Excavations at Ringlemere farm Woodnesborough 2002-2006

39 Excavations at Ringlemere Farm, Woodnesborough, 2002-2006 keith parfitt and stuart needham M etal-detecting conducted by Cliff Bradshaw in a recently harvested potato field at Ringlemere Farm, near Sandwich, during November 2001 led to the discovery of an important prehistoric gold cup buried at a depth of about 0.40m (Plate I). The only British parallel for this vessel is Plate I The Ringlemere gold cup as discovered KEITH PARFITT AND STUA RT NEEDHAM 40 Fig. 1 Map of north-east Kent showing the location of Ringlemere in relation to the Wantsum Channel. EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 41 provided by the Rillaton gold cup, recovered from an Early Bronze Age cairn on Bodmin Moor in Cornwall during the nineteenth century (Smirke 1867; Needham 2000). The vessel from Ringlemere was discovered on a low, but quite distinct, rise in the middle of the field, which had every appearance of being a previously unknown round barrow. With only the Rillaton vessel in Britain and four or five more parallels in gold on the Continent, it quickly became clear that the Ringlemere cup was of both national and international importance. A Steering Group, consisting of a partnership of local and national archaeological organisations, was brought together to consider an appropriate response to the discovery and it was soon agreed that the find-spot required detailed investigation. Accordingly, an initial programme of field-walking, metal-detecting and geophysical survey, followed by excavation was organised (Parfitt 2003), and from this has evolved the Ringlemere Ancient Landscape Project, led by the Canterbury Arch-aeological Trust. The full excavation of the monument producing the gold cup was made a collaborative project between the Trust and the British Museum and excavations on this site were completed in the summer of 2006, with the support of the KAS, amongst other bodies. Ringlemere lies some 3.75km west of Sandwich, in the parish of Woodnesborough, about 1.5km west of the parish church, NGR TR 2938 5698 (Figs 1 and 2). Ringlemere Farm is some 400m to the south-east, with Black Pond Farm on Fleming Road 150m to the south-west (Fig. 2). The mound is situated at an elevation of between 10 and 13m above od and in subsequent fieldwork has been designated Monument (M)1. T he Gold Cup The image presented by the Ringlemere cup is striking, partly because of the obvious quality of the original workmanship, but also because of the severe crumpling it has recently suffered from agricultural machinery (Plate I). No physical restoration of the cup has been attempted because of concern that opening the severe buckles might alter the ancient metal structure. However, close inspection of the original, together with a virtual restoration by computer, has revealed the original shape of the vessel. By analogy with other Bronze Age cups made from precious materials and fresh consideration of their associations a date of around 1950-1700 bc may be suggested. As restored, the Ringlemere cup would have stood about 124mm tall (Needham et al. 2006; Fig. 3). The body of the vessel has all been raised out of a single piece of gold, the handle is a separate piece of sheet metal attached top and bottom by four rivets passing through tab extensions. Starting at the base, there is a small, neat omphalos [dimple]. The wall KEITH PARFITT AND STUA RT NEEDHAM 42 initially rises in a smooth gently convex profile, which is maintained for the ribbed part of the lower body. The wall is essentially vertical by the time it reaches a pronounced shoulder located almost two-thirds of the way up. Due to the subsequent damage, this shoulder is not obvious from any casual inspection. Above the shoulder, the upper body presents a strongly concave profile, sweeping out to a well flared rim. Most of the upper body is ribbed, but the uppermost band returns to a smooth metal profile. Immediately beneath the rim is a row of sixty-two dots punched from the outside of the vessel and interrupted only at the ribbed handle. Fig. 2 Plan of the Ringlemere site showing the location of the excavated henge-barrow (M1), adjacent (unexcavated) ring-ditches and other discoveries. EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 43 G eology and Topography of the Site The Ringlemere site lies at the foot of the North Downs dip slope, towards the bottom of a long north-east facing slope, which constitutes the southern side of the broad, shallow valley of the Durlock Stream (Figs 1 and 2). The underlying geology here is head brickearth, with some gravel, overlying Thanet Beds clay. Today, the Durlock Stream begins at a spring which rises in the immediate environs of the site and flows for about 8km westwards to join the Wingham River, which in turn empties into the Little Stour near Ickham (Fig. 1). A ridge of Eocene sands separates the Durlock valley from the south-western edge of the former Wantsum Channel, which once divided the Isle of Thanet from the Kentish mainland and was a much used water-way in ancient times (Fig. 1). Ringlemere lies just over 4 km from the Wantsum shore, which seems significant in terms of the Continental connections of Early Bronze Age items from the site – the gold cup and also two pieces of worked amber. The Wingham River may once have formed a broad inlet opening off the main Wantsum Channel and its lower reaches might have been usable by ancient vessels with shallow draught, perhaps similar to the Bronze Age boat found at Dover, just 16km to the south of Ringlemere (Clark 2004; Fig. 1, inset). Fig. 3 Reconstruction drawing of the Ringlemere gold cup. KEITH PARFITT AND STUA RT NEEDHAM 44 Field-walking, Metal-detector and Geophysical Surveys, 2002-2004 Initial field-walking of the site was undertaken in 2002. This involved detailed surface artefact collection over an area centred upon Bradshaw’s mound (M1). All material of archaeological significance was collected and bagged by individual 5 metre squares. Prehistoric calcined flint and worked flint was found to be spread across the entire survey area with noticeable concentrations being plotted around the mound. Other finds included occasional Roman, medieval and post-medieval pottery and some fragments of Roman tile but there were no significant concentrations of these. In 2003 and 2004 the survey area was extended across the valley, using a slightly less intensive search pattern based on 30 metre squares. By Easter 2004 more than 130 such squares had been surveyed. It is now apparent that an unbroken scatter of prehistoric struck flint and calcined flint is present across the entire area examined, although its density is generally less than recorded in the area of M1. Gridded metal-detector surveys have also been undertaken. Typically for the region, these yielded a light scatter of late Roman coins, together with other artefacts of Roman, Anglo-Saxon, medieval and post-medieval date. There are also two important prehistoric finds. At a point about 150m to the north-east of M1, a rare cast bronze brooch ���of 􀀫allstattD2􀀒Hallstatt D2/ D3 type was discovered �����������������������������������������������(Fig. 2; Parfitt2005).Thetypeiswellknown(Parfitt 2005)�������������������������.Thetypeiswellknown. The type is well known on the Continent but there are very few close parallels from Britain; it is almost certainly an import. About 150m to �������������������������thesouthofM1anumberthe south of M1 ���������anumbera number of bronze fragments may be derived from a Late Bronze Age hoard or metalworking area (Fig. 2). Geophysical survey has suggested that M1 is, in fact, the focal point of a more extensive prehistoric ceremonial landscape, now effectively invisible on the surface due to centuries of plough erosion (Fig. 2). An initial survey, covering some 1.4ha around the cup’s find-spot, was undertaken by staff from English Heritage’s Centre for Archaeology. This revealed the approximate outline of monument M1, together with two smaller ring-ditches (M2 and M3) situated immediately to the south-west (Martin 2003). In 2003 further resistivity surveys were undertaken by Aaron Birchenough in adjoining areas (Birchenough 2004). This produced more significant information and at least six more ring-ditches have been identified (Fig. 2). These vary considerably in both size and morphology; three of them have also been recognised on aerial photographs of the area. The Excavation of Monument 1 Geophysical survey and excavation have now confirmed that the low mound initially identified by Bradshaw is a man-made structure of EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 45 prehistoric date. Following the initial survey work in 2002, it was clear that the only prospect of establishing a useful context for the Ringlemere gold cup would be to excavate a sizable area around the find-spot. Accordingly, English Heritage provided funds for the excavation of an area measuring 10 by 30m on the north-western side of mound M1 (Trench 1; Parfitt 2003). This work confirmed the presence of surviving mound material, encircled by a substantial ditch. The mound sealed an earlier soil profile and cut-features, associated with large quantities of struck flint, calcined flint and Late Neolithic Grooved Ware pottery. A wealth of data and artefacts was recovered which demonstrated that further work on the site would be of considerable value, even though the precise context of the gold cup was still unclear. Since cultivation was continuing to erode the mound, with the tines of the sub-soiler cutting down into the pre-mound land surface, total excavation of the surviving monument seemed highly desirable. However, the large area of the monument, the survival of some stratified deposits and the scale of the ditch meant that, with limited funding, the work would have to be spread over a number of seasons. Eight separate excavations were undertaken on the monument between the spring of 2002 and the summer of 2006 (Needham et al. 2006; Trenches 1-8; Plate II), with a total of Excavations in 2005 Plate IIii KEITH PARFITT AND STUA RT NEEDHAM 46 463 days of fieldwork recorded in the site diaries. All the vulnerable prehistoric levels of M1 were investigated and virtually the whole of the interior plan of the enclosure has been revealed. Just over two-thirds of the encircling ditch was emptied; this includes exposing an entranceway on the north side. Five deeply buried sections of the enclosure ditch are not imminently threatened by continued ploughing and have been left for future generations to investigate. Mesolithic Activity A small proportion of the large quantity of prehistoric flintwork recovered from Ringlemere may be dated to the Mesolithic period, although no associated features or implement concentrations have been identified. The presence of fresh running water, in the form of the nearby Durlock Stream, would no doubt have made the area attractive to Mesolithic people. Finds of Mesolithic date are not at all common in north-east Kent and are largely confined to isolated surface discoveries of axes and adzes. The only significant local site is that at Lower Farm, Finglesham, located some 5.5km south-east of Ringlemere (Parfitt and Halliwell 1983). The Late Neolithic Henge Monument Sometime around the middle of the third millennium bc a circular ditched henge enclosure was constructed close to the source of the Durlock Stream. The enclosure was 41.5m in diameter internally and 50m externally. The ditch itself survived between 4 and 5m across and was up to 2m deep, with a broad, flat bottom (Fig. 4). Analysis of laminar sediments in the base of the ditch has shown that they were water-laid, implying that the ditch had held water, at least in wetter seasons. This is likely to be a reflection of the nature of the clayey subsoil rather than any specific design feature, however. The higher ditch fills in some excavated segments show that more material was slipping in from the outside than the inside, suggesting the former presence of an external bank, of which no trace has survived later ploughing. The bank was evidently made from the spoil from the ditch, for it had a proportion of gravel derived from the natural bands of such material that are sealed under the brickearth. None of the deposits filling the ditch were rich in artefacts and many of the layers were sterile. On the south side, however, two floors of deliberately laid flint cobbles were found lying on the base of the ditch. These included a small amount of humanly struck flint but, overall, the date of the construction and filling of the ditch remains to be ascertained. On the northern side, a deliberate break in the circuit formed an entrance causeway about 2.50m across. A pit discovered here must once have held a large post, almost centrally placed in the gap between the ditch EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 47 terminals, serving either as a blocking structure or to constrict entry to the monument (Fig. 4). A significant number of features occurred within the enclosed area, including variously sized pits, post-holes, stake-holes and hearths (Fig. 4). At the centre lay a small rectangular timber structure, formed of two L-shaped slots which originally must have contained timber up-rights (Plate III). This structure perhaps represents a shrine or ‘cove’, measuring c.2.4 x 1.2m in plan. The main axis was n-s and it was apparently originally open to the west. A central gap in the east side could have offered a narrow window facing east – in effect a portal. The surviving depth of the post-emplacements suggests that this structure could have been about 2m high. The central cove was surrounded by more than 180 other features (Fig. 4). These were most numerous on the south-western side of the Fig. 4 Plan of pre-barrow features associated with the henge monument. KEITH PARFITT AND STUA RT NEEDHAM 48 enclosed area, where the hearths are located. Indeed, pits and post-holes seemed to be clustered around the hearths. An arc of post-holes, possibly relating to a large building, occurred in the area of the south-east hearth. All of this implies that there had once been considerable activity within the enclosure. The excavated features are regularly associated with large quantities of flintwork and Late Neolithic Grooved Ware pottery. There is also some Beaker ware, including three complete vessels found in pits on the eastern side of the enclosure. The particular contents of several pits suggest that they include special ‘placed’ deposits. Charcoal from a pit containing Grooved Ware pottery near the north entrance gave a radio-carbon date of 2890-2600 cal bc (2 sigma; Beta-183862). It seems possible that some of the other features excavated may be connected with an earlier, pre-enclosure Neolithic settlement on the same site but further analysis of the recorded information is needed before this can be fully determined. Despite the large quantities of lithic and ceramic material recovered, no corresponding assemblages of prehistoric faunal remains had survived due to the slightly acidic nature of the soil. Considering all the excavation evidence, the original monument may be reconstructed as a penannular enclosure, with substantial ditch and an external bank, encircling a level interior of almost 42m diameter. A single entrance points roughly northward and this alignment is respected by the central timber cove. From the overall form of the monument and The central cove of the henge monument Plate IIIiii EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 49 the large quantity of Grooved Ware pottery recovered, there can be no doubt that the site represents a henge of Class 1 (single entrance) type (Atkinson et al. 1951; Harding and Lee 1987), apparently the first to be fully excavated in Kent (Needham et al. 2006). The Bronze Age Barrow A round 2000 bc a mound was raised in the middle of the old henge enclosure, burying all the earlier features. Exceptionally for the heavily ploughed landscape of east Kent, the base of this barrow mound survived at Ringlemere, with a maximum thickness of 0.50m remaining at the centre (Fig. 5). A core of soft, decayed turf was enclosed by an outer deposit Fig. 5 Plan of the barrow mound showing the associated façade trench (F. 1027) and the later pit (F. 1024) which probably held the gold cup (remnant mound shaded). KEITH PARFITT AND STUA RT NEEDHAM 50 of orange-brown clay. The turf core contained much residual domestic rubbish in the form of struck flint, calcined flint and broken pottery (mostly Grooved Ware with small amounts of Beaker), clearly derived from earlier activity on the site (as described above). The outer mound appears to have originally extended as far as the lip of the enclosing ditch but it had been partially cut away by later terracing in most areas (see below). Within the make-up of the outer mound, an absence of material derived from the distinctive lower green clay and gravel deposits through which the ditch was cut implies that the material of the mound did not include up-cast from the ditch. Given the character of the mound, it initially seemed likely that Early Bronze Age burials would be associated. This prospect was encouraged by the occurrence of the gold cup, since some of the parallels are from graves. However, complete excavation of the barrow has failed to locate any formal prehistoric burials. Rather than a burial mound it now seems more likely that the barrow was created as a low platform to support a new timber structure replacing the original cove. This was represented by a 3.68m long trench dug through the centre of the turf core, close to the site of the earlier structure (Fig. 5, F. 1027). It probably held a timber façade, again respecting the n-s axis. Later, a large irregular pit (Fig. 5, F. 1024) was dug into the top of the turf platform nearby. The western side of this pit clipped the façade bedding trench, but it is possible that the façade posts were still standing. Indeed, it is possible that the façade and pit were functionally linked, one dug quickly after the other was erected. The overall dimensions of the pit, at about 3.30 by 1.45m, are rather large for a grave and, in the absence of any skeletal material, other ritual functions seem more likely (see below). There is no reason to think that the new pit was intended to contain a replacement façade, although a thin horizontal layer of decayed wood was recorded in its filling. Importantly, this produced part of a rare Early Bronze Age amber pendant, which finds its best parallels in Brittany. Towards the south end of F. 1024 was a small intrusion, a pit of about 0.30m diameter. It was filled with loose soil and some decayed vegetation and appeared to be of very recent origin. We now believe that this could have been Bradshaw’s original excavation to unearth the gold cup. If so, the find-spot is placed 1.50m to the south-west of the estimated centre of the monument and some 8m from Bradshaw’s stated find-spot. By the time the gold cup came to be buried the monument had perhaps already been in use (possibly intermittently) for around 500 years. Yet, not long after, the site seems to have been abandoned, leaving the mound with its encircling ditch as a monument to the Ancestors, largely ignored by later inhabitants of the region. The burial of the cup, together with the amber pendant fragment, may represent a final act before the site changed in function to become a more passive focal point in a growing EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 51 monument complex. Possibly, the vessel had previously been used in rites and ceremonies performed around the façade on the platform. Sometime after the old henge enclosure ditch had become completely filled, ploughing extended across it, eventually cutting into the tail of the barrow mound to produce a terrace or negative lynchet on all sides except the north. The terrace was found to be particularly well developed along the north-eastern (downhill) side before swinging away, perhaps to avoid an area of natural gravel outcropping on the north side of the monument. The terracing must have been produced through centuries of ploughing, which probably began during the late prehistoric period. Amongst the first features to be removed would have been any outer bank surrounding the monument. The terrace, itself, subsequently became filled with down-washed soil, presumably partially derived from the ancient mound. The Early Anglo-Saxon Cemetery Before excavation work commenced at Ringlemere, the possibility had been considered that the ancient mound might have served as the focus for a subsequent Anglo-Saxon ‘flat’ cemetery because such post-Roman re-use of prehistoric burial sites is now becoming an increasingly familiar pattern in east Kent and beyond (Parfitt and Brugmann 1997, 4). Nor did it seem to be pure chance that the Ringlemere site was overlooked from the north-east by the important sixth-century Anglo-Saxon burial site at Coombe, some 750m away (Davidson and Webster 1967). Excavations on the south side of the mound in 2004 (Trench 5) led to the discovery of thirteen burials of Anglo-Saxon date. Trench 6, in 2005, revealed another, much larger group of graves, tightly packed around the south-west and west side of the mound (Fig. 6). Virtually all the graves had been inserted into the cultivation terrace which surrounded the barrow. This was already in existence and partially infilled with down-washed soil when the first burials were made. Over fifty burials have now been excavated in all, with every indication that more lie beyond the investigated area. Most are inhumations but there are also a dozen cremations – a burial rite rarely encountered in east Kent during the early Anglo-Saxon period. A small oval pit with heavily burnt sides located adjacent to cremation Grave 2 may represent the base of an associated cremation pyre flue, although no datable evidence was recovered from the feature, which had been partially disturbed by burrowing animals. Preliminary inspection of the cremation urns and grave goods associated with the inhumations indicates that many of the burials are of fifth-century date. Objects recovered from the inhumations include fine glass vessels, beads, brooches, silver rings and pins, buckles and various iron objects, including knives. Some exceptionally important early burials appear to be KEITH PARFITT AND STUA RT NEEDHAM 52 Fig. 6 Plan of the early Anglo-Saxon cemetery around the barrow, sunken-floored building and later Medieval features (remnant mound shaded). represented at Ringlemere, although no weapon-graves were discovered. The graves and their contents will require detailed study but it is already clear that this new cemetery site will be of considerable significance to Kentish Anglo-Saxon studies and more widely. Cut into the north-western side of the mound, a sunken hut of well-known Anglo-Saxon type was discovered in 2003 (Fig. 6, F. 1333). The abandoned hut-pit was probably in-filled during the seventh century and the building presently appears to be an isolated structure, unrelated to EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 53 the cemetery. Evidence for Anglo-Saxon rural settlement in east Kent remains remarkably scarce, making the new Ringlemere find of some particular interest. Medieval Activity Following the abandonment of the Anglo-Saxon cemetery, cultivation of the terraced area around the ancient barrow resumed, which seems to have led to the truncation of a number of the grave cuts. On the north-eastern side of the surviving barrow mound, a medieval gully (Fig. 6, F. 2302) was found to run across the base of the terrace, to delimit a rectangular enclosure around the remnant mound, presumably as part of a more extensive field-system. T hroughout the historic period, generations of local rabbits, foxes and badgers had found that the barrow mound provided an ideal place to dig their burrows and numerous examples of these have been located, cutting through the prehistoric levels and the Anglo-Saxon graves. Indeed, the site could have been maintained as a formal warren during the medieval period. Excavation of the burrows produced more archaeological material. The most important find was a second piece of prehistoric amber, part of a pommel from a small Early Bronze Age knife. Its original context remains a mystery but the piece indicates the former presence of at least one more prestigious prehistoric item on the site. conclusion Eight excavations at Ringlemere, spread over a period of five years (2002-2006) have enabled us to save from plough destruction one of very few complete henge-interior plans. Moreover, the structural evidence is accompanied by considerable quantities of Grooved Ware cultural debris and lesser amounts of earlier and later material (including Beaker and the Early Bronze Age gold and amber artefacts) tied to a fascinating sequence of monument construction and modification, occupation and religious observance. The post-excavation phase promises rich insights into this strategically crucial but generally poorly researched region of southern Britain for the later prehistoric period. The discovery of a sunken hut and a previously unknown Anglo-Saxon cemetery, containing some remarkably rich and chronologically early graves, including cremations, represents an added bonus, extending the archaeological significance of this site well into the early historic period. acknowledgements The Ringlemere project serves as a fine example of what can be achieved KEITH PARFITT AND STUA RT NEEDHAM 54 by professional and amateur archaeologists working in close co-operation with responsible metal-detectorists and local residents. Thanks are due first and foremost to the landowners – Andrew, Robert, Christopher, Jeremy and other members of the Smith family at Ringlemere Farm – who readily allowed access to their ground and have taken a keen interest in the progress of the work from the first. The excavations have been funded by substantial grants from the Kent Archaeological Society, the British Museum, the British Museum Friends (Townley Group), the British Academy, English Heritage and the BBC, together with a donation from Cliff Bradshaw, the finder of the site. A t English Heritage, David Miles, Peter Kendall and Sarah Jennings provided assistance and support throughout the early stages of the project. Thanks are also due to the staff at Kent County Council’s Heritage Conserv-ation Group, particularly Simon Mason. Roger Bland and Michael Lewis of the Portable Antiquities Scheme attended to the reporting requirements of the gold cup under the terms of the Treasure Act and have given much support and encouragement to the subsequent fieldwork. Unpaid volunteers have conducted the bulk of the excavation work, led by a small number of full-time, salaried staff from the Canterbury Archaeological Trust. Of the full-time team, Grant Shand, together with Enid Allison and Richard Helm have made substantial contributions to the project, whilst Barry Corke has acted as site records officer, surveyor and draughtsman throughout. CAT’s Director, Paul Bennett has given much support, guidance and practical help. Volunteers from various local archaeological societies, including those of Lenham, Maidstone, Otford and Thanet, and most particularly the Dover Archaeological Group, have made a valuable contribution to the work, together with a number of other KAS members and local helpers. Of the individual volunteers, Don and Tassa McGregor, Leslie Smith, Tina Parfitt, David Holman, Richard Hoskins, Geoff Halliwell and John Smythe deserve a special mention for their long-term commitment to the project, whilst Cliff Bradshaw tirelessly worked on his barrow between 2002 and 2004, undertaking both metal-detecting and excavation as required. Geoff Halliwell kindly funded some initial radio-carbon dates from the site. At the British Museum, the excavation campaign would not have been possible without the wholehearted support of Leslie Webster, Keeper of Prehistory and Europe, the BMbm’s Research Board and the Townley Group. Valuable specialist assistance and advice has also come from Sonja Marzinzik (Anglo-Saxon curator), Saul Peckham (photographer), Tony Simpson (scientific technician) while several members of the Conservation Department are involved in the painstaking removal of fragile Anglo-Saxon artefacts from lifted soil blocks, notably Denise Ling, Fleur Shearman and Duygu Cleere. EXCAVATIONSxcavations at RINGL EMERE FARM, WOODN ESBO ROUGH , 2002-2006 55 bibliography Atkinson, R.J.C., Piggott, C.M. and Sandars, N.K., 1951, Excavations at Dorchester, Oxon., Oxford, Ashmolean Museum. B irchenough, A., 2004, One man and a Magnetometer: a geophysical investigation into the extent and morphology of monument construction at Ringlemere Farm, Woodnesborough, Kent (undergraduate dissertation, Bournemouth University, April 2004). Clark, P. (ed.), 2004, The Dover Bronze Age Boat (English Heritage, Swindon). Davidson, H.R.E. and Webster, L., 1967, ‘The Anglo-Saxon burial at Coombe (Woodnesborough), Kent’, Medieval Archaeology, 11, 1-41. H arding, A.F. and Lee, G.E., 1987, Henge Monuments and Related Sites of Great Britain, BAR (British Series), 175 (Oxford). M artin, L., 2003, Woodnesborough, Kent, Report on Geophysical Surveys January 2002 (English Heritage Centre for Archaeology Report 33/2003). Needham, S.P., 2000, ‘The development of embossed goldwork in Bronze Age Europe’, Antiquaries Journal, 80, 27-65. Needham, S., Parfitt, K. and Varndell, G. (eds), 2006, The Ringlemere Cup: Precious Cups and the Beginning of the Channel Bronze Age (British Museum Research Publication 163, British Museum Press). Parfitt, K., 2003, ‘Bronze Age Discoveries at Ringlemere Farm, Woodnesborough’, Archaeologia Cantiana, cxxiii, 390-391. Parfitt, K., 2005, ‘Rare Early Iron-Age Brooch: Ringlemere Farm, Woodnes-borough’, Archaeologia Cantiana, cxxv, 382-384. Parfitt, K. and Brugmann, B., 1997, The Anglo-Saxon Cemetery on Mill Hill, Deal, Kent, Society for Medieval Archaeology Monograph Series No. 14, London. Parfitt, K. and Halliwell, G., 1983, ‘A Mesolithic Site at Finglesham’, Kent Archaeological Review, 72, 29-32. Smirke, E., 1867, ‘Some account of the discovery of a gold cup in a barrow in Cornwall, AD 1837’, Archaeological Journal, 24, 189-95.