Investigations at a Prehistoric Romano-British and Early Medieval Site at Little New House Farm Headcorn Aldridge

INVESTIGATIONS AT A PREHISTORIC, ROMANO-BRITISH AND EARLY MEDIEVAL SITE AT LITTLE NEW HOUSE FARM, HEADCORN

neil aldridge

The archaeological site was first identified in 1994 and is located with-in the claylands in the parish of Headcorn in the Low Weald, an area also known as the Vale of Kent. Although Kent as a whole is very rich in archaeological remains this part of the county has received less archaeological study which has tended to be concentrated on the lighter soils of the Chalk and Greensand escarpments to the north-east. There archaeological field techniques such as air photography reveal sites more readily in the expansive arable fields rather than in the smaller enclosed fields and woodland areas of the Wealden zone.

The apparent lower density of early settlement in the Low Weald has previously been assumed to be due to its less inviting heavy clay soils although this viewpoint is now changing particularly with regard to the apparent prevalence of prehistoric and, in particular, Roman iron working sites revealed in fieldwork undertaken by the writer and others (Fig. 1). This has been taking place over the past 15 years in Headcorn and some surrounding parishes (Aldridge 1996; 2003; 2004; 2005a; 2005b; 2006) and especially by the longer transects taken by recent pipeline archaeology across the Weald (Network Archaeology 2001). The findings from this and other fieldwork is now being recognised as of significant regional importance and is modifying our perception of the occupation of the Weald in later prehistory and in the Romano-British period. A recent investigation demonstrated the possible association of Roman sites with later landscape boundary features in the parish of Benenden (Pollard and Aldridge 2008).

The fieldwork and excavations reported here were undertaken at Little New House Farm by the writer with the assistance of Mr W. Coomber, a son of the landowner, who had previously made the initial find of a Roman coin from the site. In earlier fieldwork on land at Little Poplar Farm some 4km to the north in the parish of Ulcombe (farmed by the same family) a significant Roman and early prehistoric landscape had been identified. A tract of arable land 1km in length and 0.5km wide had been extensively fieldwalked over a number of seasons. In work led by the writer a number of previously unknown Mesolithic, Iron Age, Romano-British and Medieval sites have been recorded. (This work will form a separate paper for a future volume of Archaeologia Cantiana.)

The site lies 1km due south of the parish church at Headcorn and some 260m to the south-west of the modern dwelling house at Little New House Farm (NGR 8313 4317) (Fig. 2). There is a view northwards from this relatively elevated site (30m aod) to the Greensand escarpment bordering the Weald proper some 6km distant. The land on which the site lies is equidistant between the flood plain of the River Beult to the north and the lesser Hammer Stream to the south. This part of the parish has deposits of a discontinuous bed of thin laminated sandstone occurring within the yellow clay which fractures readily into flat fragments. At the base of the hill to the south-west of the site are deposits of an iron-pan concretion which forms in the waterlogged soils of this area and is often referred to by farmers locally as ‘crowstone’ or ‘shrave’. This may well have been a source for the raw material needed for iron smelting on the site (see below) although it would be lower grade ore.

This hill at New House Farm and Little New House Farm, together with another somewhat further south are known together locally as the ‘Marl Banks’. There are, however, no obvious indications of any attempt to exploit clay marl here in the form of surviving marlpits. The field when under investigation from 1993 was being used as arable after being in use as an orchard for at least 50 years prior to that.

The Investigation

Attention was first drawn to the site after the chance find of a single Roman coin in January 1993 by a member of the Coomber family. It was after the coin was shown to the writer for identification that he became directly involved in setting up and directing further archaeological fieldwork and has been responsible for undertaking the investigation which extended over a period of some four years from 1993-1997.

After the finding of the coin a preliminary surface artefact collection across the southern slope of the hill revealed 1st/2nd century Romano-British pottery sherds, burnt clay, iron slag and charcoal in several locations. The first coin find had by this stage been identified as a sestertius of Marcus Aurelius, dating to ad 173-174 and fieldwalking produced a second coin this being of Trajan ad 97-117 together with a Roman lead steelyard weight from the same general area.

The subsequent fieldwork consisted of opening a series of small test trenches in areas with significant concentrations of occupation material located by fieldwalking in order to ascertain the date, nature and extent of the site. The use of a soil auger proved helpful and indicated where possible occupation levels were situated. The position of at least two buried ditches which appeared to extend across part of the site were also confirmed by sections dug later to prove their existence as well as to obtain evidence for dating purposes.

This report is limited in extent by the small-scale nature of the invest-igations and excavations which were necessarily restricted by the short windows between the crop cycles. A metal detector survey was carried out across the site at intervals and this produced the majority of the coin finds which had all suffered serious corrosion from the acidic local soil. It was also found that animal bone survival on the site was very poor and only cremated human bone could be recovered for useful research purposes.

Trench A A small test trench was opened in the area of the initial coin find (Fig. 3) and where there also appeared to be the greatest density of Roman occupation material; this revealed an intact, compacted Roman horizon averaging 75mm thick, lying at a depth of 0.30m below the topsoil with in-situ Roman domestic material. The trench was then enlarged which clearly indicated that an archaeological horizon with occupation material including pottery sherds, fired clay and burnt wood was present. A more substantial piece of fired clay with cut ‘stepped’ sides and grass or reed impressions was found and this may have formed part of an oven or kiln?

Trench B A second test trench extended from Trench A westwards (Fig. 3) which indicated that a compacted occupation layer continued at a depth of 0.28m consisting of a thin grey humic deposit containing charcoal and pottery sherds, pre-dominantly of a probably locally produced grog-tempered fabric now termed ‘East Sussex ware’ (similar to Patchgrove ware). It occurs on Iron Age and Roman sites in the Kent and Sussex Wealden zone; and was produced in East Sussex and probably within the Kentish Weald throughout the 1st-3rd centuries ad.

The extension to the excavation also revealed a gully 0.60m wide and 0.40m deep with a grey silty fill; it had been cut into a deposit of the thin bedded laminated sandstone which here lay barely 0.42m below the topsoil. The gully contained further occupation material, pottery sherds including some Samian ware, a flagon neck, fired clay, an iron nail and a possible iron hook. There were also a few fragments of Roman tegulae, charcoal, and very badly decayed animal bone (probably sheep or goat?).

At the surface of the assumed occupation area lying on the southern side of the gully were slight indications of a structure? in the form of two possible postholes together with a shallow horizontal slot. This suggests that possibly a Romano-British structure of timber had been sited close to the south side of the gully. There was a noticeably greater concentration of occupation debris in the westernmost section of the gully (which was not traced to its terminus).

A large area of the hill slope had in some places been subjected to considerable agricultural disturbance that included the wholesale grubbing of the mature trees of the fruit orchard in the early 1990s which with subsequent subsoiling meant that only partial archaeological horizons and cut features were likely to survive to any great degree.

A third coin, a worn unidentified Roman sestertius of 1st-3rd-century ad date was stratified in the upper fill of the gully. This together with other coin finds and the pottery, particularly the Samian sherds, all suggests a 2nd/3rd-century date for these features.

A significant find from the Romano-British period gully were the broken fragments of an elaborate enamelled Roman military belt plate of the 2nd/3rd cent ad thus providing additional site dating evidence (see below).

Trench C A ditch was located in the eastern part of the site by the use of the soil auger and was subsequently sectioned in a small trial excavation. It dated to the late Iron Age and appeared to extend south-west downslope from the other excavation trenches which lay some 12m north of Trench C although its whole extent was not fully investigated.

The ditch, where sectioned, was some 1.5m wide and 0.60m in depth with a grey silty fill containing Iron Age pottery in the primary fill and a spread of Roman pottery in the upper layers.

A complete Iron Age bowl (Plate I), was excavated from the base of the ‘Belgic’ period ditch at the point where the ditch appeared to change alignment near to the position of the Iron Age round-house located in Trench G (see below). It is possible that this vessel may have functioned as some form of votive offering.

Trench D Fieldwalking had identified another area which consisted of a spread of reddened soil and charcoal together with associated Roman pottery sherds some 18m west of Trench B. A trench 3 x 3m was excavated by hand to investigate this area further. A Roman horizon was identified 0.28m below the topsoil which consisted of a red burnt area with many fragments of fired clay together with an area of charcoal. The red burnt area was difficult to interpret as it had been disturbed by agricultural subsoiling to improve drainage. The fragments of fired clay may possibly represent the remains of an oven. It was uncertain whether this was part of a domestic structure or an industrial feature.

It was noted that some of the burnt clay pieces bore impressions of plant material in their surfaces. A few fragments of the burnt clay material also had circular pierced holes in them. The site was dated from sherds of Roman pottery, including Samian ware, together with a few fragments of Roman tile including two tegulae. There were also a number of fragments of quernstones associated with the burnt area. There were a few pieces of animal bone from the feature and some of the sherds of pottery had also been subjected to burning. There was no evidence of iron slag or other waste material from metal working in this part of the site or wasters from a pottery kiln. It appears most probable that the hearth may be part of another Romano-British building.

Trench E Another ditch was partially traced by augering and then sect-ioned by hand. This was initially thought to be contemporary with the Roman site. However, the pottery sherds found are entirely of Medieval date, 12th/13th-century. The ditch is also close to the findspot of a Medieval lead seal matrix found as part of the metal detector survey (Aldridge 1995b).

Trench F A number of Roman cremation burials were located after fieldwalking an area of ploughsoil some 74 m downslope to the south-west of the known Roman occupation area (Fig. 4). A number of conjoining pottery sherds associated with cremated human bone fragments were identified within the ploughsoil. A more comprehensive investigation was undertaken here with the opening of Trench F.

There were two cremation burials excavated from the trench with a third much disturbed cremation subsequently recorded, probably part of the same cemetery (see Appendix for further details and summary of specialist report). All of the site had been subjected to disturbance during the grubbing out of the orchard.

It is possible that the cemetery had been sited on the periphery of the settlement (which would have been normal Roman practice) though without more extensive investigation this remains uncertain. The Roman pottery spread in the ploughsoil did not appear to extend much beyond the area of the cemetery.

Trench G This trench, 14m in length and 2m wide, was excavated by machine to the north-east of Trench B to establish the extent of the Romano-British occupation area and to ascertain whether the ditch traced by the soil auger continued to the north of Trench B. It revealed an earlier prehistoric occupation layer with stratified Iron Age pottery at 0.30m below the topsoil. There was the apparent curvilinear edge of a shallow drip gully relating to a probable round-house extending across the line of the trench (Fig. 5). The fill of the drip gully? contained much Iron Age pottery together with carbonised wood fragments. The ‘interior’ of the round-house contained at least three postholes? with further Iron Age pottery within their fills. There were no finds of Romano-British pottery fabrics from any of these features suggesting that this was, indeed, an earlier structure. The features also appear to be sited within the Iron Age ditch traced earlier by augering although the ditch itself was not located.

Iron Smelting debris on southern slope of field

On the lower part of the hillslope below the occupation areas there was a considerable spread of iron smelting waste including much iron tap slag. It was not possible to identify any in-situ furnaces on the site, however; although there is no direct dating evidence it is most probable that the waste material is associated with either the Iron Age or Roman sites although a later date cannot be entirely ruled out.

The exploitation of the iron ore deposits of the Weald for smelting into iron was taking place long before the Roman invasion but it became a much more important industry after this though on present evidence the industry appears to have stopped around the mid third century ad. This cessation may have been the result of iron being obtained from elsewhere or such external influences as the decrease in military involvement.

Prehistoric Lithic material

Fieldwalking produced a quantity of lithic material both waste flakes and finished artefacts. This appears to be mainly Mesolithic and Bronze Age in date and indicates earlier prehistoric activity in the general area. Further evidence for a Bronze Age presence comes from the finding of a palstave which was recovered during the metal detector survey of the site during 1994 (Aldridge 1995a). The find spot was located at NGR 8318 4329 close to the crest of the hill. The palstave is 143mm in length, the blade edge is 0.55mm wide and the total weight is 238g. The implement dates to the Middle Bronze Age and is of the low flanged type.

finds

Enamelled Roman military belt plate. At a depth of 0.10m and stratified within the centre of the gully fill were found five pieces of a broken copper-alloy artefact. It was some while before this object could be identified owing to its fragmented condition. The five pieces were eventually seen to have formed a sub-rectangular openwork hollow cast copper alloy belt plate (Fig. 6a) with the following features:

- at either end pelta-shaped terminals with decayed brown and yellow millefiori enamel decoration

- rectangular side panels also decorated with similar surviving enamel decoration

- a central bar, 0.42 mm in length and the sides 0.53mm in length, consisting of two three-dimensional out-turned boars’ heads flanking an oval plate of blue champleve enamel.

When complete the original length of the plate would have been 0.85mm and the total original width 0.37mm.

There are other examples of similar military type belt plates known from across the Roman empire. Some of them have the same boar’s head central bars, as from South Shields, Newcastle upon Tyne, where a series of five inter-linked belt plates were discovered on the Arbeia fort site (Fig. 6b), whilst there are others such as were found at the Limes Roman military frontier site on the Danube at Carnuntum near Vienna with a central bar with four oval enamel roundels and the same side plates of millefiori enamelling. An example from the site of Dura-Europos in present day Syria is of the same design though missing the central bar. The majority of the other examples have all been found on sites with at least some military connection (Bidwell 1996; Allason-Jones and Miket 1984; Bishop and Coulston 1993).

Other examples can be cited from Kingsclere Hants, a metal detector find recorded as a Portable Antiquities Scheme Record; Dorchester, Dorset, a nineteenth-century find from the centre of the Roman settlement; and Vindolanda, Northumberland, from a frontier fort of the Severan period, 193-235 (Portable Antiquities Scheme 2000-1; Puttnam 1989; Birley 1977).

Copper alloy stud, probably Roman? This find was made by metal detector adjacent to Trench B within the ploughsoil. It is a small copper alloy stud 12mm in diameter with engraved? decoration on one face (Fig. 6c). Metal studs are found on many Roman sites and it is often not possible to tell what they may have been used for although this one may have been a fitting in a belt? In view of the find of the military belt plate this artefact may also have a similar military origin?

Coins A total of twenty-six Roman coins were found on the site the majority of which were located by one of the metal detector surveys made over the course of the fieldwork. The coins had all suffered serious corrosion from the acidic soil and the details of only four coins could be obtained:

- a sestertius of Marcus Aurelius, 173-174, found close to Trenches A and B.

- sestertius of Trajan, (Coin 2), 98-117, from this same part of the site

- another coin of Marcus Aurelius found in the area of Trench D

- one of Otacilia Severa, 244-249 (Coin 25), from the area of Trench D, the latest Roman coin found.

There were no coin types from the later Roman occupation.

The dates of the coins provides further weight to the theory that the majority of the Romano-British Wealden sites with iron-working evidence all appear to cease operation after the mid third century a striking pattern reflected elsewhere in the region.

Lead and Bronze Weights, Roman? A total of ten probable weights were found on the site:

- two small copper alloy weights found in the gully fill of Trench B; fieldwalking across the occupation zone produced a further eight examples

- a lead and iron Roman steelyard weight of typical Roman form was found in the ploughsoil close to Trench A with a further two small weights of lead in the same area.

- five small weights, all of lead, found around Trench D. This was the trench which contained a hearth feature of uncertain usage, though the weights may indicate subsequent other functions for this part of the site. A number of them certainly bore signs that they had possibly been folded around an object to seal it.

The number of weights recovered in the detector survey is quite significant and may point to a workshop or a place of trade?

Modern Finds

World War II 1944 – debris from a crashed aircraft. A number of pieces of spent ammunition from an American Mustang aircraft which had crashed and burnt out in the field on 12 May 1944 at 11.15hrs were also recovered during the metal detector survey. The aircraft was part of the US forces based at the nearby Lashenden Advanced Landing Ground at present-day Shenley Farm, Headcorn.

dating and general discussion

Although Kent as a whole is rich in archaeological remains the clay lands of the Low Weald have been relatively little studied and the findings from the fieldwork of 1993-1999 in Headcorn are of significant regional interest. Subsequent archaeological surveys and investigations, particularly around the southern margins of Ashford at sites such as the Romano-British settlement at Westhawk Farm and the Iron Age settlement at Brisley Farm in Kingsnorth, have shown that there was apparently more widespread occupation in the Weald at this time than has previously been thought.

Archaeological evaluations along the routes of two pipelines across the Weald (Network Archaeology 2001) have also helped to reinforce some of the findings and conclusions postulated by the writer as a result of the mid 1990s investigations at Headcorn. The majority of the sites found within the Low Weald had at least some connection with early iron working.

The main period of the occupation of the site at Little New House would appear to be the Romano-British with the peak from perhaps the early 2nd to mid 3rd century ad. This is evidenced from the few closely datable coin finds made during the excavations, along with the detector surveys and by the Roman pottery finds, especially the sometimes closely datable Samian ware. The majority of the pottery from the site is of ‘Patchgrove’ type native wares (which was also found to be the most prevalent at the Roman settlement at Westhawk Farm, Ashford), the usage of which continued in the Weald into the third century.

The small finds also provide dating particularly the possible military find of the enamel belt plate and the fibula from Cremation 2, the former dating from the mid 2nd century ad and the latter from the late 1st century. The finding of the belt plate in a sealed archaeological context may perhaps suggest some degree of military influence at the site, possibly connected with iron working? Certainly almost all the other similar finds in Britain and Europe have been made on Roman military sites. There have not been any local finds of Roman CLBR stamped-tiles from any nearer than Cranbrook, these being thought to suggest that this naval unit operated in the Wealden zone in order to supervise iron producing sites. Perhaps here we have an indication that the Roman Army was involved at similar sites on the northern fringes of the Weald? (It is worth noting in this context that the excavations at Ulcombe mentioned above revealed a Hod Hill type iron spearhead and a copper alloy phalerae.) Alternatively the Headcorn site may simply have been settled by a retired Army veteran?

The evidence of pre-Roman activity from a ditch and a probable Iron Age round-house confirms a continuance of occupation on the site from late prehistory. This may be further evidence for significant occupation in the Weald at an early period as found at other sites in Ulcombe. The resources of this area, notably the relatively low-grade iron deposits, may well have been the principal impetus to exploit the area utilising the existing knowledge of the Belgic tribes already occupying sites such as this. The excavations have provided clear evidence for a Roman presence in this part of the Low Weald. However, only larger scale fieldwork in the vicinity will provide answers to the many questions raised by it. There were only a few fragments of Roman tegulae found during the excavations yet it is unlikely that any structures here were tiled. Perhaps the tiles had originated from another Roman site close by as yet un-discovered? It is perhaps worth recording in this context that a Roman site (just 2 km south-east of Little New House Farm) was located as part of the Transco gas pipeline construction in Frittenden parish in 2001 at Ayleswade and this also produced some 26 pieces of Roman brick and tile (TQ 8366 4125, Transco Plot 49/5ii). In the discussion regarding this Roman material and the fact that it occurred with iron slag the excavation team concluded that ‘the site of a Roman building and the remains of an iron smelting area are likely to be close by’ (Network Archaeology 2001). Prior to nineteenth-century restoration Frittenden Church had Roman tiles built into the walls.

The contemporary Roman road access to the site is uncertain as no contemporary road was located during the fieldwork or subsequently. There is a possibility that some of the local droveroads which have been assumed to be dated to the Anglo-Saxon period may in fact be much earlier and perhaps sections of them may have formed trans-Weald communication crossed by the known main Roman roads in the area.

The nearest Roman highway is located 4.5km to the west at Staplehurst where the north-south Rochester-Bodiam road traverses the Weald. To the north the nearest road is the Sutton Valence-Hamstreet Roman road located 4.20km from the site. The more extensive second/third-century Roman settlement discovered at Little Poplar Farm, Ulcombe, by the writer (at around the same time as the Headcorn site) would have been sited within 250m of this road.

acknowledgements

Thanks are due to the members of the Coomber farming family of Headcorn and Ulcombe who first drew the writer’s attention to the site and permitted and encouraged further fieldwork on their land over an extended period enabling the site to be more fully investigated. The fieldwalking and excavations were largely undertaken by the writer together with Mr W. Coomber who rendered much assistance in often far from ideal conditions. The metal detector surveys were undertaken by Mr W. Coomber and the late Mr G. Buss.

The writer’s thanks are also due to the Fieldwork Committee of the KAS for providing a financial grant to enable a full report to be produced on the Roman cremation burials. This was undertaken by the late Trevor Anderson, osteo-archaeologist, and by J. Andrews, dental surgeon, also of Canterbury to whom my thanks are due.

APPENDIX

Description of the Roman Cremation Burials

Cremation Burial 1

Pottery sherds were noted lying on the ploughsoil after cultivation together with cremated human bone fragments. A small excavation revealed the outline of an in-situ Roman pottery burial vessel of soapy Patchgrove-type fabric sited within a contemporary burial pit (Fig. 4a).

The uppermost part of the vessel which was barely 0.28m below the topsoil had been damaged by earlier ploughing activity. However, the lowermost portion was found intact although the burial vessel had become cracked while still containing the remaining burial deposits. This burial was accompanied by a Greyware beaker on the north side and by a Greyware ‘dogbowl’ type dish on the western side. The main vessel contained a probable adult female burial (see report below), no grave goods were found in this burial. The pottery is probably of late 1st/2nd century ad date.

Cremation Burial 2

The excavation was extended farther to the west and revealed a second intact Roman cremation burial with a primary burial vessel of a Redware fabric which although intact had been damaged by soil pressure combined with the corrosive nature of the acidic soil. This burial together with Cremation 1 were both lifted intact within a block of soil and removed for more detailed examination off-site.

Burial 2 contained a single adult burial together with a bronze fibula dating to 80-100 ad of ‘Dolphin’ type. The Redware burial vessel probably dates from the 2nd century ad.

Cremation Burial 3

A disturbed third cremation burial was located from a number of sherds from a large grog-tempered vessel scattered in the ploughsoil. A small test trench produced a few fragments of cremated human bone, an iron nail, a beaker base and three pieces of a probable blue glass vessel. The pottery from this vessel is similar to that from Burial 1, native style Patchgrove type fabric.

Summary of Report by T. Anderson and J. Andrews on the Cremation Burials (A full report has been deposited in the KAS Library in Maidstone Museum.)

The two relatively intact cremations both contained single adults. Just less than half of the bones by weight could be identified although larger fragments were present in Cremation 2.

Cremation 1

Bone weight 620gm, burial was possibly female (pelvic morphology). Age is grown. The bone fragments were white in colour and this indicated that this cremation took place in a well oxygenated environment. Most of the material is highly fragmented. Exfoliation of the tooth crowns supports the view that the temperature was in excess of 500 degrees C. Pathology revealed osteophytes on one cervical and two lumbar vertebrae.

Cremation 2

Most of the bone fragments are white and highly fragmented. However the spine and pelvis are represented by larger fragments. Age, grown, vertebral end plates fused together with distal hand phalanx.

Other finds: iron nail, three fragments joining with bone fragments adhering.

[The bone remains associated with the burial 3 were too fragmentary to be included in the osteo-archaeological report.]

references

N. Aldridge, 1995a, ‘Bronze Age palstave axe from New House Farm Headcorn’, Archaeologia Cantiana, cxv, 460-462.

N. Aldridge, 1995b, ‘Medieval Lead Seal matrix from New House Farm Headcorn’, Archaeologia Cantiana, cxv, 462-3.

N. Aldridge, 1996, ‘Iron Working Sites in Kent’, Wealden Iron Research Group Bulletin, 2nd Series, No. 16.

N. Aldridge, 2003, ‘Wealden Archaeology’, Kent Arch. Society Newsletter, no. 58, 2-3.

N. Aldridge, 2004, ‘The Early History and Archaeology’, in Ulcombe’s Story - a Millenium History of Our Village, Ulcombe History Society, 17-22.

N. Aldridge, 2005a, ‘A Belgic Cremation Cemetery and Iron Bloomery Furnace at Jubilee Corner Ulcombe’, Archaeologia Cantiana, cxxv, 173-82.

N. Aldridge, 2005b, ‘Early Kent Iron Working Sites’, Wealden Iron Research Group Bulletin, 2nd Series, no. 25, 9- 19.

N. Aldridge, 2006, ‘The Roman Road from Sutton Valence to Ashford’, Archaeo-logia Cantiana, cxxvi, 171-83.

L. Allason-Jones, and R. Miket, 1984, Catalogue of Small Finds from South Shields Roman Fort, Soc. Antiquaries Newcastle, 3.10, 3.11, pl. vi.

T. Anderson and J. Andrews, 2003, ‘Report of the Romano-British Cremation burials from Little New House Farm Headcorn’, deposited in the KAS library.

P.T. Bidwell, 1996, The Roman Fort of Arbeia at South Shields, p. 25. Fig 18 depicts a more complete example consisting of four plates connected by a chain arrangement.

R. Birley, 1977, Vindolanda, a Roman Frontier Post of Hadrians Wall, plate VII.

M.C. Bishop and J.C.N. Coulston, 1993, Roman Military Equipment, Batsford, p. 152; fig 108.1 from Carnuntum Austria, and 4 from Dura-Europos, Syria.

Network Archaeology, 2001, ‘Archaeological Watching Brief, Transco Gas Pipeline Hadlow-High Halden, unpubl. report.

E. Pollard and N. Aldridge, 2008, ‘An Early Boundary, probably Anglo Saxon, associated with Roman Sites in Benenden’, Archaeologia Cantiana, cxxviii, 301-07.

Portable Antiquities Scheme, 2000-1, Roman Belt Plate, Kingsclere, Hampshire.

B. Puttnam, 1989, Dorset County Museum, Dorchester, Dorset. Catalogue No. 1884.9.137. Illustrated in Roman Dorset, 1989, Dovecote Press, p. 21.

Fig. 1 Location of Romano-British Sites in the Kentish Low Weald area. 1. Roman Road from Maidstone to Hamstreet (amended route). 2. Ulcombe/Jubilee Corner (discovered 1993). 3. Little New House Farm, Headcorn (1993). 4. Little Farningham (re-investigated 2000). 5. Benenden (2006). 6. Bodiam (Sussex), possible Roman port. 7. Smarden/Romden ironworking site. Dotted lines indicate probable Roman roads.

Fig. 2 Plan of Site and Cremation Cemetery; location of Trenches A-G. A/B. Roman occupation area and Roman Gully. C. Iron Age ditch. D. Possible site of Roman hearth/kiln. E. Medieval Ditch. F. Roman Cremation Cemetery. G. Iron Age Round-House.

Fig. 3 Plan of Trenches A, B, C, D, E, and G.

PLATE I

Belgic bowl from primary fill of Late Iron Age ditch at Little New House Farm, Headcorn. (100mm diameter, 97mm height.)

Fig. 4 Plan of Roman Cremation Cemetery (Trench F).

Fig. 4a Plan and Section Drawing of Roman Cremation No. 1.

Fig. 5 Plan of Features relating to Iron Age Round-House (Trench G).

A. Drip Gully. B, C, D Postholes.

Fig. 6a Roman Military Belt Plate.

Fig 6b A Roman belt from Arbeia (South Shields) (after Bidwell) to show arrangement of linked plates to form a cingulum. Its central bars have comparable zoomorphic boars’ heads as seen in the Headcorn example.

Fig. 6c Roman decorated copper alloy stud (possible military belt fitting?).