was formed from a loamy dark grey silty clay, probably representing the upcast from the excavation of the ditch. The later erosion of the bank is likely to have formed the upper fill in the ditch. Both the bank and ditch were clearly visible prior to the machining of the western trench.
Ruffets Old Quarry (Plot 43) (NGR 567970 135332): two trenches were excavated through the Ruffets Old Quarry, one through the south-facing slope on the northern side of the quarry, and one through the quarry base, with the aim of investigating the quarry’s history and use (Fig. 7). Small-scale quarries have been little studied but offer the opportunity to understand a variety of industrial practices and associated processes, including the nature and extent of the geology, the methods of extraction, the extent and nature of on-site dressing or treatment, and the methods of handling the extracted material.
[fg]jpg|Fig. 7 Piercefield Shaw quarry (EFF 60) and Ruffets Old quarry (EFF57).|Image[/fg]
The underlying natural geology in both trenches contained frequent outcrops of fine-grained silty sandstone. The quarry appears to have been worked into a seam of larger sandstone blocks, which would be ideally suited to use as building material.
The investigation has provided little evidence concerning quarrying practices and associated processes and no archaeological artefacts were recovered. This may, in part, be due to the fact that significant disturbance to the quarry was caused by the advance site clearance works. The size of the quarry suggests small scale extraction for use in local construction. Documentary evidence shows that the quarry came into existence after 1840 (Bannister 2001), and its usage may coincide with the construction of the West Lodge in 1842.
Piercefield Shaw Quarry (Plot 28) (NGR 568310 135910): a total of six test pits were excavated here, located to coincide with features identified during an earlier archaeological survey (Babtie 2002). The aims of the investigation were largely the same as those defined for Ruffets Old Quarry.
The underlying natural geology observed in the test pits consisted of slightly silty clay. The quarry face lay on the north-east side of the study area, cutting into the natural upslope and it appears that the raw material was removed from the quarry via a narrow entrance in the southern slope. Piercefield Shaw is located right on the edge of an outcrop of Wadhurst Clay, close to its boundary with deposits of Tunbridge Wells Sand. The clay does not appear to have been rich in stone deposits and it would appear unlikely that this was the resource being exploited.
Test pits were located through earthworks in the base of the quarry, which consisted of raised mounds or banks (Fig. 7). The recorded sections suggest that [pg142][pg143]these features were the result of the deposition of waste material from the quarrying process.
The results of the investigation suggest that Piercefield Shaw was not a quarry for building stone or for the extraction of ironstone, but rather a clay pit. This interpretation is supported by the presence of brick and tile works at Scotney from the 18th century, and a number of former clay and marl pits in the vicinity named in the cartographic evidence.
Scotney Castle Park Drive (Plot 43): (NGR 567881 135248): three trenches were excavated across the western end of the park drive, to the south-east of West Lodge (Fig. 8), with the aim of investigating how it was built, and the nature and source of the materials used. Of particular interest was the question of whether material salvaged from the ruination of the Old Castle had been used in its construction.
[fg]jpg|Fig. 8 Scotney Castle Drive (EFF59).|Image[/fg]
A number of changes were made to the layout of the grounds at Scotney Park during the 19th century. Work started on the main drive from the new mansion in 1837 with the levelling of the route through the woods and with the planting of firs at the entrance. This ran through Colliers Wood to Lamberhurst Down, where it joined the Hastings Turnpike Road. It was designed to allow visitors to Scotney Park to experience the ‘Picturesque’ landscape as soon as they entered the park (Bannister 2001, 11). The West Lodge at the junction with Hastings Road was designed and built by Edward Hussey in 1842. The stone for this building may have come from the quarry at Ruffets.
The trenches demonstrated that the construction of the original drive had involved the excavation of a cutting through the existing topsoil and subsoil over part of its length, and the construction of an embankment elsewhere. The depth of the cutting decreased as the drive headed downhill towards West Lodge and the Hastings Road, in order to reduce its gradient relative to the slope. A series of levelling deposits of sand and clay were then used as make-up layers for the drive surface. Some of these deposits were derived from the excavation of the cutting, and some from the enlargement of the pond in Colliers Wood, which took place at about the same time as the drive was constructed (ibid.). Further deposits of clay were also used.
Evidence for the removal of trees during the original construction of the drive was encountered, but no worked stone, which might have derived from the ruination of the Old Castle, was observed during the investigation.
Phase 3 monitoring works
The third phase of works, the monitoring of all topsoil stripping for the road construction, identified no significant remains or deposits outside those investigated in previous phases of work. A number isolated features and finds scatters were recorded:
Ten unstratified flints were recovered along the length of the route, indicating a general background of prehistoric activity.
In Plot 13, a shallow sub-circular scoop, measuring 1m in diameter and 0.07m in depth, was investigated. A single abraded sherd of flint tempered pottery was recovered, suggesting a prehistoric date.
[pg144]The cutting of the new river channel confirmed the existence of an earlier channel of the river Teise, which measured 6m in width (see above).
In Plot 36, two undated postholes and the northern boundary of the plot were recorded.
Sherds of medieval pottery were recovered from the natural spring in Plot 37.
The effects of modern agricultural practices, particularly deep ploughing, were noted in all the areas investigated. Plough attrition is likely to have affected the potential survival of archaeological features and deposits along the whole road corridor.
A Medieval Iron-working site on Spray Hill discovered
The evaluation of the findspot of a sherd of medieval pottery in Plot 40, undertaken as part of the Phase 1 works, revealed the presence of a medieval iron-working site on the northern slope of Spray Hill, between 76m and 88m aod. The site is located to the north of Scotney Castle drive, centred on NGR 56803 13540 (Fig. 9). It lies in an area intended to be excavated for a cutting along the road corridor, the spoil providing infill material for an embankment across the valley of the river Teise, to the north.
[fg]jpg|Fig. 9 Spray Hill excavation ; site location.|Image[/fg]
Spray Hill sits in an area of Tunbridge Wells Sand. In the Lamberhurst area, the Tunbridge Wells Sand is in excess of 70m thick. It comprises fine grained silty sandstones with appreciable amounts of siltstone and stiff silty clay, underlain by thick bands of thinly laminated mudstone with some sand and siltstone. The main outcrop encountered at Spray Hill, is extensively faulted (Babtie 2001).
The site lies immediately south of the Kilndown fault line and approximately 200m north of the Hook Green fault line. Seasonal springs, indicative of perched water tables at geological interfaces, are located to the north-east of Spray Hill Farm, to the west of the present site (ibid.).
A fieldwalking survey of the road corridor by the Field Archaeology Unit, University College, London (Barber, 1992), produced no information about this plot as it was under pasture and had not been ploughed. During an initial site walkover by Heritage Network staff in 2003, an unabraded sherd of medieval pottery, initially dated to the 12-13th century, was discovered. In June 2003, this area was re-machined under close archaeological supervision using a JCB, and a further four evaluation trenches were excavated in the surrounding area.
The evaluation demonstrated the existence of a network of linear features. Iron working slag was identified in the fills of some of these features, but no further dating evidence was recovered. Cartographic evidence indicated that the site had formed part of the medieval deer park, but gave no indication of any field system or settlement that corresponded to the features recorded during the evaluation (Bannister 2001a).
An open area excavation of the extent of the site affected by the road construction was carried out between July and August 2003. The particular aim of the project, in the context of the paucity of archaeological data for the Lamberhurst area and the Weald in general, was to gather information which might have the potential to contribute to the following research topics: The change and development of land use throughout the medieval period; [pg146][147] The definition of medieval settlement patterns across the region; The nature and extent of industrial and craft production; The investigation of datable pottery assemblages and the establishment of regional pottery sequences.
Site Layout
During the 13th century, a series of small rectangular plots was established on the north-facing slope of Spray Hill, running on an approximate east to west alignment and enclosed by ditches. The alignment of ditch groups [40255], [40254], [40257] and [40256] demonstrate that these divisions follow the natural contours of the hill (Fig. 10).
[fg]jpg|Fig. 10 Spray Hill excavation: context locations.|Image[/fg]
The ditches divide the excavated area into three plots, designated A, B and C. Only a small fraction of the eastern area, Plot C, was revealed within the limits of the road corridor, and no archaeological features were recorded. A proportion of the western area, Plot A, was revealed, and a small series of gullies were recorded, suggesting that there was only limited occupation activity in this location.
Virtually the whole of Plot B, in the centre of the site, was investigated. This area covered approximately 2,480m[fn2], and formed the focus of activity on the site. The pottery assemblage indicates that this activity extended into the 14th century, but the limited nature of the assemblage, and lack of domestic rubbish recovered from the boundary ditches, suggests that the general level of occupation on the site was low. The pottery assemblage is discussed in detail at Appendix.
The quantity of iron slag recovered in the basal fills of the boundary ditches indicates that iron working was being undertaken while the ditches were open. The upper fills are indicative of natural silting following the abandonment of the site.
The site may have been supplied with water from a natural spring located approximately 50m to the north west of the centre of the site, and use may also have been made of water collected from the boundary ditches.
An east-west track through Colliers Wood, linking Lamberhurst Down and the Scotney Park Estate, appears to respect the orientation of the northern boundary ditch.
Evidence for iron working
The recorded evidence indicates that small-scale iron production was taking place using a bloomery furnace. There was no evidence of other metals being exploited on site.
Iron working appeared to be concentrated in three areas (Fig. 11). Area 1 was located outside the ditched enclosure, against the southern baulk of the excavation area to the south of Plot B. It comprised two features, pit [40214] and a ‘banjo’ shaped feature [40158], interpreted as a roasting oven. Area 2 was located in the south-eastern corner of the Plot B, approximately 20m to the north of Area 1, and comprised pit groups [40267] and [40266]. Area 3 was located in the centre of Plot B, approximately 40m to the north-east of Area 2, and consisted of pit group [40264].
[fg]jpg|Fig. 11 Spray Hill excavation: distribution of slag.|Image[/fg]
As all the features across the site were heavily truncated, with features getting [pg148][pg149][150]shallower the further down the hillside they survived, all of the elements of the original above ground structures, and some of the shallower cut features associated with the iron production have been lost. However, some evidence for the furnace structures and tap pits were preserved as a result of the cooling slag taking on their shape.
Area 1: the ‘banjo’ shaped oven [40158], which may have been used for roasting the ore before the smelting process, consisted of a sub-circular chamber with a linear flue pointing to the south-east. This feature measured 1.10m wide, 0.15m deep and 2.30m long. It contained two fills, (40159) and (40160), which displayed evidence of burning in situ. A large amount of slag and hammerscale was recovered from the fills. Adjacent to the oven was a large sub-circular pit [40214], which measured 5.5m wide, 0.50m deep and 6.5m long. A large amount of burnt ore was recovered from its fill, supporting the interpretation of the ‘banjo’ feature as an ore roasting oven.
Iron ore fragments were encountered on the site. It is likely that the ore was extracted elsewhere and brought to the site, which, importantly, had a ready supply of fuel for the ovens and furnaces close by. This fuel would have taken the form of charcoal, probably derived from oak trees growing in the adjacent Colliers Wood. Fragments of charcoal were observed trapped within the matrix of some of the slag, although they were too small and degraded to allow species identification.
Area 2: context groups [40267] & [40266] comprised four features. Group [40266] consisted of a trough-like linear pit [40010] / [40150] measuring 0.85m wide, 6m long and 0.16m deep, with concave sides and a flat base. The two fills (40151) and (40152) contained burnt clay and a significant amount of hammerscale and slag. This feature was on the same orientation as the flue of oven [40158] and may represent the remains of a furnace, surviving only as the flue.
Immediately to the west was group [40267], consisting of two pits, [40176] and [40153], which flanked an irregular shallow feature, [40155] / [40172]. This feature was shaped like a figure ‘8’ and measured approximately 4m wide, 6m long and 0.09-0.13m deep. No evidence for burning in situ was observed, though a significant amount of slag was recovered from it.
The two sub-circular pits [40176] and [40153] were located on either side of the southern end of feature [40155]. Both measured c.1.20m wide, 1.35m long and 0.40m deep, and had steep concave sides and a flat base. The fills were very mixed, containing small amounts of charcoal and slag.
Area 2 has been tentatively interpreted as the location of the primary smelting furnace. However, it is unclear whether either the trough pit [40266] or feature [40155] represent the vestiges of an actual furnace.
These features lay immediately to the north of feature [40187], which formed part of the southern boundary ditch [40254] and has been interpreted as a sump. As furnaces need to be kept dry while in operation, and it is imperative to stop water from coming into contact with hot slag, it is likely that the sump was intended to drain this area. The location of the sump appears to have been deliberately chosen, as it lay approximately 30m to the west of the lowest natural point along the southern ditch. Over 50%, in weight, of all the recovered slag from the site [pg151]came from this one feature, suggesting that it was being maintained while iron working was taking place.
Area 3: pit group [40264] comprised two pits, [40217] and [40145]. Feature [40217] consisted of a large sub-circular pit 1.6m wide, 0,27m deep and 3.7m long. Its two fills, (40218) and (40219), contained the highest density of slag, hammerscale and charcoal of any feature on the site (Fig. 12). This material appears to have been deposited during the iron-working process, rather than dumped as later backfill. The adjacent eastern end of ditch [40258] also contained significant amounts of similar material. This may be where unworked blooms were reheated and worked into ingots or simple products such as nails. Nevertheless, smithing appears to have been relatively small scale on this site.
[fg]jpg|Fig. 12 Spray Hill excavation: distribution of hammerscale.|Image[/fg]
Burning in situ was observed in a small pit [40145] measuring 0.90m wide, 0.32m long and 0.18m deep. This feature was truncated by the later ditch [40258]. It is unclear whether this area formed part of a structure or was directly related to the feature [40217].
Discussion
Relatively little is known about Wealden iron production in the medieval period. Although a number of documents relate to the purchase of iron, to bloomeries and to smiths, these references are only occasional. They are, however, concentrated in the 13th and 14th centuries, the period when the present site was active (Cleere & Crossley, 1995). As well as a demand for iron from local farms and manorial estates during these centuries there are records of large scale purchases of Wealden iron products for military needs, for example 30,000 horseshoes and 60,000 nails in 1254 (ibid.). In addition, the Crown was known to buy blooms to be worked by its own smiths.
Contemporary documents suggest that iron from Kent was not considered to be of as good a quality as the blooms produced in the Forest of Dean, or imports from the Continent. The slag recovered from Spray Hill suggests that the process used for extracting the iron from the ore was not very efficient and that the iron being produced here was not very pure. It would, nevertheless, have been acceptable for the manufacture of nails, horseshoes, arrowheads and ploughshares. There is no known documentary evidence of a bloomery/ smith on Spray Hill.
Domestic Occupation at Spray Hill
Two areas of possible domestic occupation were identified, represented by context group [40107] in Plot A, and groups [40258], [40259], [40261], [40263], [40264], and [40265] in Plot B. The vast majority of the pottery assemblage (80% by weight) was recovered from ditch features close to these two locations, suggesting a discrete pattern of disposal close to occupation areas (see Appendix).
Limited evidence for structures was recorded, possibly as a result of deep ploughing across the site, which may have destroyed structural evidence such as beam slots or postholes. Alternatively, the relatively short span of occupation may indicate that any dwellings were insubstantial, leaving little trace in the archaeological record.
[pg152][pg153]Group [40107] in Plot A comprised a series of small truncated linear features, with a maximum depth of 0.20m, and a single posthole (see Fig. 10). These were located against the north-western baulk and appeared to continue northwards, beyond the limit of the excavation area. Gullies [40123] and [40108] formed a right-angle, possibly representing one side of a beam-slot structure. The single posthole, cut [40121], was located immediately adjacent to the butt end of [40108]. A concentration of pottery sherds was recovered from the fills of these gullies, reflecting a similar discard pattern to that noted in Plot B.
Group [40265], to the north-east in Plot B, represents a second possible dwelling, consisting of a small sub-circular area terraced into the slope of the hill. This was surrounded by a network of small ditches, groups [40258], [40261], [40259] and [40260], apparently providing drainage. Two pits associated with iron working, group [40264], were sited on the north-eastern corner of this area.
Five postholes were also identified on the site, cuts [40161], [40163], [40167], [40169] and [40202]. These features orbit group [40265], but do not make a coherent pattern. It is likely that further postholes have been removed by later ploughing. They hint at the presence of a wooden structure sited within Plot B, but do not indicate its size or nature.
Two large sub-circular pits, group [40263], were located to the north of the terraced area, but within the space demarcated by the drainage ditches. These contained silty sandy fills and pottery sherds, suggesting that they may have been intended for domestic waste.
A cluster of five shallow sub-circular pits, [40185], [40204], [40206], [40209] and [40212], was recorded approximately 10m to the south-east of the proposed occupation area, in the south-east corner of Plot B. The layout of these features may suggest a small post-built structure. No artefacts were present within their fills.
Decline of the site
The pottery recovered from the site demonstrates that occupation did not continue beyond 1350 (Barber, this report). As the earliest evidence on the site dates from the first half of the 13th century, it would appear that activity on the site was limited to a timespan of around 100 years.
The abandonment of the site may have been the result of climate deterioration, which, from the end of the 13th century until 1350, was partly responsible for crop failures and livestock disease during the early half of the 14th century. As a result much marginal land became untenable (Platt, 1978). Iron production on Spray Hill may also have ceased as a result of the ‘Black Death’, which, from 1349, decimated the population and would have left a shortage of people to work in the industry and in the fields. Competition from producers of better quality iron elsewhere may also have contributed to the decline of the site.
Occupation on the site would have certainly ended during the 14th century when it was incorporated in the newly enclosed deer park. The southern boundary of the park runs across the brow of Spray Hill, immediately to the south of the site.[pg154]
Conclusions
The programme of archaeological investigations on the line of the A21 Lamberhurst Bypass had three principal aims:
to evaluate areas of previously defined archaeological potential in advance of the start of the construction works, and clear any which proved to be of significance; to investigate specific archaeological features forming part of the National Trust’s Scotney Park estate; to maintain a watching brief on all topsoil stripping operations in order to identify and record any unexpected archaeological features and deposits before they were destroyed.
The advance evaluation works recovered environmental evidence from the line of the new river diversion channel, which suggests that the region was heavily wooded in the earlier prehistoric period, and that there may have been some clearance of the woods on the higher ground in the Neolithic/Bronze Age. The works also identified a previously unknown iron-working site on the northern slope of Spray Hill. This appears to have been small in scale and was focused on the use of a bloomery furnace. The pottery recovered from the site in the course of its subsequent excavation, has been dated to the period 1225-1350, when the Wealden iron industry was flourishing.
The investigation of the Scotney Park pale has confirmed that the boundary consisted of a ditch and bank. This has been damaged in places by later agricultural and quarrying activity. The excavated ditch sections appear to demonstrate a uniform profile, consisting of slightly concave sides and a rounded base. This accords with other profiles recorded by the National Trust (NTSMR 142063) and suggests that the boundary was established as a single entity. The lack of datable artefacts from the ditch fills implies regular maintenance of the ditch in the course of its functional life.
The investigation of Ruffets Old Quarry suggested that it was used for the small- scale extraction of building stone but no evidence for quarrying practices or any associated processes was identified. Piercefield Shaw appears to have been a clay pit rather than a quarry. No archaeological evidence was recovered from either site to provide definitive dates for their exploitation.
The investigation of the Scotney Park drive demonstrated that a level road surface had been achieved by a combination of cutting and embanking, but no evidence for the use of demolition debris from Scotney Castle to build up the road bed was recovered.
Along the length of the bypass route there was a general lack of artefact evidence that might reflect occupation, suggesting that, at least since the 14th century, the land to the east of the village of Lamberhurst was largely agricultural or pastoral. No evidence for human activity pre-dating the later medieval period was recovered, indicating that during the prehistoric, Roman and Saxon periods this area may have been considered unsuitable for settlement.[pg155]
Acknowledgements
The present project was carried out by on behalf of May Gurney Construction by The Heritage Network Ltd under the direction of URS Corporation Ltd. The Heritage Network is pleased to acknowledge the advance archaeological investigation works undertaken by the Field Archaeology Unit of University College, London on behalf of Kent County Council, and by the Babtie Group on behalf of the Highways Agency.
The fieldwork for the various stages of the present project was undertaken by Gary Bishop, Tom Burns, David Kaye, Pauline Phillips, Paul Riccoborii, Geoff Saunders, Karin Semmelmann, Adam Single, Alice Thorn, and Mark Winter, under the supervision of Chris Turner. Finds processing, assessment and reporting was coordinated by Helen Ashworth. The illustrations for the report were prepared by Karin Semmelmann and Chris Turner, and the report was edited by David Hillelson. The Heritage Network would like to express its thanks to Peter Lingwood, CEC; Ron Henry, Sean O’Connell and Paul Kendall, URS Corporation; Nigel Steere, Matt Childs and Steve Bowen, May Gurney; Ian Greig, Archaeology South-East; Geoff Potter, Compass Archaeology; Reg Houghton and Dorothy Meads, Wealden Iron Research Group; and Casper Johnson and Wendy Rogers, Kent County Council Heritage Conservation Group, for their co-operation and assistance in the execution of the project.
Medieval pottery from Spray Hill by Luke Barber, Archaeology South-East
The excavations at Spray Hill recovered a total of 320 sherds of pottery, weighing 1,689g, from 23 individually numbered contexts. The pottery is generally in poor condition with small abraded sherds (average sherd size 5.3grams), frequently affected by acidic ground conditions, dominating the assemblage. No large context groups are present: the largest two being from [40149] (54 sherds weighing 592g) and [40184] (67 sherds weighing 192g), though many of these sherds are from single vessels in each case ([40149] contains 37 sherds from a single pitcher). Despite this, the assemblage as a whole is considered to be of some interest due to the scarcity of excavated pottery from this part of the Weald. As such, the aims of the pottery report have been to outline the range of fabrics present and to help with dating both individual features and the overall occupation based on the total assemblage.
The pottery was divided into fabric groups based on a visual examination of tempering, inclusions, decoration/finish and form. All the pottery was duly recorded by fabric (number and weight) for each context and each assemblage spot-dated. (This information is lodged with the archive.)
The material spans the mid 13th to mid 14th centuries and is represented by a number of fabric types typical of the Wealden area. Feature sherds are scarce in the assemblage and the majority are illustrated on Fig. 13.[pg156] [pg157]
The Fabrics
Although the fabrics belong to one general period it should be noted that there is some chronological progression and overlap between a number of them. Unfortunately the current site did not produce large enough context assemblages to analyze changing fabric ratios. (The percentages given below are for the number of sherds in whole site assemblage by sherd count.) F1: Moderate fine to medium sand with rare to sparse shell to 2mm (25.3%) Mainly cooking pots and bowls. Suggested date – 13th to early/mid 14th century. Illustrated vessels (Fig. 13):
[fg]jpg|Fig. 13 Spray Hill excavation: illustrated pottery|Image[/fg]
1 Small bowl. Mid grey core with dark grey/black surfaces. Externally sooted. (Context 40124)
2 Cooking pot. Mid grey core, buff margins and light brown/grey surfaces. (Context 40136)
3 Cooking pot. Mid grey core with mid/dark grey surfaces. Externally sooted. (Context 40136)
4 Bowl. Dark grey core, dull red margins and dark grey/black surfaces. Externally sooted. (Context 40190).
5 Cooking pot. Dark grey core, brown margins and dark grey surfaces. Externally sooted. (Context 40201).
F2: Moderate to abundant coarse sand (30.6%)
Only cooking pots noted. Suggested date – 13th to early 14th century.
F3: Moderate to abundant medium sand (33.1%)
Mainly cooking pots but some bowls and occasionally coarse jugs with sparse glaze. Suggested date – mid 13th to 14th century.
F4: Well-fired moderate fine to medium sandy blackware (3.8%)
Some sherds have the odd shell inclusion. Surfaces deliberately fired evenly dark grey/black. Forms – unglazed cooking pots and bowls. Probably Limpsfield products (Ketteringham 1989) though some sherds could be from Winchelsea (Barton 1979). Suggested date – later 13th to 14th century.
F5: Sparse to moderate fine to medium sand (3.4%)
Usually oxidised. Forms – jugs with thin external green glaze. Possibly Rye or Ringmer products. Suggested date – 13th to 14th century.
F6: Fine sand with iron oxide inclusions to 1mm (3.6%) Usually oxidised. Forms – cooking pots. Suggested date – mid/late 13th to 14th century. Illustrated vessel:
Cooking pot. Dull orange throughout with blackening on rim and splashes of glaze on the underside of rim. (Context 40149).
[pg158]The small size of the context assemblages makes the precise chronological ordering of individual groups impossible. However, a number of cross-joins between contexts are present, particularly for the central features within the main enclosure, suggesting much of the material may have been deposited over a relatively short space of time. Cross-joins include fragments from a F3 pitcher from Contexts [40149] (37 sherds weighing 390g), [40190] (18 sherds weighing 48g) and [40021] (six sherds weighing 20g). Sherds from the same F2 jug/pitcher were also noted in Contexts [40183], [40184] and [40190]).
The earliest pottery appears to relate to the middle of the 13th century when coarse sand and sand and shell appear to have made up the main coarseware fabrics. The absence of solely shell-tempered or flint-tempered wares strongly suggests the assemblage does not extend back to the 12th or very early 13th centuries. Flint tempered wares were also absent from Bayham Abbey (Streeten 1983) which was founded in the 13th century. Flint-tempered wares are known from a number of Wealden sites where they are generally given a 12th- to early 13th- century date (Barber forthcoming) and it is thus quite likely their absence indicates a start date after 1225 and perhaps as late as 1250.
The bulk of the medieval pottery, and thus presumably the most intense activity, relates to the mid/late 13th to mid 14th centuries. During this period the sand and shell F1 and coarse sand F2 proportionally decrease with the increase in medium sand tempering and fine sand tempered jugs (F5). During the 14th century the quality of finish and increase in firing temperature begin to increase but the wares are variable and defy close classification.
The source of most of this material is probably from local, as yet undiscovered, kilns in the Weald. However, jugs possible from Rye and Ringmer appear to be represented and F4 is very similar to both Limpsfield types (Ketteringham 1989) and ‘Winchelsea Black’ (Barton 1979, 118-20). This fabric, whether from Limpsfield or Winchelsea was well represented at the aisled hall at Salehurst (Gardiner, Jones and Martin 1991, 92). A typical range of vessels appears to be represented: the assemblage is dominated by cooking pots, though glazed, and frequently unglazed, jugs/pitchers are also represented. With the exception of a hollowed handle from a skillet (Context [40149]) no other forms were noted. Unsurprisingly, given the site’s isolated nature, there are no imported vessels in the assemblage. This lack of imports is not a reflection of the site’s status as very low quantities of imported material has been noted at Bayham and other sites (Streeten 1983, 99).
The latest material from Spray Hill appears to relate to the mid/later 14th century. None of the higher fired ‘ringing’ wares of the late 14th to 15th century are present suggesting the site was abandoned by this time.
Bibliography
Alcock, J., 2000, Life in Roman Britain, London: Batsford/English Heritage. Babtie, 2002, Lamberhurst Bypass, Archaeological Fieldwork, Babtie Group. Bannister, N., 2001, A21 Lamberhurst Bypass, Heritage Evaluation. National Trust.
Bannister, N., 2001a, Scotney Castle Estate, Archaeological and Historic Landscape Survey: Vol. I, Main Text, National Trust.
Bannister, N., 2001b, Scotney Castle Estate, Archaeological and Historic Landscape Survey: Vol II, Sites and Monuments Inventory. National Trust.
[pg159]Barber, L., 1992, An Archaeological Evaluation of the Proposed Route of the Lamberhurst Bypass, Kent, Institute of Archaeology, UCL.
Barber, L., forthcoming, ‘The Pottery’, in D. Meads, S. Broomfield and L. Barber, Excavations at Huggetts Farm, High Hurstwood, East Sussex.
Barlow, F., 1988, The Feudal Kingdom of England, 1042-1216, 4th edn, London: Longman. Barrett, J., 1980, ‘The Pottery of the Later Bronze Age in Lowland England’, PPS, 46,
Barton, K., 1979, Medieval Sussex Pottery, Chichester: Phillimore.
Burnham, B.C., 1996, ‘Celts and Romans: Towards a Romano-Celtic Society’, in, Green,
M. (ed.), The Celtic World, 121-141, London: Routledge.
Cleere, H. and Crossley, D., 1995, The Iron Industry of the Weald, 2nd edn, Cardiff: Merton Priory Press.
Cunliffe, B., 1978, Iron Age Communities in Britain, London: Routledge Kegan Paul. Darvill, T., 1987, Prehistoric Britain, London: Batsford.
Gardiner, M., Jones, G. and Martin, M., 1991, ‘The Excavation of a Medieval Aisled Hall at Park Farm, Salehurst, East Sussex’, Sussex Arch. Coll., 129, 81-97.
Ketteringham, L., 1989, ‘Two Medieval Pottery Kilns at Limpsfield Chart’, Surrey Arch. Coll., 79, 125-145.
Martin, D. and Martin, B., 2011-12, ‘An Archaeological Interpretive Survey of the Old Castle, Scotney: Parts I and II’, Archaeologia Cantiana, cxxxi and cxxxii.
Millett, M., 2000, Roman Britain, London: Batsford/English Heritage.
Money, J., 1960, ‘Excavations at High Rocks, Tunbridge Wells, 1954-1956’, Sussex Arch. Coll. 98, 173-221.
Moon, J., 1992, ‘Hansfleote: A Late Anglo-Saxon Settlement of the High Weald’, unpubl. report.
Muir, R., 2000, The New Reading the Landscape; Fieldwork in Landscape History, University of Exeter Press.
Percival, J., 1981, The Roman Villa, London: Book Club Associates.
Pittman, S., 2012, ‘Elizabethan and Jacobean Deer Parks in Kent’, Archaeologia Cantiana, cxxxii, pp. 53-82.
Platt, C., 1978, Medieval England, London: Book Club Associates.
Reynolds, P.J., 1996, ‘Rural Life and Farming’, in Green, M. (ed.), The Celtic World, 176- 209, London: Routledge.
Stace, C., 1991, New flora of the British Isles. CUP.
Streeten, A., 1983, Bayham Abbey, Sussex Arch. Society, Monograph 2.
Sumbler, M.G. (comp.), 1996, British Regional Geology: London and the Thames Valley, London: HMSO.
Wealden Iron Research Group, 2003, Experimental Bloomery Ironmaking [online]
Available from: http://users.argonet.co.uk/users/tonysing/WIRG/Expt/index.htm Williamson, T., 2002, The Transformation of Rural England; Farming and the Landscape 1700-1870, University of Exeter Press.
Wood, M., 1990, Domesday, a Search for the Roots of England, London: BBC Classics.[pg160]