Later Bronze Age Cremation at West Cliff, Ramsgate

LATER BRONZE AGE CREMATION AT WEST CLIFF, RAMSGATE

gerald moody, nigel macpherson-grant and trevor anderson†

In early 2002 significant archaeological deposits dating from the Neolithic/Early Bronze Age, Iron Age, Romano-British, Medieval and Post Medieval periods were encountered in evaluation trenches excavated by the Trust for Thanet Archaeology in advance of a development of new houses within the garden of the former Bon Secours Nursing Home at Ramsgate (Fig. 1; NGR TR 3711 6433; Boast 2002). The features were concentrated within a walled garden in the north-west part of the site and south-east of the main building. In July/August 2002 an open-area excavation was targeted at these areas (Areas 1 and 2); nothing was found in Area 2 but in Area 1 the earliest features found were of later Bronze Age date – a group of pits containing pottery vessels and a shallow ring ditch, capped with a mound or cairn built of nodules of flint and chalk.

Because of their importance for the region this report describes only the later Bronze Age features along with a summary of the report on the ceramic evidence by Nigel Macpherson-Grant and a discussion of the human remains by the late Trevor Anderson. These extracts are part of a larger archive of reports generated by the archaeological works on the site. As the development involved the conversion of the existing nursing home into dwellings an historic building survey was carried out on the standing building which is of historical importance to the area because of its association with the childhood of Queen Victoria (Seary 2002). Iron Age and Romano-British features were also excavated and a full summary of the site and detailed reports on other aspects of the site are available in the Trust for Thanet Archaeology archives.

The former Bon Secours nursing home site is located on a promontory bordered to the east and west by two dry valleys, the terminal of the eastern valley forms the basin of Ramsgate Harbour at its southern end (Fig. 1). The area has been substantially truncated by coastal erosion and a cliff face now forms the southern limit of the promontory between Pegwell and Ramsgate.

The Bronze Age Cremations

The Bronze Age cremation features consisted of a group of pits containing truncated ceramic vessels, most of which had been inverted when deposited. The soil fills of the vessels contained small quantities of cremated human bone. A small ring gully, probably contemporary with the pits, may have provided a focus for the disposal of cremating human bodies. The gully was filled with soil and flints derived from a dispersed mound or cairn. A curvilinear ditch encountered on the eastern edge of the excavation (Fig. 1, Ditch A) might have been contemporary with these features, pottery from its lowest fills dated to the later Bronze Age and it was cut by the terminal of a later Roman ditch (Ditch B). The ditch could be part of a round barrow, or possibly an enclosure ditch associated with a settlement which may have influenced the positioning of the cremation deposits.

The Ring Gully

The ring gully was roughly circular in plan with a diameter of approximately 4m (Fig. 2 and Plate I). The feature had been heavily eroded and on the eastern side the profile was very shallow and irregular (Fig. 2, Sections 4 and 5). The rest of the gully had a variable profile, predominantly with steeply sloping sides breaking at the base to a narrow curve (Fig. 2, Sections 1-6). The gully varied in width from 0.6-0.8m and ranged in depth from 0.29m to 0.17m at the shallowest point. In the primary silt at the base of deposit 1061 (Fig. 2, Section 3) twenty sherds from a small thick-walled, Deverel-Rimbury type bowl of later Bronze Age date were found (Fig. 3, 1).¹ The bowl was deposited in the open gully and suggests an association between it and the surrounding pit groups which is not demonstrated by the stratigraphy (Plate II).

Densely packed flint nodules were present within the fill of the gully, set in a matrix of sandy silt or silty clay (Fig. 2, Sections 1-6). These were continuous with a dense spread of flints and chalk fragments that extended over and beyond the outer limits of the gully and were probably derived from an eroded mound or cairn that originally sealed the gully (Fig. 2). Flints were dispersed throughout the fill and they defined a fairly regular circle where they were retained within the upper edges of the gully [1067] indicating that it had only been partially filled when the mound began to slide into the cut. A ditch of the later Romano-British phase (not illustrated) had removed a small section of the flints in the fill where it cut through the feature. A ‘spur’ of flints in the later ditch fill extended to the north-east. As many as six or seven further contemporary thick-walled Deverel-Rimbury bucket urns were represented as sherds within the upper fills of the gully (one shown at Fig. 3, 2) and in the spur of the later ditch cut (one shown in Fig. 3, 3). These vessels may have been associated with the function of the gully, or perhaps further eroded cremation groups.

The Pit Deposits

In close proximity to the ring gully, five roughly circular pits were excavated (Fig. 2, A-E). Four pits were located to the north and north-east of the gully [cuts 1020, 1024, 1030 and 1093] and a single pit lay to the south-west [cut 1077]. In all but one of the group [cut 1093], the substantial remains of pottery vessels were recovered (Fig. 2, E) all but vessel 13 were inverted, with rim sherds generally present at the base of the cuts (Fig. 5, 13). A cluster of small fragments of base sherds from another vessel remained within cut 1093 suggesting that a similar single vessel or group had been present but was almost entirely lost. Each of the truncated vessels was lifted with its contents intact and excavated under laboratory conditions; despite the heavy truncation by later activity the deposits filling the vessels were shown to contain cremated human bone in variable densities.

Pit deposit A In a roughly circular pit [cut 1020], 0.6m diameter and 0.22-0.25m deep, three inverted flint-tempered jars were packed closely together (Fig. 3, vessels 4, 5 and 6; Plate III). Vessel 4 contained 140g in total of human bone distributed throughout its fills [deposits 1021, 1172, 1174, 1176 and 1178]; although there was a concentration of cremated bone in the middle of the pot [deposit 1174].

Cremated bone in vessel 5 amounted to 53g in weight and one element was identifiable as belonging to an individual at least 10 years old, the bone was again distributed throughout the deposit [1022, 1173, 1175, 1177, 1179 and 1180] with unevenly distributed concentrations throughout the vessel.

Vessel 6 contained 255g of bone, probably from an adult. A sample of the human bone from this vessel was radiocarbon-dated by the University of Waikato (WK 15116, see below). Two deposits were excavated from within the vessel [1126 and 1181] with a further deposit [1183] representing material from the main fill of the pit that had been worked into the vessel by worm and plant action.

A silty deposit [1023] that formed the uppermost fill of the pit sealed the vessels and their fills. A sample of this deposit produced a few bone fragments as well as a few charred wheat and barley grains.

Pit deposit B A very truncated pit [cut 1025], 1.12m diameter and only 0.1m deep contained another two to three fragmentary vessels (Fig. 4, 7-10). One vessel (Fig. 4, 8) also containing 15g of cremated human bone among mixed deposits of chalk nodules and silt from the later erosion of the feature [deposit 1025].

Pit deposit C Initially it appeared that a single inverted vessel (Fig. 4, 11) was contained within a roughly circular pit [cut 1030] with a maximum diameter of 0.75m. At the western edge the base of the pit was shallow and flat, with the appearance of a platform adjacent to the pot. The vessel itself was contained within a more restricted basal depression wedged in with flint nodules contained in the chalky deposit filling the pit [1029]. Human bone and small amounts of charcoal amounting to 99g were recovered from the fill [1031]; 80g of the material was contained in a single layer in the northern half of the vessel. A limb fragment within the cremated material suggested the presence of an adult. Examination of the loose sherds of pottery associated with the vessel and contained within the pit fill showed that a further two vessels (Fig. 5, 12 and 13) were represented, although their original association with vessel 11 cannot now be determined.

Pit deposit D The pit [cut 1077] was elliptical in plan with the longest axis measuring 0.56m and the shortest 0.5m; the cut survived to 0.35m in depth. The cut had a curved profile with the sides slightly undercutting the upper edge. A basal deposit [1076] of sandy clay incorporated flint nodules packing vessel 13 in the cut; unusually in an upright position rather than inverted like the rest of the vessels described. The two deposits [1182 and 1080] filling the vessel were composed mainly of flint fragments and chalk with occasional fragments of charcoal. Only two fragments of cremated bone (1g total weight) were recovered from the upper deposit [1080]. Additional sherds from a globular urn and possibly another bucket urn were present in the fill.

Pit deposit E Pit E is represented by an irregular elliptical pit [cut 1093] 0.7m wide along the longest axis and 0.45m along the shortest. The cut survived to only 0.1m deep and contained 20 sherds of Deverel-Rimbury type flint-tempered pottery from one or possibly two vessels of similar form to the vessel shown as 14 (Fig. 5, 14) within its clay silt filling. The pottery is contemporary with that contained in the other nearby pits suggesting that a similarly structured cremation deposit had been entirely destroyed.

Radiocarbon dating

A sample of one of the groups of human bone fragments [1026 not illustrated] found within the fill of the vessel shown as Fig. 3, 6 was submitted to the University of Waikato for an AMS Radiocarbon age determination (Result 3139 ± 34 bp Wk 15116). The calibrated result indicated ranges between 1520 bc and 1310 bc at 95% probability, and 1490 bc and 1320 bc at 68% probability.

the later bronze age pottery

The following summary of the ceramic material from the cremation groups has been extracted from a full ceramic report by Nigel Macpherson-Grant which provides further details on parallels for the vessels, and is available in the site archive. A general summary of the ceramic assemblage is presented as Appendix 1.

The pottery from the ring gully falls into two groups:

Group 1 is material that is directly associated with its primary use and material that represents disturbed cremations or re-deposited sherds introduced during later activity.

Group 2 pottery comes from the five sub-circular features assoc-iated with the ring gully; four associated with cremated bone (Pit deposit A-D) and one apparently without any human bone (Pit deposit E).

With the exception of the one or two fineware jars from Pit deposit E and another from Pit deposit D, all the pottery is coarsely flint-tempered, with no significant variations in fabric (other than minor differences in grit size and density between individual vessels), and a mixed range of oxidized, reduced or differentially-fired vessel surface firing colours, typical of most prehistoric assemblages.

Group 1 - Pottery from the Ring Gully

Directly associated

A small cluster of mostly conjoining and unworn sherds, representing approximately one-third of a small thick-walled bowl (Fig. 3, 2), was recovered from the basal ditch fill 1061 on the south-west side of the ring-ditch [Cut 1060], sealed by a layer of flint nodules presumably derived from the original central mound. No other pottery was recovered from primary fills in the other ditch sections. The position and unworn condition of this bowl indicate that it is probably associated with the primary use of the gully.

Indirectly associated

Two small sherds (including Fig. 3, 2) and some worn scraps were recovered from ditch section ¹⁰⁶⁴ on the north-east side of the ring-ditch. Another sherd was recovered from the general ring-ditch fill ¹⁰⁶⁷. Another 6 sherds (including Fig. 3, 3) came from the fill of the later Roman field-boundary ditch ¹⁰³⁴, where it cut across the ring-ditch on its north-eastern side. Overall, six or seven thick-walled bucket urns are represented. The oxidized fabric of one of these is similar to the vessel shown as Fig. 5, 12, and it may be derived from plough-reduction of Pit deposit C. It is possible that the other sherds come from a plough-reduced, primary inner, or secondary external, cremation deposit.

Group 2 - Pottery from the pit deposits

Pit deposit A

A sub-circular pit containing three cremation vessels, all containing cremated bone. Two (Fig. 3, 4-5), inverted and substantially complete but with their bases ploughed away; the third (Fig. 4, 6), was crushed but presumably also originally inverted. No other vessels were represented. A sample for C-14 analysis was taken from the human bone associated with vessel 6.

Fig. 3, 4. Almost straight-sided fairly narrow mouthed, thick walled bucket urn with off-set shoulder. Decorated with single rather irregular horizontal row of finger-tip impressions along shoulder off-set and two similar vertical rows of impressions down the body, from the off-set on one side only (Plate IV).

Fig. 3, 5. Bucket urn with slightly rounded shoulder and very slightly convex shoulder-rim zone. Slight lip evertion probably created during application of decoration. Three horizontal rows of impressed finger-tip decoration at the shoulder and rim externally and another inside the rim.

Fig. 3, 6. Large, closed-form, thin-walled, probably weakly convex-shouldered, bucket urn (cf. Fig. 4, 11 below). Two horizontal rows of impressed finger-tip decoration externally – one just beneath the rim, the other slightly above maximum girth. The latter has been applied discontinuously – first in one direction, then in another.

Pit deposit B

The severely plough-reduced sub-circular pit contained a few medium sized but mostly small sherds, many with heavy unifacial damage, together with a small amount of cremated bone. The crushed and severely abraded sherds were originally seen as one collapsed vessel although the high degree of disturbance made it uncertain whether the urn was originally inverted. Subsequent examination showed that two vessels were represented – a large urn (Fig. 4, 7) and a smaller vessel (Fig. 4, 8). In addition there was a single rim sherd (Fig. 4, 9) from a large urn and another base (Fig. 4, 10) from a vessel similar to vessel 8. Despite any potential for differential firing, 7 and 9 are unlikely to be from the same urn; however, even with their markedly different profiles, 8 and 10 may just be from the same pot. Irrespective, between 2-3 vessels are represented.

Fig. 4, 9. Thick-walled bucket urn rim, decorated with fingertip impressions.

Fig. 4, 7. Base from large bucket urn.

Fig. 4, 8 and 10. Base sherds from one (or two) small tub(s).

Pit deposit C

This roughly sub-circular pit superficially contained a single inverted vessel (Fig. 4, 11). Examination showed that in fact three vessels were represented. Thirteen sherds came from the upper fill – two from its plough-damaged base and eleven from another vessel (Fig. 5, 12). Another 12 sherds came from its lower fill – 8 from vessel 13 and 4 from another pot (not illustrated). All these sherds came from within urn 11 and there were no other discrete deposits of pottery from the cremation pit itself.

Fig. 4, 11. Large, closed-form, weakly convex-shouldered, bucket urn with two roughly applied lugs (one on either side) at shoulder. Rim top decorated with continuous finger-tip impressions, shoulder decorated (slightly above maximum girth) with one horizontal row of impressions, after application of the lugs.

Fig. 5, 12. Fairly small bucket urn with upper-body profile similar to 6 (possibly also shoulder decoration). Extant decoration : irregular finger-tip impressions around rim top.

Vessel 14. Vessel not illustrated, Fig. 5, 14 shown for comparison only. Bucket urn represented only by bodysherds.

Pit deposit D

Small sub-ovoid pit with sherds from 2-3 vessels and a few scraps of cremated bone. The majority of the sherds came from cremation vessel Fig. 5, 13 which, unusually, was not inverted. One other vessel was definitely represented by fragments from a globular urn with offset shoulder and a few sherds just possibly from another vessel were present.

Fig. 5, 13. Bucket urn with near-vertical straight-sided lower-body walls.

Pit deposit E

This shallow, plough-reduced pit containing a small quantity of sherds. No cremated bone was recorded. All the sherds were small or only moderate-sized and worn to varying degrees – some with heavy unifacial wear. One vessel of the general character of the illustrated parallel (Fig. 5, 14) is definitely represented, just possibly another.

Fig. 5, 14. One, possibly two, globular urn(s), represented by frag-mentary off set shoulder sherds only. The illustrated parallel is from Kimpton, Hampshire.

Dating

General parallels, for most of the Bon Secours formal and decorative ele-ments, can be found amongst broadly contemporary southern or south-eastern English later Bronze Age assemblages. On the basis of general nat-ional dating trends but, more specifically, associations with local metalwork hoards, the main floruit of the regional Kentish Deverel-Rimbury tradition could be placed between c.1500-1100 bc, with some overlap at either end (Macpherson-Grant 1992, 62-3). This does not seriously conflict with a recent re-appraisal of both the scientific and metalwork dating that could be applied to Bronze Age traditions (Needham 1996, figs. 2-3) – except for a shift in the tradition’s floruit to between c.1600-1150 bc, again with an overlap at either end. This modification has been applied to the main typological dating given in Appendix 1; Table 1.

On a detailed inter-assemblage comparison, however, there are very few close vessel equations. The closest are to Bridge, Monkton or Kimpton, via the tub forms of 8 and 10, distinctive in their own way but a simple and long-lasting type, and rather more tenuously via the urn 6. More specifically, the form and lugs of bucket urn 11 are broadly similar to Kimpton F8 but again the equation is not that close. Another indirect link is between the globular urn fragments from Pit deposits D and E and two sherds from the Monkton Area 7 Deverel-Rimbury cemetery, a rim from a globular urn and a thin-walled probable bucket urn with an off-set shoulder. The most distinctive element from Bon Secours is the decoration on the bucket urn 4. It does have very general parallels with contemporary East Anglian material, but its thin vertical band of thumb-presses down one side only (rather than balanced and on both or more sides) is peculiar and not really akin to the exuberant decorative thumbing from Essex and Suffolk.

Rather more usefully there are broad generic links, via the bucket urn 12, to the Netherhale Farm assemblage from western Thanet, with its likely link to the sophisticated dot-and-ring decorated Birchington bowl (Macpherson-Grant op. cit., 60 and fig. 6). It is uncertain whether this truly globular form occurs alongside the more jar, or tub-like ‘globular’ urns normally associated with the Deverel-Rimbury tradition, or replaces them. Irrespective, it has a hoard-associated date of between c.1300-1100 bc. If comparisons between scientific and typological dating can be relied upon, then the end of the Bon Secours C-14 dated span may overlap with the beginning of the likely currency of these bowls. If this equation is valid then it is all the more useful to note the presence at Bon Secours of sherds from at least 2-3 traditional globular urns with (as at Kimpton Phase E) fairly characteristically off-set shoulders. This association between Kimpton and the Bon Secours dating usefully places the latter within the C-14 date obtained for Kimpton Phase E and earlier than either the Birchington hoard bowl or the cremation urns from Bridge.

The Bon Secours radiocarbon determination is the second for Deverel-Rimbury material from the former Isle of Thanet, the first being obtained for soot from the body of a Cornish Trevisker Ware jar from Monkton Area 3, Ring-ditch X (Bennett et al. 2008, 54). At 94.5% probability it compares well with the slightly broader Monkton date and falls well within the quoted published dates for Kimpton, Hampshire (Phases D and F), but outside that for Bridge, Kent. The chronological relationships between the scientific and typological dates quoted in Appendix 1; Table 1 are self-evident but it is worth stressing that the present radiocarbon date suggests that, at least part of the formal content of the Bon Secours assemblage pre-dates the typologically-dated currency for forms like the Birchington bowl and the Netherhale Farm assemblage, it may have a pre-c.1300 bc currency emphasis and therefore is within the earlier part of the Deverel-Rimbury tradition. Establishing whether this difference represents a diagnostically useful formal/chronological watershed between the earlier and later phases of the regional Deverel-Rimbury tradition would depend on obtaining a larger series of scientific dates coupled with a comprehensive review of local and Kentish Deverel-Rimbury assemblages.

the human bone

The following discussion of the human remains was extracted from a full report by the late Trevor Anderson who provided considerable assistance in lifting and subsequently processing and reporting of the cremation groups excavated at the development at the Bon Secours nursing home site. The vessels in each of the cremation groups were lifted with their contents intact and the deposits within the vessels was carefully excavated and recorded under laboratory conditions allowing the dispersal of the bone and soil to be measured accurately. Further details are held within the archive, which also includes the records of the excavation of the vessels and the full description of separate deposits and human remains found.

The Material

Pit deposit A

Vessel 4 This inverted vessel contained 140gms of cremated bone in five distinct horizontal layers identified in the fill. The majority of the bone (105gms), including the largest fragments (up to 25mm in length) and three teeth roots, was recovered from two layers in the upper-mid portion of the vessel, originally nearest to its base before it was inverted. The uppermost layer of bone, was 2cms thicker on the north side of the vessel. The base of the lower layer of bone was horizontal. The material consisted mainly of small (under 10mm) eroded white bone fragments, with approximately 16%, by weight (22gms), predominantly blue in colour. The majority, almost ninety per cent, of the fragments could not be firmly identified but five tooth roots, all incisors, were recognised. The apices of three central incisors were available for examination indicating that they derive from an individual over 10 years of age (van Beek, 1983, table 3). Sex diagnosis was not possible.

Vessel 5 The inverted vessel containing 53gms of cremated bone. The fill was separated into six distinct layers, three-quarters of the material, including the largest fragments (up to 30mm in length), were found in the upper and mid portions of the vessel, nearest to the base of the vessel. The material consisted mainly of small (under 10mm) eroded white bone fragments. A single rib fragment and a mandibular incisor root were identified. Age and sex diagnosis was not possible.

Vessel 6 An inverted vessel containing 255gm of cremated bone, although the bone was scattered throughout the vessel but two separate layers were distinguished. The material consisted mainly of small (under 10mm) eroded white bone, with less than 2%, by weight (4gm), predominantly blue in colour. The majority, some eight-five per cent, of the fragments could not be firmly identified. The fragment of a cervical vertebra, in which the end-plate was fully fused, suggests that this bone was from an adult. Sex diagnosis was not possible.

Pit deposit B

This pit contained a roughly circular deposit of pottery fragments, classed as a single vessel although fragments of two or three other vessels were identified. The fill was sub-divided into a northern and a southern half.

Vessels 7-10 Only 15gm of bone was recovered. The material consisted mainly of small (under 10mm) eroded white bone. In the northern half of the fill, one fragment 25mm in length was recovered. Age and sex diagnosis was not possible.

Pit deposit C

This pit contained part of an inverted fragmented pottery vessel and sherds from two other vessels. The uppermost portion had been truncated and was not recovered. The fragmented vessel had cracked and separated into two approximately equal halves. The two halves were separated and the section was cleaned, photographed and planned. The pottery was carefully removed from the internal fill. The fill was separated into four quadrants and wet-sieved without further excavation.

Vessel 11 A total of 99gm of cremated bone was recovered from the four quadrants and the cleaned section. The majority of the bone (c.80gm), was recovered from a single layer, largely confined to the northern half of the vessel and located near the rim of the inverted vessel. The material consisted mainly of small (under 10mm) eroded white bone fragments, with only 3%, by weight, predominantly blue in colour. The largest fragment was 30mm in length. The majority (90%) of the fragments could not be firmly identified but a mandibular and a maxillary incisor root were recognised. The latter apex was intact and complete, indicating a minimum age of 10 years (van Beek, 1983, table 3). A lower limb fragment is probably from an adult. Sex diagnosis was not possible.

Pit deposit D

This pit contained an upright pottery vessel. Fragments of the broken and damaged base were found mixed in the fill, which was divided into two deposits.

Vessel 13 Only two minute cremated bone fragments were recovered from sieving the excavated fill. Age and sex diagnosis was not possible.

Deposition of the human bone

The relatively insignificant amount of bone recovered, as well as its uniformly high fragmentation, must lead us to question if any of the samples represent the deliberate deposition of a cremated human body. Certainly, the minute amount of bone recovered from the only upright vessel (13) indicates that it was not intentionally used as a container for a cremated body. In this case, the two small fragments, both in the upper layer, entered over the centuries as the surrounding soil infiltrated the vessel. Only fragments of the lower portion of vessel 8 was recovered, the fact that it contained charcoal and animal bone mixed with very little human bone (15gm), supports the view that this also is post-depositional contamination from the surrounding soil.

The other four inverted vessels are largely complete, missing only their uppermost portions. The available vessels were not crushed and the absent pottery was not recovered within the fill. The vessels were damaged at the same level, truncated mechanically after deposition and the broken sherds appear to have been scattered and were not recovered. In rare cases there is evidence that the base of an inverted vessel was deliberately removed to permit the insertion of cremated bone (Patchett 1950) or that the vessels were damaged during the backfilling of the grave pit (Barnatt 1994). Assuming that the pots were complete when deposited over cremated bone, it would originally be located as a pile of fragments at the lowest portion of the otherwise empty vessel (Bristow 1998, appendix 2: sites: 774, 883, 1149). Once an inverted vessel was damaged, or if it was upright and uncovered, the surrounding soil could infiltrate. The action of burrowing animals and earthworm activity would lead to the bones being mixed within the newly infiltrated soil (Anderson and Fell 1995). In one vessel earthworm activity was confirmed by the presence of not only their circular tunnels but also by the fact that dead desiccated, modern worms were recovered. The degree of soil movement can be identified in vessel 13, where large fragments of the damaged base were displaced into the fill of the urn.

Computerised tomography of soil-filled urns suggests that the bone fragments tend to remain in the lower half of the vessel (Ibid.). Only, in the case of vessel 12 is there a definite concentration of bone at the lowest level, this may represent a deliberate deposit of cremated human bone under an inverted urn, similar to that seen at Ringwould, Dover (Bristow 1998, appendix 2: site 883; Grinsell, 1992). If this interpretation is correct, the modest amount would indicate that it represents only a token deposit. Similarly small, symbolic, deposits have been recorded under inverted urns at Woodsford, Dorset (Bristow 1998, appendix 2: site: 1679). The upper fill of vessel 12 contains large fragments of charcoal, apparently pyre debris. The fact that it is not mixed with the clean layer of bone, suggests it may be pyre material that has infiltrated from the surrounding soil, urns apparently deliberately inverted over pyre debris are known from Wiltshire, Dorset and Cornwall (Bristow 1998, appendix 2: site: 588, Calkin 1962; Christie 1960).

Deposits containing multiple inverted pots appear to be a rare finding. At Paitford, Hampshire, two urns (both complete and inverted over cremated bone) were found side-by-side near the centre of a bowl barrow (Preston and Hawkes 1933). In the group of three vessels on this site, the bone was largely in the upper or upper-mid portions (vessels 4 and 5) or else scattered throughout the fill (vessel 6). Charcoal was found in two of the urns (vessels 4 and 6) and the latter also contained calcined flint. This coupled with the highly fragmented nature of the bone, as well as the small sample size, suggests that all three vessels contain post-depositional infiltration of pyre debris.

The fact that the uppermost portions of the vessels are largely devoid of bone does not negate the argument that pyre debris has collapsed into the vessels from above. It must be remembered that the freshly broken urns would present as a void for the overlying and surrounding soil to collapse into. Certainly, there is evidence from other Bronze age sites that pyre debris was scattered over cremation urns which had been buried under barrows (Bristow 1998, appendix 2: site 520, 773, 1082; McKinley 1997). Also urns, apparently containing only pyre debris; either ash (Bristow 1998, appendix 2: sites 587, 606, 771; Clay 1928; Green et al. 1982; King 1989; Matthews 1976) or charcoal (Bradley and Richards 1979-1980) have been recovered.

Consideration was given to the possibility that the urns had originally contained a deliberate cremation deposit and had been buried elsewhere in an upright position. After several years they would be full of post-depositional soil as well as the original bones, near the base (Anderson and Fell 1995). If they were then removed and re-buried in an inverted position the cremated bone would be lost when the bases were truncated by later damage and disturbance. However, if this was the case, the expected pattern of large bones confined to the uppermost levels with smaller bones below them (Anderson and Fell, 1995) was not seen. Also, there was no evidence that any of the vessels contained alien soil. The careful excavation and recording of the internal bone layers in vessel 4 suggest that it, similar to the other two urns, was originally buried vertically. Apparently, post-depositional disturbance was responsible for it tilting towards the other vessels. Evidently, some bone had entered while the damaged vessel was vertical and the latest, uppermost, bone layer had accumulated after the vessel had been displaced, the vessels may have been subjected to repeated episodes of damage (apparently moving from north to south).

All the evidence negates the possibility of secondary burial for the vessels. Primary burial of inverted urns over cremated bone material is well documented from Bronze Age Britain (Bristow 1989, appendix 2). Indeed, the practice is a very frequent finding in Kent, urns inverted over cremated human bone being known from Chartham (Iffin Wood); Dover (Capel-le-Ferne and Ringwould), Ramsgate (Lord of the Manor), Tonbridge and Malling (Wouldham) and Wickhambreaux (Grinsell, 1992; Perkins and Macpherson-Grant 1981, 16). Based on a long-lived large cemetery in Hampshire, there is some evidence that the practice of protecting cremated bone with upturned vessels became most frequent in the later middle Bronze Age (Dacre and Ellison 1981).

The largest amount of bone (255gm) was found in vessel 6. Unlike the other two vessels, the bone was scattered throughout the urn and not in definite layers, suggesting that it represents the remnants of a deliberate cremation deposit. However, it is the largest vessel and, consequently has the greatest volume for soil and bone to infiltrate. Indeed, the ratio of bone to fill is slightly lower than that of vessel 4 which has less bone. Also, if it was inverted over an incomplete deliberate deposit of cremated bone, one would expect to see the larger bone elements near the rim or on the ground surface on which the urn had been placed.

The paucity of cremated bone in each vessel is emphasised by the fact that in modern crematoria a skeleton will yield 1,600-3,600gms of bone (McKinley, 1989). In archaeological contexts, it has been estimated that approximately 40-60% of the burnt bones were recovered and deposited within the cremation grave (McKinley 1997); there is some evidence that a higher level of recovery was obtained if the bones were placed within a pottery vessel, as opposed to being un-urned (McKinley 1997). At Twyford Down, the mean weight of cremated bone recovered from within pottery vessels was 1,642.4gms (McKinley 1997). Cremation graves, under the centre of Bronze Age barrows, consistently display a higher recovery rate (McKinley 1997). A broadly similar level of recovery is known in Roman cremation material – mean bone weight 1,317gm, with a range from 859 to 1,959gm (Anderson 1998). At the same site, Each End, Ash, the bone/fill weight ratio for seven vessels, ranged from 6.1 to 23.3, with the mean being 14.8. Which is in strong contrast to the low bone/fill ratio (0.04-3.05) seen here.

Careful scientific micro-excavation of the soil from within the vessels has permitted us to establish the exact location of the cremated bone within the vessel. This evidence, coupled with the paucity of the bone, its highly fragmented nature, as well as the recovery of burnt wood and flint, suggest that the contents of the urns do not represent the deliberate deposition of a cremated body. One vessel (13) was definitely empty of bone when deposited. Only one vessel (12) appears to have been inverted over a symbolic, token, sample of cremated human bone. The other vessels, all damaged and including three in one pit, contain post-depositional soil infiltration, which possibly includes pyre debris that had originally been scattered over the vessels.

general discussion

The small group of features at the Bon Secours site is one of the few later Bronze Age sites to be excavated by modern archaeological methods either on the Isle of Thanet or in the east Kent region (Perkins 1999). The evidence provided by the careful excavation of the contents of the urns suggests that if they represent a cultural practice particular to this period, the actions involved were complex and not always uniform.

It appears that the majority of the vessels in the primary deposit were inverted, usually in groups within a pit not much larger than the vessel group. In some cases they may have been deliberately placed to cover a deposit of existing pyre debris. In most cases it appears that pyre debris overlaid the vessels, possibly forming the primary filling of the pits, spread over the vessels. The infiltration of human bone and charcoal from the pyre as recorded seems to have been the result of a secondary process, associated with the disturbance and natural decay of the original assemblage. In response to the relative lack of cremated bone held within the urns, Trevor Anderson suggested they may have been used in association with a token deposit taken from pyre material from elsewhere. It is possible the inverted urns could have covered some other perishable material and the pyre material infiltrated after truncation. The presence of a single upright pot among the remaining inverted vessels hints at a variation in rite that may indicate that a traditional practice could occasionally re-invented, however this may be overextending the interpretation of the evidence of this site.

If the debris from the cremations of human bodies was deposited in-formally over the vessels, the bone may be derived from a communal pyre or at least from cremations that cannot with certainty be assigned to each of the structured groups of vessels. This communality and dispersal may account for the sherds of similar urns found within the assemblages, possibly representing disturbed material from earlier phases. The evidence suggests that the cultural significance of the act of depositing the groups of vessels should be separated from that of the process of cremation and disposal of individual human bodies. It may be that the association between the vessel assemblages and the processing of human remains is spatially coincident but not culturally connected. Perhaps the structuring of the pottery groups was a commemorative action that took place within the limits of an area reserved for the act of cremation by a larger cultural group. The ring gully capped with its cairn, sealing a similarly deposited whole vessel, may demonstrate the independence of this set of actions.

The association between the vessel groups and the gully and their location within with a space reserved for cremation suggests the model of social organisation proposed by Ellison (1979) for the later Bronze Age, representing a cremation urn field of limited extent, reserved for a family or clan group and directly associated with a settlement. The pit groups and cairn are close to a curvilinear ditch that may be an element of a contemporary settlement enclosure or possibly a round barrow of earlier Bronze Age date. It appears that cremations superseding the tradition of burial into round barrows in the later Bronze Age (Green 2000). Ellison cautions against this simple model as there is evidence that the barrow tradition continues, as does a hybrid rite of the raising of symbolic barrow mounds into which urns or inhumations are inserted. The changing pattern of disposal of the dead may reflect issues of social stratification as well as a general change in cultural attitudes. At the same time there is evidence that the landscape was being structured and divisions were becoming formalised and demarcated with structured boundaries. As the divided landscape accumulated more complex social functions, cultural restrictions over rights of use may have increased. These restrictions may have influenced the capacity of a community to raise monumental round barrows as freely as was once possible in traditional locations (Parker Pearson 1993). Existing round barrows may have been curated in the landscape as enduring monuments to ancestors of a family, clan or community (Moody 2008) and these places may have been regarded as appropriate locations for the developing rite of cremation, perhaps by those who found themselves disadvantaged by political geography of the restructured landscape. The ring gully itself is too small to be considered a conventional round barrow but examples of very small round barrows are emerging in Thanet, particularly the ring ditch containing multiple inhumations discovered at Bradstow School, Broadstairs in 2006 (Hart 2006). The cairn covering the ring ditch has affinities with larger flint cairns re-used to insert cremation burials of this period. It is possible that the ring gully and cairn could be considered a ‘token’ barrow mound around which cremated material and vessels were deposited.

Appendix 1. Summary of the ceramic assemblage

Including pottery from features not discussed in this report, a total of 1,243 sherds (weighing 11,401gm) of all periods were recovered from the evaluation and excavation phases of this site. The overall multi-period assemblage inevitably contained a mixed range of wear patterns – from both small or medium-sized worn sherds (redeposited in contemporary contexts or residual in later ones) to large relatively unworn sherds – the latter solely from contexts associated with the cremation groups. Sherds were macroscopically identified and dated according to the regional Fabric Type Series used by the Canterbury Archaeological Trust.

The uncertain Later Prehistoric element includes 4 sherds in a coarsely grog-tempered fabric that could be from either earlier Bronze Age urns or from LIA ‘Belgic’-style vessels. Since most of these are from the lower and middle fills of Ditch A, also containing mostly Deverel-Rimbury-type sherds it is tempting to suspect that this enclosure might be of mid-later second millennium bc date but there is far too little material to be confident, and since the rest of the mid-upper fills contain Iron Age and later fabric types, these sherds and the enclosure’s construction-date have been placed into the Late Iron Age – particularly in view of the typically S-profiled jar and curvilinear-decorated sherds from its upper fill and Evaluation Trench 18 that cut this feature.

The presence of some activity during the earlier Iron Age (Phase 4) is confirmed by the presence of re-deposited sherds from the mid-upper fills of the ditches A-B – particularly a scrap from an angle-shouldered bowl and some worn thicker-walled (non-Deverel-Rimbury type) flint-tempered coarseware sherds. Some of the more worn uncertainly allocated Later Prehistoric sherds may stem from this phase. Irrespective the overall count is low and should indicate agricultural activity marginal to any associated settlement area.

The relatively high number of ‘Belgic’-Early Roman sherds in Table 1 are attributed to a large number of sherds from a single local Conquest-period ad Thanet silty ware jar. An imported Gallo-Belgic Terra Nigra platter from the Marne-Vesle valley, with English currency dated to between c.10 bc- ad 50 (pers. comm. A.J. Savage, Canterbury Archaeological Trust), and represented by a sherd from the upper fill of the later Roman ditch B.

Roman pottery recovered from the site as a whole indicates that the main phase of occupation in the immediate area lasted until the late second century, though a single sherd of Late Roman Oxfordshire colour-coated ware indicates does indicate continuity of, at least, land-use until the late third or fourth century ad. There is no definite activity in this area until the Medieval period, represented mostly by small worn sherds in Canterbury Tyler Hill sandy ware. The condition of these sherds suggests agricultural activity between c. ad 1200-1350 – with most sherds probably arriving in place via manure spreads. The degree of later activity is self-evident (Table 1).

endnote

¹ Within the text the number given to the vessels (i.e. vessel 1) refers to their order in the illustrations, Figs 3-5. A concordance with the original vessel numbering and small find numbers assigned to the pottery in the excavation and in the subsequent analysis is contained in the archive.

bibliography

Anderson, T., 1998, ‘The Human Bones’, in Hicks, A., ‘Excavations at Each End, Ash’, Archaeologia Cantiana, 118, 91-172 (pp. 119-130).

Anderson, T. and Fell, C., 1995, ‘Analysis of Roman cremation vessels by computerised tomography’, Journal of Archaeological Science, 22, 609-17.

Barnatt, J., 1994, ‘Excavations of a Bronze Age unenclosed cemetery, cairns and field boundaries at Eagleston Flat, Derbyshire’, PPS, 60, 287-370.

Bennet, P., Clark, P., Hicks, A., Rady, J. and Riddler, I. 2008. At the Great Crossroads. Prehistoric, Roman and medieval discoveries on the Isle of Thanet 1994-95, CAT Occasional Paper No. 4.

Boast, E.J., 2002, ‘Former Bon Secours Nursing Home, Junction of London Road and Pegwell Road, Ramsgate, Kent. Archaeological Evaluation’ (unpubl. TTA report).

Bradley, J. and Richards, J.C., 1979-1980, ‘The excavation of two ring ditches at Heron’s House, Burgfield’, Berkshire Archaeology, 70, 1-7.

Bristow, P.H.W., 1998, Attitudes to Disposal of the dead in Southern Britain 3500BC – AD43, BAR 274.

Brown, N., 1999, ‘The Archaeology of Ardleigh, Essex: Excavations 1955-1980’, East Anglian Archaeology, 90, 76-116.

Calkin, J.B., 1962, ‘The Bournemouth area in the middle and late Bronze Age with the Deverel Rimbury problem re-considered’, Archaeological Journal, 119, 1-65.

Christie, P.M.L., 1960, ‘Crig-A-Mennis: a Bronze Age barrow at Liskey, Perranzabuloe, Cornwall’, PPS, 26, 76-97.

Clarke, A., 1982, ‘The Neolithic in Kent’, in P.E. Leach (ed.) Archaeology in Kent to AD 1500, CBA Research Report No. 48, London.

Clay, R.C.C., 1928, ‘The excavation of an oval barrow in Hadden’s Hill Plantation, Bournemouth’, Antiquaries Journal, 8, 87-89.

Dacre, M.W. and Ellison, A.B., 1981, ‘A Bronze age urn cemetery at Kimpton, Hampshire’. PPS, 47, 147-204.

Dunning, G.C. 1966, ‘Neolithic Occupation Sites in East Kent’, Antiquaries Journal, 46, 1-25.

Ellison, A .,1979, ‘Deverel-Rimbury urn cemeteries, the evidence for social organ-isation’, in Settlement and society in the British later Bronze Age, BAR 83.

Gibson, A.M., 1990, ‘The Beaker from Cottington Hill, Ebbsfleet, Ramsgate, Kent’, report no. 4 for TTA.

Gibson, A.M., 1995. ‘The Neolithic Pottery from Cross Roads and St Richard’s Road, Deal, Kent’, report no. 21 for the CAT.

Gibson, A.M., 1996, ‘The Neolithic and Beaker Pottery from Laundry Road, Minster, Kent’, report no. 37 for TTA.

Gibson, A.M., 2001, ‘Ramsgate Harbour Approach Road Assessment of The Neolithic and Early Bronze Age Pottery from the Causewayed Enclosure’, report no.68 for the CAT.

Gibson, A.M., 2003, ‘The Neolithic Pottery from the Causewayed Enclosure at Kingsborough Farm, Isle of Sheppey, Kent’, report no.72 for AS-E.

Green, C.J.S., Lynch, F.M. and White H., 1982, ‘The excavation of two round barrows on Launceston Down, Dorset (Long Crichel 5 and 7)’, Proc. Dorset Natural History and Archaeological Soc., 104, 39-58.

Green, M., 2000, A Landscape Revealed, 10,000 Years on a Chalkland Farm, Tempus.

Grinsell, L.V., 1992, ‘The Bronze Age barrows of Kent’, PPS, 58, 355-384.

Hart, D., 2006. Excavations at Bradstow School, Dumpton Park Drive, Broadstairs, Kent. A post-excavation assessment report and proposals for publication, unpubl. AS-E report.

King, A.C., 1989, ‘A Bronze age cemetery at Oliver’s Battery, near Winchester and some related finds’, Proc. Hampshire Field Club and Arch. Soc. 45, 13-23.

Macpherson-Grant, N., 1968, ‘Two Neolithic Bowls from Birchington, Thanet’, Archaeologia Cantiana, 83, 249-50.

Macpherson-Grant, N., 1980, ‘Archaeological Work along the A2 : 1966-1974’, Archaeologia Cantiana, 96, 133-183.

Macpherson-Grant, N., 1992, ‘A Review of Late Bronze Age Pottery from East Kent’, Canterbury’s Archaeology 1991-1992, 55-63.

Matthews, C. L., 1976, Occupation Sites on a Chiltern Ridge. Excavations at Puddlehill and Sites near Dunstable, Bedfordshire: Part I Neolithic, Bronze Age and Early Iron Age, BAR 29.

McKinley, J., 1989, ‘Cremations: expectations, methodologies and realities’, in Roberts, C.A., Lee, F. and Bintliff, R. (eds), Burial Archaeology: Current Research Methods and Developments, BAR S211, pp. 65-76.

McKinley, J., 1997, ‘Bronze Age barrows and funerary rites and rituals of cremation’, PPS, 63, 129-145.

Needham, S., 1996, ‘Chronology and Periodisation in the British Bronze Age’, in Randsborg, K (ed.), Acta Archaeologica, 67, 121-140.

Parker Pearson, M., 1993, Bronze Age Britain, Batsford/English Heritage.

Patchett, F.M., 1950, ‘Cornish Bronze Age pottery. Part II’, Arch. Journal, 107, 44-65.

Perkins, D.R.J., 1999, ‘A Gateway Island’, unpubl. Ph.D. Thesis, University College, London.

Perkins, D.R.J. and Macpherson-Grant, N., 1981, The Isle of Thanet Archaeological Unit Interim Excavations Reports 1977-1980, Thanet.

Preston, J.P. and Hawkes, C.F.C., 1933, ‘Three late bronze age barrows on the Cloven Way’, Antiquaries Journal, 13, 414-454.

Seary, P., 2002, ‘Building Recording at Bon Secours Nursing Home, Ramsgate, Kent’, unpubl. CAT report.

van Beek, G.C., 1983, Dental Morphology: an Illustrated Guide, Bristol.

Fig. 1 Site Location Plan.

PLATE I

Ring gully before excavation showing flints in fill

Fig. 2 Plans and Sections of Flint-Filled Ring Gully.

PLATE II

North-west facing section 1060 showing Vessel 1 sealed under flints

Fig. 3 Vessels 1-6.

PLATE III

Pit deposit A viewed from south

Fig. 4 Vessels 7-11.

Fig. 5 Vessels 12-14.

PLATE IV

Pit deposit A under excavation showing finger tip decoration on Vessel 4

TABLE 1. PERIOD SHERD TOTALS

Period	No. Sherds
Early Neolithic	4
Early Bronze Age	1
Later Bronze Age	1,023
Early/Middle Iron Age	15
Indigenous Late Iron Age	33
Late Iron Age ‘Belgic’ Style	19
‘Belgic’ Early Roman	78
Roman	32
Medieval	8
Post Medieval	2
Uncertain	28
Total	1,243