The Impenetrable Forest? Pre-historic and Romano-British Settlement in the Weald, an account of some recent Fieldwork
INTRODUCTION
During the last five years, a programme of fieldwork has been undertaken across a transect of the Weald between Ulcombe and Frittenden. The underlying geology is principally a heavy clay in the floodplain of the River Beult. There are, however, a number of ridges created by outcrops of freshwater limestone and iron-bearing shale-like rock. This part of the county has previously received little archaeological attention. It has only proved possible to carry out this survey over an extended period through the goodwill and cooperation of a number of landowners. Ploughland has been walked, often repeatedly, in varying soil conditions. This has resulted in the recurring and recovery of a considerable quantity of archaeological material.
RESULTS OF FIELDWORK SURVEY.
Virtually every arable field produced prehistoric flint material, including low-lying areas that seemed less obvious sites for early activity. Two extensive flint working sites, both on southwest-facing slopes, may indicate more permanent settlement. The Bronze Age is represented by the find of a single palstave axe, (published in Arch Cant, Vol. CXV, 1995). Iron Age pottery has been found in five locations in the survey area. Two sites have provided evidence to date them to the Middle Iron Age, (6th century B.C.), the pottery is some of the earliest yet found in the Weald. From the spoil of a recently dug ditch in Headcorn came an assemblage of prehistoric sherds. Amongst these were a number with combed decoration found in association with fired clay and iron slag. Similar material had been found earlier at a site 800 meters away during the excavation of a pond. As well as pottery, this site contained at least one contemporary iron smelting furnace. This furnace has been dated from ceramic evidence to 500 B.C. and may possibly be some of the earliest evidence for iron working in the county. The process is thought to have arrived here from Continental Europe around 800 B.C. A further furnace, identical in form, was found during a fieldwalking survey 1 kilometer away in Ulcombe. It possibly dates from the 1st century B.C. When excavated, it was found that a small cremation cemetery had been sited above it fairly soon after its last use. Three of the burials had been inserted into the upper part of the furnace, and one of these contained a brooch identical to one found in a cremation at the Swarling Farm Belgic cemetery. This was investigated by J. P. Bushe-Fox in 1921 and also produced evidence for iron smelting in close proximity to the burials. The stratification there though had been somewhat disturbed by a later gravel quarry. Together with the 'Aylesford Cemetery', which was examined in 1886 by Arthur Evans, these cemeteries form the typesites for the 'Aylesford Swarling' culture. Three Romano-British sites have also been discovered in the parishes of Headcorn and Ulcombe, and their excavation has formed a major part of the survey. Two of the sites are in close proximity to the prehistoric iron working areas, and this suggests a continuity of occupation. Site 1 consisted of a rectangular structure, of timber construction, (Fig 1).
Excavation revealed the ground plan of the building in the form of postholes and sill beam outlines. It measured 9m X 4m, with an annexe and porch on the east side 6m X 3m. The internal floors were overlain with domestic debris, particularly the annexe area where the material included waste pieces of animal horn used for implement handles. Iron tools included knife blades and a sickle. An oven and a large domestic hearth, as well as evidence for internal room revision, were also found within this structure. Beneath this building lay a ditch containing Flavian pottery, late 1st century. A sherd of Samian ware had part of a Latin personal name scratched into its surface, and a large yellow and white glass bead from the porch area is of the Iron Age tradition. A second-century date for the building would though seem probable from ceramic and coinage evidence. Site 2 adjoined Site 1 and consisted of a sunken floored domestic structure, (Fig 2), dating to the mid to later 1st century A.D.
It was not possible to ascertain the length of the complete building as it had immediately been disturbed to the east. However, it measured 4m in width with an internal entrance ramp. The interior contained a well-preserved clay oven, a fragment of window glass, and a considerable quantity of pottery. Similar Flavian period ceramics link this building to the ditch found beneath Site 1. Comparison can be made with this structure and the sunken floored houses found recently at Monkton on the Isle of Thanet. During the drought conditions experienced during the early summer of 1997, some of the survey area was photographed from the air by helicopter. Two of the photographs indicated a possible early field system situated on the hillslope directly above the Romano-British sites 1 and 2. The features appeared as a series of interlocked sub-rectangular ditched enclosures in a field of wheat. During earlier field walking across this area, a number of abraded Iron Age and Romano-British sherds were found together with a fragment of a quernstone. The crop marks seemed to indicate that one of the field ditches coincided with that found earlier under the building at Site 1. It was sectioned by machine after the harvest and found to contain further Flavian pottery. The field system is therefore associated with the Romano-British sites and could perhaps contain elements of earlier enclosures. Waste industrial material in the form of iron slag is prevalent across both sites. Several probable hearths have been noted after ploughing. One has been excavated but consisted of little more than a circular patch of burnt clay with scattered pieces of furnace lining, together with charcoal and slag. A few pieces of 1st-2nd century A.D. pottery confirmed a Roman date; however, any structure had been removed by ploughing. The sites seem to have been primarily agricultural, with any iron smelting taking place on a fairly limited scale. On the assumed edge of the site was a cremation cemetery where three burials were excavated. A third Romano-British site was located to the south of Headcorn. Again, it lay on a south-facing slope and was found by fieldwalking. A trial trench, laid out across the zone of greatest pottery density, located an occupation horizon and a small ditch. There were slight traces of a possible timber structure, but the area has been subject to intensive agriculture, and no building plan was discernible. A large quantity of stratified pottery was found together with coins, nails, a knife blade, and a lead steelyard weight. However, the most surprising find came from the upper fill of the ditch. It is a decorated military belt plate dating from the early third century, (Fig 3).
This type of belt fitting has been found across the Roman Empire from Vindolanda to Dura Europos in Syria, usually on military sites. On either side are plates of yellow and black 'millefiori' inlay with a zoomorphic creature as the central bar. To the west of the site was a possible corn drying oven. When excavated, it was found to contain a number of fragments of quernstones together with a further lead steelyard weight. A substantial ditch was traced around two sides of the site by augering and selective excavation at several points. In one of the sections was a considerable deposit of burnt clay, iron slag, and cinder. Just beyond the southern boundary was a small cremation cemetery, where 3 burials were recovered. Preliminary analysis of the pottery from this site, (around 1,100 sherds), shows that some 77% is grog tempered ware, ie, 'Patchgrove' native type material. The remainder is made up of sandy wares, North Kent material and Samian. The site is probably another farmstead, there is nothing to suggest a military connection despite the presence of the belt plate.
Summary:- This fieldwork has provided evidence for significant pre-historic occupation in this part of the Weald. Iron smelting was taking place as early as the Middle Iron Age and continued through the 1st and 2nd centuries A.D., despite the lack of significant geological deposits of iron bearing rocks as found farther west in the High Weald. The Greensand ridge to the north of the survey area has been assumed to represent the southern limit of occupation by the Belgic tribes in Kent, the archaeological evidence though appears to argue against this assumption. The known limits of Belgic influence have therefore been pushed significantly farther south.The finding of three, probable agricultural, settlements dating from the mid 1st to early 3rd centuries A.D. is the first definite evidence for non-iron related exploitation of the Weald. Similar field systems with associated occupation sites have only been found at one other site in Kent at Highstead. These RomanoBritish sites may represent a continuance of settlement between the pre-conquest period and the post-conquest. Rural Settlements producing a preponderance of native type pottery wares appear to have been utilising rectangular timber dwellings from early on in the Roman occupation. Mixed agriculture was evidently possible even on the heavy clay soils of the Weald. There is a considerable amount of material from these sites to process and research before they are published in full, this account is therefore an interim report until then.
Neil Aldridge.