Ringlemere: investigation of prehistoric ring-ditches, M2 and M3

1 RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 keith parfitt Metal-detecting in a field at Ringlemere Farm near Sandwich in 2001 led to the discovery of a spectacular Early Bronze Age gold cup (Parfitt 2003; Needham et al. 2006). The vessel was found on a low rise that was soon identified as the remains of a round barrow. Full excavation of this monument (M1) was made a collaborative project between the Canterbury Archaeological Trust and the British Museum. Investigations were completed in 2006 and an interim report, including a description of the gold cup, followed in Archaeologia Cantiana cxxvii (Parfitt and Needham 2007). That account is now superseded by a detailed monograph containing a revised interpretation and chronology of the complex sequence of prehistoric remains revealed (Parfitt and Needham 2020). Ground survey and aerial photograph analysis conducted across the surrounding area showed that other barrows had formerly existed nearby; apparently a whole barrow cemetery had been all but erased from the landscape by centuries of ploughing. In order to better understand the main monument, some examination of these other sites, which survive only as buried ring-ditches, was important. Ringditch M3 was fully excavated in the summer of 2007, with an extension cut across the ring-ditch of the adjacent M2. The results of these investigations form the basis of the report below. The Ringlemere site is located at the foot of the North Down dip slope, towards the bottom of a long north-east facing slope which constitutes the southern side of the valley of the Durlock Stream. The site lies some 3.75km west of Sandwich, in the parish of Woodnesborough, about 1.5km west of the church. Ringlemere Farm is some 400m to the south-east (Figs 1 and 2). The underlying geology in the area is head brickearth, with some gravel, overlying Thanet Formation clay. Today, the Durlock Stream begins at springs which rise in the immediate environs of the site and flows for about 8km westwards to join the Wingham River (Fig. 1). A ridge of Eocene sands separates the Durlock valley from the south-western edge of the former Wantsum Channel, which once separated the Isle of Thanet from mainland Kent, and was a much-used waterway in ancient times. excavation of monument m1 – summary Extensive excavations undertaken after the discovery the Bronze Age gold cup led to the recognition of the remains of a large, previously unrecorded round barrow, KEITH PARFITT 2 Rive r Stour River Wantsum Little Stour Great Stour Durlock Stream Wingham River Wingham Sandwich Woodnesborough Eastry Deal RINGLEMERE Ramsgate ISLE OF THANET Minster Reculver Broadstairs Margate Former Wantsum Tidal Channel Former Wantsum Tidal Channel Pegwell Bay 170000m 160000m 640000m 630000m Under 10m. Over 10m. Over 20m. Over 30m. Over 40m. Over 50m. Over 60m. Over 70m. Over 80m. Over 90m. 2.5km Based on an Ordnance Survey map with the permission of the Controller of Her Majesty's Stationery Office, © Crown Copyright. Licence No. AL100021009 Monument (M)1. (NGR TR 2938 5698, centred.) The base of this plough-reduced mound was encircled by a substantial ditch about 42m in diameter, with an entrance causeway on the northern side. Centuries later, the surviving mound had served as the focus for an early Anglo-Saxon cemetery. The barrow mound protected a well-preserved land surface below, containing evidence for extensive earlier activity, and making it clear that the site had a long history of use before the barrow was erected. The excavated remains were divided into nine broad phases of activity. These are summarised below with full details presented in the published monograph (Parfitt and Needham 2020). Phase 1: Mesolithic and Early Neolithic, pre-3500 bc: activity during these periods was largely represented by a scattering of flintwork. An excavated arc of stakeholes seemed to relate to a sub-circular structure of some kind. Surviving Fig. 1 Map of north-east Kent showing location of Ringlemere i n relation to rivers and the local topography. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 3 charred wood from of one the stakes gave a radiocarbon date of 3770-3645 cal bc, implying an Early Neolithic date. Another radiocarbon date of 4040-3804 cal bc, for a small collection of burnt ?human bone fragments recovered from a nearby pit probably provides further evidence for site activity during the Early Neolithic, although the bone itself was perhaps re-deposited. Phase 2: Late Neolithic and possibly Chalcolithic (Grooved Ware), c.3000-2200 bc: during the Late Neolithic a period of intense activity began, associated with a complex of more than 230 cut-features. These consisted of variously sized pits, hollows, post-holes, and stake-holes, with three hearths and two possible 10 15 20 25 30 35 10 15 M2 M5 M6 M7 M8 M9 M10 M3 M1 Scattered Late Bronze Age hoard Black Pond Farm Fleming Road Ringlemere Farm Ringleton Manor Durlock Stream Track (disused railway) The Lodge 292 293 294 295 296 297 570 569 568 567 571 572 573 100m Based on an Ordnance Survey map with the permission of the Controller of Her Majesty's Stationery Office, © Crown Copyright. Licence No. AL100021009 Fig. 2 Ringlemere Early Bronze Age cemetery, as established by geophysical survey and aerial photography. KEITH PARFITT 4 graves. Structural arrangements identified included a horseshoe setting (early); a ‘cove’; a trapezoidal building; pit alignments and post screens, constituting an ever-changing layout. Several structures were the foci for activity that involved the generation and deposition of large quantities of cultural debris, in the form of calcined flint, struck flint and distinctive Grooved Ware pottery. Lesser quantities of non-local stone (including axe fragments), fired clay and burnt bone were also recovered. The Grooved Ware-associated activity possibly continued into the Chalcolithic Period, when Beaker pottery appears on the site and new structural arrangements occurred. Phase 3: earliest Early Bronze Age (Beaker), c.2200-1950 bc: most of the Beaker pottery found came from the buried soil profile rather than in feature fills. However, three pits containing whole Beaker pots were discovered. It is possible that two of these were graves, although no skeletal material had survived. Some other features excavated could also be of Beaker date but this cannot be certain. Fragmentary evidence for the existence of a ditch enclosing the site may be tentatively identified. If not datable to late in the Grooved Ware sequence, this early ditch could be associated with the Beaker-phase layout. It was subsequently replaced by a much larger ditch during Phase 4. Phase 4: mature Early Bronze Age, c.1950-1500 bc: around the beginning of the second millennium bc the existing enclosure was transformed by the construction of a broad mound in its interior. This had a turf inner core and a clay capping or skirt, but lacked evidence for any immediately associated burial. The new monument was further enhanced by the renewal of the enclosure ditch on a grander scale than before, with an external bank. The ditch seems to have closely followed its predecessor and enclosed an area measuring between 43.75m (n-s) and 41.50m (e-w), with a narrow entrance maintained on the northern side. On top of the mound a timber façade was erected aligned north-south, respecting a key alignment of the previous phase. Later, a large pit, possibly a grave, was dug alongside the façade; this had contained one or probably two amber objects and the gold cup. Phase 5: Middle Bronze Age – later Iron Age, c.1500-100 bc: this phase covers the lifetime of the monument after its primary use and active reworking, but before agriculture made any significant inroads into it, spanning from the middle of the second millennium bc through to late in the first millennium bc. Activity appears to be sporadic during this phase. The monument’s outer bank may have been deliberately slighted now, perhaps in connection with the beginnings of cultivation around the site. Phase 6: Late Iron Age and Romano-British period, c.100 bc-ad 400: the environs of M1 became an arable landscape with ploughing progressively levelling its encircling earthworks, truncating the top of the ditch and encroaching upon the edges of the mound itself. A negative lynchet was formed around the monument. Down-washed soils began infilling this lynchet during the Late Iron Age/Romano-British period. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 5 Phase 7: Anglo-Saxon cemetery and contemporary land use, c.ad 400-700: on the south-west flank of the barrow mound, a hiatus in cultivation allowed this area to be used as an Anglo-Saxon cemetery during the fifth and sixth century ad. Later, a sunken-featured building was cut into the northern side of the mound but the adjacent lands are likely to have remained under cultivation. Phase 8: Medieval land use, c.ad 700-1300: a rectangular ditched enclosure was created around the mound on its north-eastern side. This mound became a rabbit warren surrounded by cultivated land; infilling of the downslope lynchet continued. Phase 9: Post-medieval to modern cultivation, c.ad 1300 – present: plough reduction of prehistoric mound itself occurred with continued lynchet infilling on three sides. the barrow cemeter y Aerial photograph analysis, geophysical survey and excavation established that the large round barrow represented by M1 was one of a number that had formerly existed in this area (Fig. 2). Now represented only by surviving sub-surface ringditches, these other barrows (designated M2-M9) fall into two discrete alignments along the valley-side − Upper barrow alignment and Lower barrow alignment. Together, they seem to constitute the remains of a substantial prehistoric barrow cemetery, perhaps originally containing eight mounds, but now levelled by the plough. This barrow cemetery probably developed adjacent to the pre-existing ceremonial site represented by M1 (Fig. 2). Upper barrow alignment (M2, M3 and M5): south-west of M1, at a slightly higher elevation, a localised outcrop of natural gravel supported a row of three ring-ditches (M2, M3 and M5; Fig. 2 and Fig. 3). These run roughly north-south, extending about 100m along the valley-side and each most probably originally enclosed a round barrow. At the southern end of the row, unexcavated M5 appears to be a double ring-ditch, with an outer ditch around 28m in diameter and the inner one of about 16m. These two concentric ditches are probably successive, suggesting that this monument, like M1, has more than one main phase of development. Details of the adjacent M2 and M3 follow below. Lower barrow alignment (M1, M6, M7, M8 and M9): a little further down slope, geophysical survey (Birchenough 2006, fig. 4) and aerial photographs indicate the existence of a second row of at least four ring-ditches (Fig. 2, M6, M7, M8 and M9), extending south-east from M1. Running along the valley-side for more than 100m, none of these smaller monuments has been excavated. Just under 60m from M1, the largest is M6. This appears to have a ditch approximately 24m in external diameter. Worthy of note on the geophysics is a strong circular anomaly within it, perhaps a pit. On the northern side a somewhat ephemeral magnetic response suggests the presence of another a very small ring-ditch, approximately 8m in diameter. Designated M7, this is also vaguely discernible on an aerial photograph of 1990. Its apparent placement over, or as an appendage to, the ditch of M6 suggests that it is the later of the two monuments. KEITH PARFITT 6 About 20m south-east of M6 lies M8, with a ditch diameter of some 18m. An apparent break in its circuit on the south-western side is of some interest and suggests the presence of an entrance causeway here. Inside the ditch there is an indistinct, arc-like, central anomaly represented by an area of raised magnetic response. There are also a considerable number of magnetic anomalies outside the presumed break in the ditch circuit. These are presently difficult to interpret and are not certainly of prehistoric origin. The last ring-ditch of this alignment is M9, located 40m to the south-east of M8 and about 160m from M1. It is approximately 14m in diameter and from the geophysics seems to impinge upon (or vice versa) the north-western end of an irregular, ill-defined rectilinear feature. ring-ditch m2 This ring-ditch had been previously identified on aerial photographs and geophysical surveys of the area (Birchenough 2006, fig. 4). It was also visible on the ground in the growing crop during the spring of 2002, being 28-30m in overall diameter. (NGR TR 2933 5693, centred.) A gas main had been cut across the middle of the M1 M3 M2 11 12 13 11.5 12.5 13.5 10m Fig. 3 Surface contour survey of the land around Monuments 1, 2 and 3, before excavation (contours at 0.25m intervals). RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 7 monument sometime during the 1980s but no archaeological work was undertaken then. The east-west aligned pipe-trench had passed just north of the centre of the enclosure and some damage to the monument must have occurred. Nevertheless, cropmarks appearing on a Google Earth satellite image, taken in April 2007, hint at the survival of an inner ring-ditch, about 15m in diameter, set within the main enclosure although not concentric to it. The geophysical plot, too, hints at the presence of internal features, with the possibility that the main ring-ditch may be penannular, with an entrance on the north-east side, facing M1. Two anomalies, perhaps pits or large post-holes, could mark the ditch terminals here but precise interpretation is hampered by the presence of a continuation of a later, Roman field boundary ditch into this area (see below). If M2 does possess an associated inner ditch this would suggest a complex monument perhaps broadly comparable to the adjacent M5 (see above). The presence of the gas main meant that only limited excavation on M2 was undertaken in 2007. A single trench 1.95m wide was cut across the outer ringditch on its northern side (Figs 4 and 5). This was extended for 1.50m from the inner lip of the ditch into the enclosed area, sufficient to show that no mound material survived here. Nor were any internal features revealed. The surface of the undisturbed natural gravel was found be buried at a depth of 0. 30m below present ground level, sealed by a layer of plough soil. Outside the barrow ditch to the north, the soil overlying the gravel was somewhat thicker, reflecting a general rise in the level of the surface within the enclosed area. The excavated trench established that the ring-ditch [F.2509] was of substantial proportions with an almost V-shaped profile. It was about 3.30m wide across the top and 1.50m deep (Figs 6 and 7). Convexly sloping sides gave way to a flat base between 0.40 and 0.65m across. From the drawn section (Fig. 7), it could be suggested that the infilled ditch had subsequently been re-cut, wider but to only about half its original depth. Such a recut was not specifically recognized during the excavation and more probably the recorded profile is merely the result of continued erosion of the upper ditch sides after its lower part had become filled. The lower and middle fillings of the ditch consisted of a series of weathered gravelly silts (Fig. 7, Contexts 2541-2548). These produced a combined total of ninety-two prehistoric struck flints, together with two fragments of calcined flint. The upper ditch filling comprised a light brown clay loam with rather less gravel (Context 2510), which produced a further fifty-nine pieces of prehistoric flintwork, most of which is likely to be residual. Only two sherds of pottery were recovered from the ditch filling. These came from weathered gravel deposits on its sides (Contexts 2542 and 2544) and consist of a worn piece of Roman samian ware (2544) and an Anglo-Saxon sherd dated c.ad 775/800-875 (2542). These suggest that the upper half of the ditch still remained open during the early historic period. The overall impression gained was that following its excavation (or any subsequent recutting), the ditch had been left to slowly silt up through natural processes. KEITH PARFITT 8 M1 M3 M2 Trench 1 (2002) Trench 2 (2002) Tr.3 (2003) Trench 4 (2003) Trench 5 (2004) Trench 6 (2005-6) Trench 7 (2006) Trench 8 (2006) Trench 9 (2007) 10m Fig 1.7 Annual excavated areas (1:800) Fig. 4 Plan of the excavation trenches by year. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 9 Fig. 5 General plan of Trench 9 showing excavated features associated with M2 and M3. F.2550 F.2633 F.2615 F.2613 F.2525 F.2625 F.2623 F.2635 F.2572 F.2629 F.2621 S.331 F.2631 S.308 S.332 Gr 58 Gr 55 F.2609 F.2553 F.2530 Gr 53 Gr 57 Gr 56 F.2568 F.2570 F.2619 F.2535 F.2523 F.2617 F.2517 F.2533 F.2605 F.2607 F.2611 F.2595 S.305 S.312 MONUMENT M3 Gr 54 MONUMENT M2 S.311 F.2509 S.333 F.2507 F.2507 S.325 S.330 F.2521 S.328 S.314 S.320 10 0 10 Metres S.323 S.351 S.318 S.350 S. 354 S.347 S.345 + Pottery, 2526 S.343 KEITH PARFITT 10 Fig. 6 Excavated trench across the ring-ditch of M2, looking e ast. Scale, 2 metres. 2542 2501 2502 2510 2541 2544 2545 2546 2548 2543 N S OD 13.40m F. 2509 Sec􀆟on 311 2547 0 1m Fig. 7 Section across the ring-ditch of M2. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 11 ring-ditch m3 Like M2, this ring-ditch was identified on aerial photographs, through geophysical survey and by ground observation. (NGR TR 2932 5696, centred.) Its site was fully excavated in 2007 (Fig. 4, Trench 9) and was found to consist of a continuous ring-ditch [F.2507] enclosing a fairly precise circle with an internal diameter of 15.25-16.25m (Fig. 5 and Figs 8 and 9). There was no surviving physical evidence for any central barrow mound, although one most probably existed originally. The ring-ditch [F.2507]: this feature was almost completely emptied during the course of the excavation. It was found to be between 1.00-1.80m wide across the top and in two places on the north-eastern side had been cut through by a later, Roman, boundary ditch (see below). Subsequently, two Anglo-Saxon graves (Grs 54 and 56) were also dug into its upper levels. The ditch had clearly been truncated by later ploughing. As surviving, it was 0.44-0.65m deep. The sides were convexly sloping, with a flat to slightly dished base 0.20-0.45m across. The ditch’s profile was somewhat variable around its circuit (Figs 10 and 11) but there was no evidence for any re-cutting. A series of thirty-four separate fill deposits was defined within the ring-ditch and these appeared to represent naturally weathered silts, with coarse primary gravel fills giving way to finer, more soily secondary deposits. Overall, the ditch fills produced a total of just over 560 pieces of struck flint, 77 fragments of calcined flint and 112 sherds of pottery. Amongst the pottery recovered there was just one prehistoric sherd. This is a small piece of flint-tempered ware from the upper filling on the north-west side (Fig. 11, Section 330, Context 2508). It is probably Bronze Age or Iron Age but cannot be closely dated. Two pieces of Fig. 8 General view across excavated M3, looking east. Scale, 2 metres. KEITH PARFITT 12 Fig. 9 General view across excavated M3, looking south-east towards Ringlemere Farm. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 13 early Roman grog-tempered ware came from the upper filling on the northern side (Fig. 11, Section 332, Context 2582) in the area where Roman ditch [F.2503] cut through: these are likely to be intrusive from that later feature. All the remaining pottery discovered was found in one spot on the south-west side, where a group of 109 sherds was found lying in the top of the ditch (Fig. 5, Context 2526). These represent two-three separate vessels of Anglo-Saxon date, c.ad 800-850/875. This material must have originally been deposited in a shallow pit cut into the top of the ring-ditch sometime during the ninth century ad. Ploughing had subsequently removed all but the very base of this feature. The prehistoric flintwork recovered from the ring-ditch constitutes the largest group of lithic material recovered in 2007 and amounts to over half the total assemblage. The flints themselves were scattered throughout twenty-four separate contexts, with almost two-thirds contained within its uppermost levels. There were no particularly obvious concentrations or deliberately placed deposits and the bulk of the flints appear to be residual, derived from the earlier activity in the general vicinity of the monument. The area enclosed by the ring-ditch: clearance of the modern ploughsoil within the area enclosed by the ring-ditch revealed the truncated surface of natural gravel directly below, with no surviving traces of any central barrow mound or early topsoil. The bases of twenty shallow hollows, pits and post-holes were located cut into this surface (Figs 5, 8 and 9; Table 1). All had been reduced by ploughing and none contained any definite evidence for a burial. There was no good dating evidence and half the features failed to produce any datable finds. Most are perhaps likely to predate the main monument and relate to earlier activity in the area, as was found at M1. From its central positioning, appropriate size and shape (Fig. 5 and Fig. 12), one pit [F.2525] may be suggested as representing a grave, more or less contemporary with the ring-ditch. A line of five post-holes on the north-east side, however, have been tentatively suggested as representing a fence line of Roman date, running parallel with a Roman field boundary ditch (see below). Central pit, [F.2525] and post-hole [F.2613]: almost at the exact centre of the enclosed area was a neat, oval pit filled with deposits of gravely soil [F.2525]. This measured 1.02 by 1.22m and was aligned ene by wsw. It was 0.30m deep, with steep sides and a slightly dished base (Figs 5, 11 and 12). Dug into the base of the pit at the north-eastern end, a deeper depression appeared to represent a substantial post-hole. This was D-shaped in plan and measured 0.56m (nnw- sse) by 0.28m (ene-wsw). It was 0.15m deep with steep sides and a flat bottom (Fig. 11, Section 308). It seems probable that this feature had originally held an upright wooden post positioned at one end of the pit. The D-shaped form suggests that the original post was probably a split tree trunk with the flat (split) surface facing south-west. The filling of the post-hole (Context 2540) consisted of brown silty clay containing occasional flint pebbles. It produced three small, unworked struck flints and fifteen fragments of calcined flint but no other finds. The lower filling of the main pit (Context 2602) consisted of a compact brown gritty loam containing frequent small and medium flint pebbles. Very careful excavation produced only a single fragment of calcined flint. KEITH PARFITT 14 Fig. 10 Sections across ring-ditch of M3. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 15 Fig. 11 Sections across ring-ditch and central pit [F.2525] of M3 and Roman field ditches. KEITH PARFITT 16 Table 1. Summar y of excavated features (pits /post -holes ) within ring -ditch M3 F. No. Type Shape Length (m) Width (m) Axis Depth (m) Sides Base Notes and finds 2517 P-hole Oval 0.39 0.35 E-W 0.16 Steep Rounded no finds 2523 Pit Irreg./oval 0.75-0.80 0.78 N-S 0.12 Sloping Flat 1 flint; 1 calcined 2525 Pit Oval 1.22 1.02 ENEWSW 0.30 Steep Dished Poss. grave. D-shaped post-hole in base at northern end 2533 P-hole Ovoid 0.37 0.30 NE-SW 0.16 Steep Dished 1 flint 2535 Pit/p-hole Sub-oval 0.86 0.68 E-W 0.12 Steep Flat-dished no finds 2550 Pit/p-hole Irregular 0.70-0.82 0.78 NE-SW 0.20 St’p/Slop’g Irregular 3 calcined 2568 Pit/p-hole Oval 1.28 0.61 NW-SE 0.27 Steep Irregular no finds 2570 P-hole Oval 0.53 0.29 E-W 0.18 Sloping Rounded no finds 2572 P-hole Oval 0.33 0.31 NE-SW 0.14 St’p/Slop’g Pointed Poss. fence line – nil 2613 P-hole Oval 0.60 0.51 NE-SW 0.22 Steep Dished no finds 2615 Pit Irreg. 3.45 0.78-1.46 NW-SE 0.14-0.31 St’p/Slop’g Irregular 2 flint; 4 calcined. Poss. more than one feature 2617 Pit/p-hole Oval 0.28 0.25 NE-SW 0.40 Steep Dished 3 calcined 2619 P-hole Oval 0.28 0.25 NE-SW 0.14 Steep Dished no finds 2621 Pit/p-hole Sub-oval 0.63 0.55 NW-SE 0.15 Sloping Rounded Poss. fence line – nil 2623 Pit/p-hole Oval 0.52 0.47 E-W 0.21 Sloping Pointed no finds 2625 Pit/p-hole Oval 0.82 0.42-0.55 E-W 0.17 Sloping Dished Poss. fence line; 1 flint 2629 Pit/p-hole Oval 0.48 0.43 N-S 0.11 Sloping Dished Poss. fence line – nil 2631 Pit/p-hole Ovoid 0.68 0.76 E-W 0.12 Steep Dished Poss. fence line. Shallow step on N side could be earlier posthole – nil 2633 Pit/p-hole Ovoid 1.05 0.75 NE-SW 0.18 Sloping Dished 5 calcined. Scoop on NW side 2635 Pit Oval 0.98 0.92 E-W 0.20 Sloping Irregular 1 flint RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 17 Fig. 12 Possible grave [F.2525] at the centre of M3, looking north-east. Scale, 1 metre. The main upper filling of the pit (Context 2524) comprised a light brown clay loam containing moderate quantities of flint pebbles and occasional carbon specks (more frequent towards the south-western end). This deposit also included thirtyseven more struck flints and twenty-five calcined flints. The flintwork consists of a mixed collection of waste including one core, one core fragment and thirty-five small, unworked flakes. The flints were distributed throughout the filling, with no suggestion of deliberate deposition. It remains uncertain whether the main pit simply represents a construction pit to allow the insertion of the large D-shaped post or whether it constituted a central grave, marked at one end by the post. Certainly, the proportions of the pit would have readily allowed the insertion of a crouched inhumation, such as are frequently found in a similar position within many barrows and ring-ditches. No traces of any human skeleton survived within the pit but the acidic nature of the gravel into which it was cut is likely to have led to the complete decay of any bone. Although the issue must remain unresolved, the balance of probability seems to be that [F.2525] does represent a central grave pit, originally containing a crouched inhumation marked by a timber upright. If so, the body was apparently interred without grave offerings – unless these too were of perishable organic material. KEITH PARFITT 18 Fig. 13 Sections across Roman field ditches and features excavated within M3. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 19 About 0.75m wsw of pit [F.2525] and continuing its axis another substantial post-hole was discovered [F.2613]. This would seem to mark the position of a second timber upright positioned close to the middle of the enclosed area (Fig. 5). The post-hole was oval in shape and measured 0.60m (ne-sw) by 0.51m (nw-se). It was 0.22m deep, with steep sides and dished base (Fig. 13, Section 343). The filling consisted of a brown clay loam with flint pebbles but this contained no finds. Other internal features: a number of other small pits and post-holes were located within the enclosed area (Fig. 5; Table 1). None of these can be closely dated and their purpose remains uncertain. They form no clear patterns and generally appear as a somewhat miscellaneous scatter. Many are likely to relate to activity in the area prior to the construction of M3, and as such they could be of late Neolithic or Beaker date and be contemporary with features located below M1. On the northern side, pit [F.2635] might have originally been cut by the ring-ditch but the stratigraphic relationship had been subsequently destroyed by a Roman ditch (Fig. 11, Section 331) and the only find was a single struck flint. Most of the features surviving within M3 were less than 0.30m deep, others could perhaps have been completely removed by ploughing. On the south side, pit/ post-hole [F.2617] stands out as being slightly deeper than the other features, at 0.40m (Fig. 13, Section 345). feat ures cutting, or lying outside, m3 Ring-ditch M4 (discounted): clearance of the area supposedly occupied by another ring-ditch, M4, immediately to the west of M3, failed to reveal any evidence for the existence of such a feature here. A combination of changing geology, a fortuitously positioned straight gully, [F.2611] (see below) and over-interpretation of some fuzzy geophysical survey evidence account for the original mistake. Trench 9 now disproves the existence of M4. Removal of this spurious circle has the effect of regularizing the arrangement of ring-ditches south-west of M1 to form the Upper barrow alignment, as described above. Ditch [2611]: was aligned roughly east-west and was traced for a minimum distance of 3.65m from its eastern terminal, continuing beyond the excavation limit (Fig. 5). It was 0.93-1.12m wide and 0.35m deep with steep-sloping sides and a dished base. The filling consisted of a single deposit of dark brown clay loam with pebbles. This produced four prehistoric struck flints and sixteen calcined flint fragments, all of which could be residual. The precise date of this ditch must remain uncertain but it might be broadly contemporary with Roman field boundary ditches, [F.2503] and [F.2505], further to the north-east. An L-shaped gully [F.2511] was traced in an extension to the north-west (not on plan Fig. 5). This produced thirty-two struck flints but again these could all be residual. Other miscellaneous features: irregular hollows, [F.2553] and [F.2605]: hollow [F.2553] was an irregular, ill-defined feature located immediately north of M3 (Fig. 5; Table 2). It consisted of an elongated, slightly curved, ovoid pit, with maximum dimensions of 3.18m (ne-sw) by between 0.75 and 1.22m (nw-se). It KEITH PARFITT 20 Table 2. Summar y of miscellaneous features excavated outside ring -ditch M3 F. No. Type Shape Length (m) Width (m) Axis Depth (m) Sides Base Notes and finds 2511 Gully L-shaped 7.00 and 2.00 (min.) 0.40-0.66 N-S / E-W 0.21 Steep/sloping Flat-irreg. 32 flint; 3 calcined 2521 Pit/post-hole Oval 0.45 0.40 NW-SE 0.14 Steep/sloping Rounded occ. calcined 2530 Gully segment Linear 1.35 0.43 E-W 0.17-0.20 Steep Flat 1 pot; 6 flint; 3 calcined 2553 Pit Irregular 3.18 0.75-1.22 NE-SW 0.38 Steep Irreg. 4 flint (inc. axe/adze); 4 calcined 2595 Post-hole Oval 0.33 0.32 N-S 0.18 Steep Pointed 1 calcined 2605 Large pit/ hollow ?Oval 3.60 1.06 (min.) E-W 0.43 Steep/sloping Undulating 6 flint; 4 calcined 2607 Gully segment Linear 2.98 0.73 NW-SE 0.35 Steep Rounded 1 flint 2609 Pit/post-hole Oval 0.90 0.54 NW-SE 0.23 Steep/sloping Dished no finds 2611 Ditch terminal Linear 3.65 (min.) 0.93-1.12 E-W 0.35 Steep/sloping Dished 4 flint; 16 calcined RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 21 was about 0.38m deep with steep sides and irregular base. The filling comprised deposits of silty clay containing a few flint pebbles. The main filling yielded four prehistoric flints and four small calcined flint fragments. Of particular interest amongst the flints is a tranchet axe/adze of Mesolithic appearance. The three other struck pieces are small, somewhat undiagnostic waste flakes that show no Mesolithic characteristics. It seems most likely that this feature represents some sort of naturally formed hollow, quite possibly a tree throw, into which a few flints had casually collected as residual material. Another large hollow, [F.2605], partially examined on the opposite side of the site, may have been of similar origin (Fig. 5; Table 2). Its filling of brown clay loam and pebbles contained four calcined flints and six struck flints, one of Palaeolithic date (see p. 26). Roman boundary ditch and a possible fence-line: at two points on its north-eastern side, the infilled ring-ditch of M3 was cut through by an angled ditch containing Roman material (Fig. 5, [F.2503] and [F.2505]; Fig. 9). Continuations of this ditch are visible on aerial photographs and the geophysics plots. South-east of the excavated area, the ditch extended to meet the ring-ditch of M2 on its north-eastern side, whilst at M3 itself, the ditch realigned by about forty degrees, to follow a westerly course for a further 20m. These separate alignments apparently reflect a desire to skirt around the margins of M3 (Figs 5 and 9), implying that there were still physical remains of the prehistoric monument surviving in the Romano- British landscape, even if the ring-ditch was largely invisible. The survival of a barrow mound, only later lost to the plough, is clearly implied. The original ditch [F.2503] was between 0.75 and 1.15m wide and about 0.50m deep, with sloping sides and a flat base (Fig. 11, Section 332; Fig. 13, Sections 320 and 323). Most of its exposed length was emptied, the filling consisting of a dark grey-brown loam with flint pebbles. This produced some sixty-five pot sherds, together with a number of fragments of burnt daub and a quantity of residual prehistoric flintwork. The pottery assemblage is something of a mixed group. Most of the pieces fall within the general period c.ad 50-150/200 implying that the ditch was largely full by the start of the third century ad. A number of sherds, including nine flint-tempered pieces, probably belong to the Iron Age and seem to be residual. Also residual, are two sherds of Bronze Age Collared Urn, one with cord-impressed decoration. In addition, there are eleven sherds of early Anglo-Saxon ware datable to the period c.ad 450-550/600. All but one of these came from the south-eastern sector of the excavated ditch and derive from a single vessel. They must be intrusive and connected with activity associated with the nearby Anglo-Saxon cemetery. The original ditch had been partially recut, with [F.2505] representing a redigging of the old western ditch line, which had not continued east of M3. The new ditch was cut on the precise line of [F.2503] and clearly represented its direct replacement (Fig. 5). The recut ditch terminated just west of the angle change in the line of [F.2503] – apparently only that section of ditch west of M3 needed replacement. The subsequent ditch was slightly larger than the original, being between 1.07 and 1.18m in width and about 0.64m deep; consequently, all traces of the earlier ditch had been removed. The new ditch had steep-sloping sides and KEITH PARFITT 22 a dished base (Fig. 11, Section 331; Fig. 13, Section 325). It was filled with dark brown clay loam containing flint pebbles. This produced twenty-nine sherds of pottery, together with a number of fragments of burnt daub, a small piece of chafftempered ware, a tooth from a large domestic animal, a sandstone fragment and a quantity of residual prehistoric flintwork. The latest pot-sherds date to c.ad 150-200, with some earlier Roman material and seven residual prehistoric flinttempered sherds. Overall, it seems likely that this re-cut ditch is not much later in date than its predecessor. South-west of the Roman boundary ditch, within the area enclosed by ring-ditch M3, an arc of five shallow pits (Fig. 5, [F.2621], [F.2625], [F.2631], [F.2572] and [F.2629]; Table 1) ran almost parallel to the ditch line. These lay 2-3m from the lip of the boundary ditch, spaced at intervals of between 2 and 4m. They are not closely dated; the only find was a single prehistoric flint in [F.2625]. From their positioning in relation to the Roman ditch, it seems possible that these are postholes representing a boundary fence running parallel with the ditch where it passed by the prehistoric monument. Presumably, such post-holes had originally been cut into the edge of the inferred barrow mound here and this might explain their shallow depth into the natural gravel (Fig. 13, Sections 347, 350, 354, 318 and 351). If so, the implication must be that the mound in this area was not very high by this stage. No further post-holes continuing the line beyond M3 were discovered so the interpretation must remain speculative. Anglo-Saxon burials, Graves 53-58: previous excavations at Ringlemere had established that the south-western side of M1 was occupied during the fifth century ad by an Anglo-Saxon cemetery containing over 50 burials (Parfitt and Needham 2020, 141, fig. 4.2). The full extent of this cemetery has still to be determined but six more inhumation graves (Fig. 5, Graves 53-58), together with part of a probable seventh, were discovered in 2007. These new graves appear to represent a discrete group placed adjacent to the northern side of M3. Their presence would seem to provide further evidence for the former existence of an upstanding barrow here. As previously, all the new graves were aligned roughly east-west. Two (Grs 53 and 54) were cut into the Roman field boundary ditch, which must have been filled by this time and apparently played no part in delimiting the burial area. Another burial (Grave 56) cut the ring-ditch of M3 but no graves occurred within the enclosed area. The acidic gravel subsoil meant that no skeletons survived but the size of one grave (Gr. 55) indicated that it belonged to a child. Grave items were recovered from four of the burials. Grave 53 contained five brooches and forty-two beads and must represent the burial of a reasonably well-off woman. Provisional dating of the grave-objects recovered suggests that they are again of fifth-century date. specialist reports on finds (not illustrated ) In comparison with M1 the prehistoric finds assemblage associated with M2 and M3 is quite small. This is largely explained by the non-survival of any barrow mounds or buried early topsoil here, which produced the bulk of the finds at M1. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 23 Nevertheless, a significant collection of prehistoric flints was recovered during the present excavations, mostly from the ring-ditches, together with a smaller pottery assemblage. Due to the acidic nature of the soils virtually no bone had survived. The Anglo-Saxon grave material has been kept with the main cemetery collection housed at the British Museum. The other finds from the 2007 excavation will be placed in a suitable Kent repository in due course. Pottery by Nigel Macpherson-Grant† The quantity of pottery recovered from the area of M2 and M3 was small, amounting to just over 200 sherds, and no Neolithic Grooved Ware or Beaker sherds are present within the assemblage. The largest single group of material was represented by 109 sherds of middle Anglo-Saxon ware recovered from the top of the ring-ditch on the south-western side of M3 (Fig. 5; see above). Overall, between five and six main archaeological periods are represented ceramically. These are described below. Early Bronze Age: this period is represented by just two sherds – one moderatesized and a definite example of Early Bronze Age Collared Urn-style pottery, and one small piece probably from a similar type of vessel. Both came from Roman ditch [F.2503] and are grog-tempered. Considering the relatively soft nature of the fabric, the larger sherd is remarkably unworn. It is from the collar-zone of a fairly thick-walled large-diameter vessel, the collar with typically impressed cord decoration – which in this case consists of a broad band of close-spaced horizontal lines above or below a zone of multiple-line chevrons. The sherd is datable to between c.2000-1500 bc (Gibson 1986) and is perhaps re-deposited from a disturbed cremation burial urn. Later Prehistoric: flint-tempered sherds were recovered from Context 2508 of the M3 ring-ditch (Section 330), from Roman ditches [F.2503] and [F.2505], and from the general filling of Grave 57. In all but one context these sherds were accompanied by later pottery. The highest quantities came from the Roman ditch fills. All the sherds are generally quite heavily worn – and all should be residual. Although some of this material could be as early as Middle Bronze Age, none of the recovered sherds are diagnostically typical. On the basis of available manufacturing characteristics there is a personal preference to see all of this material as belonging to the first millennium bc. Only one rim was recovered, a burnt fragment of coarseware jar with a fairly flaring and everted rim. Although the form is not utterly typical of Mid/Late Iron Age types – and could occur in an Earliest Iron Age assemblage – it is felt that this example is probably of Mid/Late Iron Age date. Late Iron Age-mid Roman: this broad phase embraces three archaeological periods but is here presented as one, representing many years of probably continuous land-use. Within this phase Late Iron Age activity is represented by ‘Belgic’-style grog-tempered sherds, eight in all, equally spread between Roman ditches [F.2503] and [F.2505]. Most are body sherds, with only one rim. The simple upright and rather primitive form of the latter suggests a first century bc date rather than later. KEITH PARFITT 24 The quantities recovered for this period are small and should represent no more than the natural bi-product of discards and stray losses from nearby habitation during the period c.50 bc-50 ad. This modest scale of activity continued into the middle and later years of the first century ad – with only one fresh, moderatesized sherd from ditch [F.2505] representing the Conquest-period. This is from a neatly-made wheel-thrown, finely grog-tempered small jar with an upper-shoulder band of broad rouletting, copying contemporary imported Gallo-Belgic white ware beakers, and datable to c.ad 50-75. The Early Roman period is initially represented by a few very worn scraps of Southern Gaulish samian, several sherds from Kentish fine wares (including two pieces from a fine oxidised thin-walled Upchurch-type cordoned beaker or jar from ditch [F.2503] and a few Romanising grog-tempered coarsewares. Most of these pieces should date to the late first century ad. However, the main phase of activity appears to be during the first half of the second century, predominantly represented by sherds from mostly rather softly-fired oxidised Romanising native grog-tempered coarseware storage and kitchen vessels but also by a few pieces from sandy ware jars and at least one Canterbury pink-buff tableware flagon. Most of this material comes from Roman ditches [F.2503] and [F.2505], with a few other sherds from the upper filling of the M3 ring-ditch (Context 2582; Section 332), gully segment [F.2607] and residual in the general filling of Grave 53. If the lower count of sherds datable to the later years of the second century is a genuine reflection of original trends, the implied nearby settlement appears to have gone into decline or shifted its activity orientations during the mid-Roman period – a few of the harder-fired, now fully Romanised, Native Coarse Ware sherds and one from a late-phase, c.ad 150-175, Canterbury product all belong to this period. Other than one rolled-rim jar sherd and one or two from fine sandy coarsewares, that might date to the third century, the recovered evidence indicates little obvious Roman activity after c.ad 200. Early Anglo-Saxon: a total of thirteen sherds from handmade jars with variably sandy fabrics were recovered from four contexts – one small, fairly worn sherd from sub-soil layer 2502 (Section 311), another moderate-sized but fairly worn sherd from the filling of gully segment [F.2530], ten pieces, virtually unworn and from the same vessel in the top of Roman ditch [F.2503] (Context 2573, Section 320) and another small worn scrap from elsewhere in the same ditch. Although the overall fabric type of all these sherds could just suggest a Mid/Late Iron Age or Conquest-period ad, or even an earlier eighth-century variant of the following Mid Saxon group of sherds (see below), at least two examples (f rom [F.2503] and [F.2530]) are burnished internally, a trait that does not occur on Mid/Late Iron Age non-bowl forms nor on Mid/Late Saxon jar forms. In addition, the samevessel base sherds from Context 2573 are near-fresh and atypical of that context’s earlier material – so that a second or first century bc date for these is unlikely. The only other period this type of handmade sandy ware vessel could occur is Early Saxon and, as such, is broadly in keeping with Early/Mid Saxon fabric types recorded from the Canterbury Anglo-Saxon sequence and elsewhere within the region. Since, as seems likely, these sherds are Anglo-Saxon, the low-angled base sherds from Context 2573 could easily come from a shouldered, biconical, jar. RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 25 Further, the single rather coarsely sandy sherd from Context 2581 of ditch [F.2503] (Section 332) appears to have traces of a single broad horizontal groove above two, possibly three, diagonal incised lines – a design-format that could easily be from a sub-biconical jar with multiple incised lines forming chevrons below one or more broad horizontal grooves – utterly typical of Myres’s Early Anglo-Saxon ‘Kentish chevron’ style of decoration and broadly datable to between the mid fifthearlier seventh century ad (Myres 1977). However, the combination here of this decoration and purely sandy fabrics would, by comparison with frequency trends from within the Canterbury sequence, suggest an earlier date-emphasis, more probably c.ad 450-550 (Macpherson-Grant 1995). Mid Anglo-Saxon: pottery of this date was recovered from two contexts – a single rim sherd from ring-ditch M2 (Context 2542, Section 311) and a moderate-sized group of rim and body sherds from the top of the ring-ditch of M3 (Fig. 5; Context 2526). The single sherd from Context 2542 is very worn and re-fired, those from 2526 represent two separate vessels, one by ninety-five sherds, and the other by only fourteen sherds. Both vessels are handmade Canterbury sandy ware jars with flaring everted rims, the first with external firing colours varying between a predominant oxidized buff to partially reduced dirty grey, the other totally reduced near-black. One sherd from the first vessel, and possibly others, has characteristic external knife-trimmed facets typical of ninth- to mid tenth-century Canterbury and at least one other North Kent pottery during this period. In Canterbury, knifetrimming is possibly present from the beginning of the Canterbury sandy ware industry, around ad 750/775. The earliest examples show a tendency for fairly light, irregular vertical trimming on certain vessels, sometimes associated with light irregular burnishing, from c.ad 800 onwards. Between c.ad 850/875-950 the degree of trimming is far more severe, frequently with fairly deep broad scalloping of the pot’s surface, each resultant facet accompanied by marked grit-scoration as the knife blade drags sand-grains across the trimmed areas. Here, the combination of form and lighter degree of knifing suggests a date c.ad 800-850, but possibly as late as c.ad 875 for both this vessel and the rim from Context 2542. For the main pot from Context 2526 many of the sherds show a moderate degree of unifacial wear externally, interiors of most being fresher and suggesting deposition and spread for most onto a flattish surface. Conversely, those from the second vessel are mostly fresh and unworn and were almost certainly buried shortly after discard. Prehistoric flintwork by Geoff Halliwell and Keith Parfitt The previous work undertaken at Ringlemere produced a combined total of more than 53,000 pieces of prehistoric struck flint. The bulk of this material (47,800 flints) came from the excavation of M1 (Trenches 1-8; Healy 2020), whilst field walking of adjacent ploughlands yielded another 5,500 pieces (Butler 2020). The excavation of Trench 9 in 2007 produced just 1,021 further pieces (12,538g), distributed throughout thirty-five separate contexts (Table 3). Over half the flints came from the filling of the ring-ditch of M3, with about 150 more from the single trench cut across the ringditch of M2. The bulk of the material is generally fresh and unpatinated. As with the previous flintwork recovered, several different industries are represented. KEITH PARFITT 26 The 2007 assemblage essentially constitutes a smaller sample of the range of lithics previously collected from fieldwalking and the excavation of M1. The bulk of the pieces are unworked waste flakes with some thirty cores. The cores are mostly quite small, broken and worked-out fragments rather than complete specimens. The largest complete core is of river gravel, has two/three platforms and weighs 351g. The range of raw flint material employed for knapping at Ringlemere has previously been described (Butler 2020, 9; Healy 2020, 149-151). The flint from the present excavation reflects the same general pattern, with local river gravel, Bullhead and Downland sources predominating. There is quite a bit of primary material present which seems to represent odd protuberances being taken off raw gravel nodules preparatory to knapping. Palaeolithic, Mesolithic and Neolithic flint: only a small amount of flintwork recovered from Trench 9 appears to pre-date the Bronze Age. Some half a dozen abraded flakes with mottled, variously coloured patinas stand apart from the bulk of the flints and appear to be much older (a few similar pieces had also been noted in previous seasons). Included amongst these is one large flake of classic Levallois type, with an extensively worked platform. There seems little doubt that this small group of quite distinctive flints derives from the top of the river gravel deposits exposed in the excavation and potentially they provide evidence for Middle Palaeolithic activity at Ringlemere around 50,000 years ago. Further work on the origins and nature of the gravels in the area is required and none of the early lithic material recovered so far has been found well-stratified within the gravel itself. Several pieces of Mesolithic flintwork were identified in Trench 9. Although blades are scarce, two Mesolithic-style blade cores come from the plough soil. A large, irregular pit located immediately north of M3 [F.2553] produced a fresh Mesolithic tranchet axe/adze. This is 128mm in length and is apparently manufactured from a black downland flint containing grey cherty inclusions. The three other flints recovered from this feature, however, do not derive from the same implement and are not readily recognizable as being Mesolithic. Several other examples of such Mesolithic tranchet axes/adzes, together with a number Table 3. Distribution of prehistoric flintwork from Trench 9 Description Number Filling of ring-ditch, M2 151 Filling of ring-ditch, M3 564 Filling of central pit within M3, [F.2525] 40 Filling of other features enclosed by M3 6 Filling of miscellaneous features outside M3 53 Filling of Roman ditches, [F.2503] and [F.2505] 112 Filling of Anglo-Saxon graves 27 Ploughsoil and subsoil deposits 68 Total (12,538g) 1,021 RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 27 of sharpening flakes, have previously been found at Ringlemere, along with other flints of this period (Butler 2020, 10; Healey 2020, 151, fig. 5. 1, L9). The few Mesolithic pieces from 2007 provide further evidence for activity in the region, although the quantity of material recovered is not suggestive of any intensive Mesolithic habitation in the immediate area. As previously, however, it may be suspected that other, less diagnostic flints of this period have been passed over, being indistinguishable from later material present (Healy 2020, 151). Definitive lithic evidence for Neolithic activity is very sparse with no diagnostic arrowheads, polished axes (even fragments), or serrated flakes present within the 2007 assemblage. Three thumbnail scrapers, two from the filling of the M3 ringditch and one from Roman field boundary ditch [F.2505], however, indicate that some subsequent Beaker flint-working was occurring in the general area. From its overall characteristics, the bulk of the flintwork recovered from Trench 9 appears to belong to the subsequent Bronze Age period. Bronze Age flintwork: based on the production techniques, often quite crude, the bulk of the flintwork recovered from Trench 9 appears to belong to the Bronze Age, especially its later stages, implying increased activity in the area at this time. Just over 10% of the Trench 9 assemblage shows some evidence of working, although there are very few well produced tools (Table 4). Scrapers are relatively scarce and make-up just 18% of all the worked material. Rather than formal tools, the bulk of the worked pieces are represented by small flakes exhibiting miscellaneous retouching or evidence of utilisation. Such pieces account for about 75% of the total worked flint recovered. The figures for these two types stand in marked contrast to those from excavated contexts in the adjacent M1, where scrapers were the dominant form, representing about 55% of the worked total and miscellaneous retouched pieces about 12.5% (Healey 2020, tables 5.1, 5.5 and 5.7). The significance of this variation remains uncertain, although it must be noted that the Trench 9 assemblage is poorly stratified and much of the material is likely Table 4. Summar y of worked flint from Trench 9 (35 conte xts ) Type No. % Axe/adze (Mesolithic) 1 0.9 Scraper: End 12 10.8 Side 5 4.5 Thumbnail 3 2.7 Other type 1 0.9 Notched flake 2 1.8 Awl/piercer 1 0.9 Misc. retouch 59 53.2 Utilised 24 21.6 Other 3 2.7 Total 111 100 KEITH PARFITT 28 to be residual, derived from eroded land surfaces in an area where there had long been regular flint production and use. The lithics from very few of the excavated contexts in Trench 9 provide any clear evidence of particular activity in the area. The single trench cut across the ring-ditch of M2 produced 151 pieces of flintwork, mostly from its middle and upper filling, but it seems probable that all this material is residual in its excavated context, the bulk having perhaps accumulated here during the historic period. Similarly, the filling of the M3 ring-ditch yielded more than 500 pieces of flintwork but no particular concentrations, such as could indicate specifically placed deposits or in situ knapping debris. Instead, the material here was scattered throughout twenty-four separate contexts around the circuit of the ditch, only four of which produced more than fifty pieces. Potentially more significant are the forty flints recovered from the filling of the pit at the centre of M3 [F.2525]. Nearly all of these came from the pit’s main upper filling (Context 2524; 37 pieces) and consist of a complete two/three platform core (158g), a fragment from another core and thirty-five unworked flakes, mostly secondary waste (24), with seven primary and four tertiary flakes. These flints seem to be derived from a short-term working area or maybe one episode where the waste was not being re-used and someone was engaged in the early stages of tool manufacture (or object production) and then removed the object for finishing or use elsewhere. This is suggested by the low numbers of tertiary flakes compared with remaining secondary pieces. There is nothing to indicate that the material had been deliberately placed in the pit and more probably it comes from earlier, unrelated knapping activity nearby. Calcined flint: the excavation of Trench 9 produced a total of 256 fragments of calcined flint (4,351g) scattered throughout forty-two different contexts. None of this material had been burnt in situ and nearly one third of the total came from the filling the M3 ring-ditch, where it was probably all residual. A further forty pieces were discovered within the fillings of the probable grave pit [F.2525] at the centre of M3. Previous excavations at Ringlemere have demonstrated the abundance and widespread occurrence of calcined flint across the site, with about 68,000 pieces collected during the excavation of M1 (Parfitt and Needham 2020, 198). Dispersed Late Bronze Age metalwork hoard by Stuart Needham In the spring of 2004, metal detecting about 175m to the south-east of M3 produced a single fragment of a Late Bronze Age socketed axe in the plough soil. A very careful search of the area (NGR TR 2940 5679; Fig. 2) led to the recovery of a group of fourteen more metal objects, mostly fragments of bun ingot with some other copper-alloy waste and a single piece of gold wire. These items were found scattered over an area about 20m across, all contained within the thickness of the plough soil. Most of the pieces are likely to derive from a single Late Bronze Age hoard dispersed by the plough. A small excavation in the area indicated that no associated sub-surface features had survived. The collection was deemed to constitute Treasure under the terms of the Treasure Act (refs 2004 T85 and 2004 T146) and has accordingly been acquired by the British Museum as part of the Ringlemere collection. A report on RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 29 the hoard by Stuart Needham was included in the Treasure Annual Report for 2004 (DCMS 2007, 39, figs 25.1-25.3). The information below is based on that report. Catalogue of metalwork recovered 1. Length of thick gold wire or rod (RFW KF 350). The piece has a neat circular section except where chopped at either end. Cutting or chiselling from opposite ends has given the ends a wedge-like profile. The piece is 37.2mm long, with a diameter of between 2.86mm to 2.60mm. It weighs 3.27g. (DCMS 2007, fig. 25.1.) 2. Socketed axe fragment of copper-alloy. Mouth intact but blade and one face missing (RFW KF 358). Class A socketed axe decorated with wing-rib design. The bases of the wings continue onto the sides as a slight horizontal moulding. Length 69mm; weight 81.4g. (DCMS 2007, fig. 25.2.) 3. Plate-like fragment of copper-alloy. All edges broken; possibly part of an artefact. Max. dimension 46.5mm; weight 23.2g. (DCMS 2007, fig. 25.2.) 4. Copper-alloy tang fragment. Blunt ended rod of sub-trapezoid section; the other end thins and expands towards a bifurcation but the two ends are immediately broken. Length 36.5mm; weight 9.4g. (DCMS 2007, fig. 25.2.) 5. Copper-alloy ingot fragment (RFW KF 349) weight 434.5g. 6. Copper-alloy ingot edge fragment (RFW KF 351) weight 286.9g. 7. Copper-alloy ingot near-edge fragment (RFW KF 352) weight 623.2g. 8. Copper-alloy ingot edge fragment (RFW KF 353); weight 325.1g. 9. Copper-alloy ingot edge fragment (RFW KF 355); weight 611.1g. 10. Copper-alloy ingot edge fragment (RFW KF 356); weight 566.2g. 11. Copper-alloy ingot near-edge fragment (RFW KF 357); weight 343.8g. 12. Copper-alloy lump, craggy form; weight 4.3g. 13. Copper-alloy lump, craggy form; weight 5.3g. 14. Copper-alloy lump, small flattish disc; weight 6.5g. 15. Copper-alloy lump, flattish, rough; weight 2.5g. The socketed axe is the only securely datable object among the group, belonging to the Ewart stage of the Late Bronze Age, c.1000-800 bc. However, all but one of the remaining copper-alloy objects would be entirely consistent with such a date; the ingot fragments are typical of the form frequently encountered in Late Bronze Age hoards, while the smaller lumps, which are prills or waste, can also be matched in some contemporary contexts. The fragment of gold rod (1) is undiagnostic. Its composition is consistent with a Bronze Age date, but it is not necessarily exactly contemporary with the base metalwork. The final copper alloy object, the tang fragment (4), has not been recognised as part of a Bronze Age type and is most likely to belong to a later period and be unrelated to the main hoard. general discussion The excavation of Trench 9 in 2007 provided important new information concerning the Ringlemere site. One of the smaller ring-ditches (M3) was completely KEITH PARFITT 30 excavated, and a second (M2) was sampled, whilst the existence of another (M4) was disproved. A few finds of early flintwork demonstrate prehistoric activity on the site long before the Neolithic/Bronze Age monuments were established. The tranchet axe/adze provides further evidence for the Mesolithic activity previously identified in the area (Butler 2020, 11; Healy 2020, 151). Perhaps of greater interest, however, is the small collection of somewhat abraded flints with a mottled patina that includes a large flake of classic Levallois type. Together these pieces suggest Middle Palaeolithic activity at Ringlemere around 50,000 years ago. The full implications of these finds require considerably more thought and study. The complete excavation of ring-ditch M3 has indicated that it represents the plough-damaged remains of a small round barrow, presumably of Bronze Age date, although dating evidence is limited. Nothing of a central mound or underlying buried soil profile remained but the enclosing ditch did survive to a reasonable depth. This would seem to have silted up slowly through natural weathering, with no evidence to indicate that the ditch had ever been re-cut. No significant prehistoric deposits occurred around its circuit. Twenty features, largely undated, were located within the area enclosed by the M3 ring-ditch. It seems likely that many pre-date the construction of any barrow here, as was found with M1. Others, notably a possible fence-line, could be later. The density of features enclosed within the smaller area of M3 was rather less than that of M1. In part, this will be due to the destruction here of the original prehistoric land surface underlying the lost barrow mound. It may also be the case that the M1 mound had been specifically erected over a spot which had long been a focus of more intense activity than the area of M3. A centrally placed pit within the M3 ring-ditch [F.2525] quite possibly represents a grave, perhaps once marked by an upright wooden post. The acidic gravel subsoil had caused the complete decay of any associated human skeleton but the overall shape and size of the pit is consistent with that of a crouched inhumation. Apart from some waste flintwork, no obvious grave offerings were found within the pit and its identification as a grave must remain unproven. If not a grave, the pit could be interpreted as a post-pit dug for the insertion of the substantial D-shaped timber upright, probably half a tree trunk, perhaps intended as the central feature of the monument. How such a post might have related to any covering barrow mound remains unclear. It could have protruded through the top of a barrow thrown-up soon after the post had been erected; or perhaps separate phases are represented, with a mound being constructed only after the post had rotted and collapsed. Work on the adjacent ring-ditch M2 was rather more limited, but the available evidence indicates that this was a larger, more complex monument, encircled by a substantial ring-ditch, possibly with a north-eastern entrance and perhaps enclosing an inner ring-ditch (p. 7, above). The relatively close proximity of a Late Bronze Age metalwork hoard to an earlier group of barrows is of some interest and implies continued activity in the general area many centuries after these monuments were erected. A positive correlation between Late Bronze Age hoards and natural water sources, such as springs, streams and rivers, is widely recognised (Yates and Bradley 2010). The Ringlemere hoard lies about 400m south of the present-day springs which are the RINGLEMERE: INVESTIGATION OF PREHISTORIC RING-DITCHES, M2 AND M3 31 source of the Durlock Stream, apparently further confirming this observed pattern (Yates and Bradley 2010, 62, fig. 7). Habitation in the region during the later Bronze Age also seems to be indicated by some of the flintwork recovered locally (Butler 2020, 12; see above, p. 27). The re-cut boundary ditch located represents the first feature discovered at Ringlemere that can be positively dated to Roman times, although casual finds belonging to this period, including pottery, tile and coins, had been found previously. Collectively, the evidence now suggests that Romano-British activity in the region might have been quite extensive, whilst material of the preceding Iron Age strongly implies that cultivation in the vicinity began well before the Conquest. A radiocarbon date and occasional pot sherds from the area of M1 suggest ploughing in that area was already underway by the Early Iron Age (Macpherson-Grant 2020). A rare Continental-style brooch dated c.550-450 bc found about 200m north-east of M1 provides further evidence for Iron Age activity in the vicinity (Parfitt 2005). The Roman ditch line located apparently served as a boundary marker. Skirting the north-eastern side of M2 and M3 (Figs 5 and 9), the clear implication must be that associated barrows were still standing in the Romano-British landscape as visible mounds and continued to influence boundary locations centuries after their original construction. The ditch itself, separated slightly higher ground from the area containing M1. It was perhaps augmented by a length of wooden boundary fence in the area of M3. The nearest previously recorded Iron Age and Romano-British habitation site in the vicinity lies at the former Hammill brick pit near the Black Pond, some 600m to the south-west of M3. Several pits and ditches were noted there, with quantities of Early Iron Age, Late Iron Age/Belgic and Roman pottery recovered (Macpherson Grant 1999). The six or seven Anglo-Saxon graves discovered adjacent to M3 provide important additional information concerning the overall size of the cemetery previously located (Parfitt and Needham 2020, 141, fig. 4.2). In all, around sixty burials are now recorded, including six cremations but the full extent and date range of the cemetery remains to be determined. The presence of graves in the immediate vicinity of both M1 and M3 further implies the continued survival of visible barrow mounds here, providing focal points for these later burials. A single seventh-century sunken featured building cut into the mound of M1 (Parfitt and Needham 2020, 141, plate 4. III), together with scattered pottery and occasional metalwork finds indicate Anglo-Saxon activity occurred across the excavated areas between the fifth and the ninth centuries. A few unworn sherds of sixth- to seventh-century date are also recorded from the brick pit near Black Pond (Macpherson Grant 1999). acknowledgements Thanks are due to the landowners – the Smith family at Ringlemere Farm – who readily allowed access and took a keen interest in the fieldwork from the first. The 2007 excavations would not have been possible without the aid of a substantial grant from the Kent Archaeological Society, together with a smaller donation KEITH PARFITT 32 from the Battle and District Historical Society, both of which must be gratefully acknowledged here. Two full-time supervisors from the Canterbury Archaeological Trust (Keith Parfitt and Barry Corke) led the project but most of the excavation and finds processing work was carried out by volunteers from various local archaeological societies and other KAS members, together with a number of students from several universities. Without their hard work, very little would have been achieved and thanks are due to all concerned. Stuart Needham has provided notes on the late Bronze Age hoard and has also kindly commented on the main text, suggesting a number of improvements. The illustrations were prepared by Barry Corke, Peter Atkinson and Mike Halliwell. bibliography Birchenough, A., 2006, ‘Geophysical surveys 2002-2003’, in Needham, Parfitt and Varndell, pp. 3-6. Butler, C., 2020, ‘Flintwork from the environs of the Ringlemere complex’, in Parfitt and Needham, pp. 9-12. DCMS, 2007, Treasure Annual Report 2004 (London: DCMS). Gibson, A.M., 1986, Neolithic and Early Bronze Age Pottery, Shire Archaeology, p. 43. Healy, F., 2020, ‘Struck flint’, in Parfitt and Needham, pp. 149-198. Macpherson-Grant, N., 1995, ‘Early to Late Saxon Pottery’, in Blockley, K., et al., Excavations in the Marlowe Car Park and Surrounding Areas, The Archaeology of Canterbury V (Part II), pp. 818-897. Macpherson-Grant, N., 1999, ‘Iron Age and Later Pottery: Eastry’, Archaeologia Cantiana, cxix, 378-380. Macpherson-Grant, N., 2020, ‘Later prehistoric pottery’, in Parfitt and Needham, pp. 244-6. Myres, J.N.L., 1977, A Corpus of Anglo-Saxon Pottery of the Pagan Period. Cambridge, vols 1-2. Needham, S.P., Parfitt, K. and Varndell, G. (eds), 2006, The Ringlemere Cup: Precious Cups and the Beginning of the Channel Bronze Age (London: British Museum Research Publication 163). 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, Woodnesborough’, Archaeologia Cantiana, cxxv, 382-4. Parfitt, K. and Needham, S., 2007, ‘Excavations at Ringlemere Farm, Woodnesborough, 2002-2006’, Archaeologia Cantiana, cxxvii, 39-55. Parfitt, K. and Needham, S., 2020, Ceremonial Living in the Third Millennium BC: Excavations at Ringlemere Site M1, Kent, 2002-2006 (London: British Museum Research Publication 217). [Reviewed in Archaeologia Cantiana, cxlii (2021), 334-36]. Yates, D. and Bradley, R., 2010, ‘The Siting of Metalwork Hoards in the Bronze Age of South-East England’, Antiquaries Journal, 90, 41-72.