Excavations at Barton Hill Drive, Minster-in-Sheppey

EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY MICK DIACK With contributions by Enid Allison, Robin Bendrey, John Cotter, Peter Couldrey, Ian Riddler and Beccy Scott Canterbury Archaeological Trust (CAT) undertook a programme of archaeological work in connection with the development of Sheppey Community Hospital, Barton Hill Drive, Thistle Hill, Minster, Sheppey (NGR TQ 9430 7220), from the 21 May to 4 July 2001. Four principal areas were machine stripped under archaeological supervision in advance of the construction of the hospital. These consisted of an access road for the new hospital, a car park area, the footprint for the hospital building and a footpath with services running alongside it. The line of this footpath was also used as a haul road during the ground works relating to the hospital's construction. The site had previously been evaluated by CAT in August I 998 (Pratt 1998a) and suggested the presence of an Early Iron Age farmstead, perhaps dating from the Late Bronze Age/Early Iron Age period, with a focus probably located in the valley between Rape Hill and Thistle Hill. Pottery recovered also indicated possible Late Iron Age/Roman, Anglo-Saxon and Norman activity in the locality. The excavation revealed a primary phase of Late Bronze Age/Early Iron Age activity (c.850-600 BC) that consisted of five linear features, presumably representing ephemeral remains of at least two enclosures, and a small quantity of pits and post-holes suggestive of contemporary occupation. A second phase of Late Bronze Age/Early Iron Age activity characterised by a number of possible sunken-featured structures was also identified. A small quantity of Roman and AngloSaxon material was recovered from the site, but no securely dated features were identified. Medieval activity (c.1050-1225) was represented by a 13m length of ditch at the southern end of the site. This ditch had been re-cut. A rectangular feature and a small pit containing quern stone fragments also dated from this period. 265 17225m. 0 25m Settle: I: I 000 17220m. I i; ij ; ; i ; i i i ·', ·,.. ·,.. Ditch I ·,i 172Um. ,., ,.,.,· ·:·􀀈 ,,,.,· .,.., . .,,.. Haul Road --.L.,· .,/' ·' ·' .,. · ,.,.,_ Northern 98 93',􀀔 9S 91 􀀡., 0 \ '· '·,.,_ yea Di1ch1 X ·, ..... ,. I /, . ,. i l i ,. '·,-, ,-,· ,/ '· '/ .,. .1 ' ' Ditch 5 Phase One: • \ ,/ Prehistoric 76 ;-· ., I Car park / f i area 􀀔\ \Ditcb6 I ,., . 42{ ' 30 /, :-. ••Z'd i.v· ' 13 •• \,f .,·' ""' ,., Ditch 3 I i ; ,., . ,. Phase Two: Prehistoric Phase Three: Medieval Phase Four: Post-medieval Un-phased ■ ffJ 1111 Ii 0 Fig. 1 Barton Hill Drive, Minster-in-Sheppey: Location and Plan of Excavation area. 266 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY Previous archaeological discoveries in the vicinity include a large, multi-period site at Power Station Road (TQ 9757 7230) located about 1.5km north-west from Rape Hill. The principal phase of occupation was of Late Bronze Age/Early Iron Age date (Pratt 1998a). Early Iron Age (or possibly Late Bronze Age) occupation has also been identified on a site near Minster Abbey (TQ 9555 7305 and TQ 9563 7305). The same site also identified Romano-British occupation (Bennett 1994; Pratt 1995). Excavations at Shrubsoles Hill (TQ 968 717) revealed unstratified Mesolithic and Neolithic flints, together with evidence for a ditched Middle Bronze Age enclosure. Late Iron Age/Roman activity was also noted, as was limited Anglo-Saxon activity (Pine 1999; 2000; 2001 ). A causewayed Neolithic enclosure is known at Kingsborough, near Eastchurch (TQ 9757 7230). The same site provided evidence for a Late Bronze Age/ Early Iron Age enclosure and Roman and medieval features (Dyson et al. 2000). Romano-British burials are known from the site of Sheppey High School, located to the north-west (TQ 9378 7265) (Leach 1969). The site is also approximately I. 75km south from Minster Abbey (TQ 957 730) which was founded by the dowager Queen Sexburga in about 664-675. Previous excavations located close to Minster Abbey identified structural remains and a small assemblage of mostly medieval pottery (Bennett 1994, Pratt 1995). The base of a sunkenfeatured building and a scatter of other Anglo-Saxon features dating from the seventh to ninth centuries are known from Mill Hill (TQ 961 727 and TQ 960 727) (Pratt 1998a; 2002) and a possible eighthcentury feature is known from Harps A venue (TQ 9500 7245) (Pratt 1998b). Two sites flanking the Old Ferry Road, to the south of !wade (TQ 8990 6760 and TQ 9020 6740) provided possible Mesolithic flint knapping debris, several Mid Bronze Age pits, a Late Bronze Age ditched field system with associated features, Late lron Age roundhouses and Roman period finds. Two phases of early medieval field ditches were also observed (Bagwell 2001 ). A Bronze Age and Early Iron Age settlement is known at Kemsley (TQ 9100 6600). This site included large collection of Devere! Rimbury pottery (Willson 2001, 32-33). The site was situated on the north facing slope of Rape Hill. The area of investigation began just short of the summit and ran approximately 150m down the slope to the north. It also followed the slope of the hill ap􀂎roximately 150m to the east, towards the point where the adjacent !h1stle Hill begins to rise up. The summit of Rape Hill is approximately +24.97m OD and in the area of development the base of the slope to the north was + 19.47 m and + 18.49m to the east. The site boundaries were defined by Barton Hill Drive to the west, a recently 267 MICK DIACK constructed road easement to the north and by field boundaries to the east and south. The Thames Estuary lies approximately 2.5km to the north, the Essex coastline being visible on a clear day. A wet ditch runs along the valley immediately to the north of the area investigated, suggesting that a small stream would once have run past the site, though this ditch was entirely dry at the time the site was excavated. The existence of a spring on the south-east side of Thistle Hill has been suggested (Pratt 1998a). The geology is heavy London Clay, as encountered over most of the Isle of Sheppey. Here the clay is slightly silty and contains rare, rounded small flints and very rare sub-rounded small to medium lumps of mudstone. The dating of virtually all features on site proved problematic due to the nature of the heavy London Clay, which has a tendency to crack as it dries out in hot weather producing considerable fissures, which allow later cultural material to intrude into earlier contexts. This intrusion is likely to be exaggerated by the movement of the clay as it dries out and becomes wet again with the changing seasons. Small fragments of modern material were found in most of the soil samples even though these were taken from freshly excavated areas and as deep within the fills as was possible. Much of the pottery recovered from the excavation consisted of very small sherds and thus prone to this sort of movement after deposition. It is also likely that some of the pottery recovered was residual. The agricultural use of the land undoubtedly included manuring and many pottery sherds were abraded, suggesting that they had come from elsewhere. The majority of the features on site were quite shallow, especially the features located towards the base of Rape Hill. It was considered likely that the features had been truncated by ploughing and would originally have been more extensive. This truncation has also made the interpretation of the archaeological features much more difficult. For ease of reference site north has been used as all the linear features described tend to run either site N-S or site E-W. PREHISTORIC OCCUPATION Phase 1 The earliest phase of activity comprised ditches, pits and post-holes found in both the (southern) car park and the northern area. The ditches appeared to form the ephemeral remains of two or more enclosures. Ditch 1 was L-shaped and was located at the southern end of site. The fills contained rare small, rounded flints, chalk fragments and 268 EXCAVATIONS AT BARTON HILL DRIVE, MJNSTER-IN-SHEPPEY charcoal. A sherd of flint-and-grog-tempered Late Bronze Age/Early Iron Age pottery together with oyster shell fragments and animal bone was also recovered. The longer part of the ditch measured 17 .00m and was aligned N-s, the shorter part was 11. 50m long. The ditch varied in size but typically was about 1.50m wide and 0.30m deep. There was a continuation of this ditch 0.65m to the east [Ditch 2]. This was approximately 7m long and from 0.60-1.00m wide. The fills of this continuation were similar to those of ditch l . A further section of ditch [Ditch 3] was possibly associated with ditch 1 and was aligned N-S, roughly parallel to the longer arm of the L-shaped part of ditch 1. This had two fills; the lower was 0.40m thick and contained very small amounts of oyster shell and a sherd of flinttempered pottery. The upper fill was 0.60m thick and contained a single sherd of flint-tempered pot and common oyster shell (many whole). This ditch extended beyond the southern limit of excavation. As seen, it was 2.50m long, 1.50m wide and 1.00m deep. The enclosure formed by these ditches is unclear, but it is possible that elements of it continue in the field to the south. In the northern part of the site there were two further ditches, both of which shared the same alignment and were probably related. Ditch 4 was E-W aligned, differing from ditches 1 and 3. The fill was dark grey, very compact, silty clay with occasional small to medium rounded and angular flint inclusions. The fill contained thirty-eight sherds of flint-tempered and four of flint-and-grog-tempered pottery. The ditch extended beyond the western limit of excavation but was at least 20.00m in length. The width ranged from 1. 70-2.25m and the depth from 0.25-0.45 m. Separated from ditch 4 by a 4.00m causeway was Ditch 5. This was 13.00m long, l . 10m wide and 0.20m deep. The fill was mid greyish brown compact slightly silty clay with occasional small rounded flint pebbles, differing from the fill of ditch 4. Located to the south and east of ditches 4 and 5 was another small (6.00m long, 1.80m wide and 0.25m deep) linear feature (Ditch 6) at right angles to the ditches to the north. Though tenuous, it is possible that these three elements formed an enclosure. The fill was light grey compact clay and contained a sherd of flint-tempered pottery. There were a number of features that appeared to be associated with these ditches. A small cluster of four pits and three post-holes were located immediately to the south of ditch 4. The pits [ 124, I 34, J 36 and 189] were on average l .40-2.50m long, l .10-l .30m wide and 0.20-0.40m deep, and contained thirty sherds of flint- and flint-andgrog- tempered pottery. The post-holes (126, 130 and 132} were on average 0.45-0.60m in diameter and 0.10-0.20m deep. No clear function is evident for these features, but they seem to be related to the putative enclosure formed by ditches 4, 5 and 6. Isolated, and in 269 MICKDIACK more or less the middle of the site, was a single pit [89]. This was 1.57 x l . I 7 m and 0.30m deep. It contained two fills and three sherds of flint-tempered pottery. South-east of this was a further, larger pit [76]. This measured 3.60 x 2.60m and was 0.43m deep. It contained a single fill that contained two sherds of flint-tempered pottery. An irregular shaped pit, a small pit or post-hole and a post-hole were located to the north of ditch 4. The pit [114) measured 3.50 x 1.80m and 0.30m deep. The fill contained nine sherds of flint- and one of flint- and grog-tempered pottery. The smaJI pit or post-hole [1 12] was 0.70m x 0.60m and 0.15m deep. It contained six sherds of flint and one of flint and grog-tempered pottery. The post-hole ( 191] was 0.40 x 0.25m and 0.25m deep. This contained a single sherd of flinttempered pottery. An unusual group of features was noted north of ditch 5 and located centrally in the northern area of site. These features were not clearly understood. There was a very irregular shaped pit [145) that measured c.1.80 x 1. 70m and 0.45m deep. It had a single fill that contained ten sherds of flint-tempered and eight sherds of flint-and-grog-tempered pottery. This had been cut along its N-S axis by two post-holes (not illustrated). The post-hole at the southern end [14 7] measured 1.00 x 0.45m and 0.50m deep and had its axis inclined to the north. It had two fills, one possibly representing a post-pipe. The post-hole contained 104 sherds of flint-tempered pottery. This post-hole had been cut by a small (0.35 x 0.20m and 0.12m deep) feature [ I 43] the fill of which contained a single sherd of flint-tempered pottery, though it is possible that this had been disturbed from the context below when the feature was cut. It is uncertain as to how this small feature fits within the stratigraphic sequence. The northern part of the irregular pit was cut by a further post-hole (168, not illustrated]; this was 0.80 x 0.40m and 0.24m deep and had a vertical axis. It contained two sherds of flint-tempered and two of flint-and-grog-tempered pottery. A further post-hole [1 74, illustrated), on the same N-S alignment as 147 and 168 was located 0.20m further north. This was 0.60m in diameter and 0.22m deep and it did not contain any pottery. It is difficult to interpret pit 145, but the post-holes were clearly related to it. The pottery from it was also contemporary with the post-holes. The features appeared to have formed a pit with an angled timber structure over it, which had been backfilled with the timbers remaining in situ. The features described above have all been assigned to the Late Bronze Age/Early Iron Age (c.850 to 600 BC) on the basis of pottery contained within their fills and would seem to represent a field system with associated small pits suggesting a farmstead. As the outline of such a farmstead is unclear, it is suggested that much of the archaeological evidence has been removed b y ploughing, or else the focus 270 EXCAVATIONS AT BARTON HJLL ORNE, MINSTER-IN-SHEPPEY of settlement lies elsewhere and the features observed here are peripheral to the settlement. Phase 2 A second phase of prehistoric activity on site was postulated from the presence of a further ditch [Ditch 7] which did not fit with the alignment of the ditches described above. Five other features have been tentatively ascribed to this second phase of activity due to similarities in their structure and the fact that four of them seem to share the alignment of ditch 7 rather then the other ditches. One of the five features has an uncertain relationship with ditch 5, but may be cutting it. All these features were located to the northern part of the site. All contained Mid Bronze Age to Late Iron Age pottery and it was not possible to separate them from the features ascribed to Phase 1 on that basis. Ditch 7 had an E-W alignment and was about 28m long and a maximum 2.50m wide and 0.45m deep. The ditch fill produced a single sherd of mid to late Anglo-Saxon Ipswich type pottery dated to AD 725-850. The presence of this pottery is problematic. Given the problems with the cracking of the London Clay and the ploughing of the site and the fact that the sherd, though fairly large (55g), was abraded it was probably intrusive. The furthest south of the five features was a subrectangular cut [98] that was 2.97m long and 1.93m wide, oriented site E-W. The break of slope was steep at the western and southern edges, shallower to the north and very shallow to the east. The cut was a maximum of 0.26m deep, with a fairly flat base. It appeared that the shallow portion to the east was forming a step to allow easy access in and out of the feature. The cut contained two fills, the first filling the lower area, below the height of the 'step'. This was light yellowish brown, firm/plastic, silty clay with rare small, irregular flints and rare charcoal fragments. This fill contained nineteen sherds of flint-tempered Late Bronze Age/Early Iron Age pottery. The second fill was mid greyish/ yellowish brown, firm but slightly friable silty clay which contained rare, small irregular flints, rare small chalk fragments and occasional charcoal fragments. This fill also contained Late Bronze Age/Early Iron Age flint-tempered pottery, in rather larger quantities (l 62 sherds) as well as two sherds of Belgic grog-and-flint-tempered pottery. Approximately 20.0m to the north of this were two further similar features that were parallel to each other, about 0.50m apart, and with a site N-s alignment. The farthest east of these was 166. This was subcircular and 3.60 x 1.70m and a maximum of 0.42m deep. The break of slope was steep at the top, more gradual towards the base on the northern, western and southern sides. The break of slope was 271 MICKDIACK gradual on the eastern side and, as with 98, this was considered to form a step. This cut contained three fills, two of which contained charcoal fragments. The latest of these fills contained seven sherds of Late Bronze Age/Early Iron Age flint-tempered pottery and one sherd of early medieval (I 150-1225) shell-tempered pottery. The second of these parallel features, located west of 166, was an irregular shaped cut [106), which was 2.80m long and 1.80m wide and 0.40m deep. This cut had a steep break of slope which then became shallow to form a flat ledge, typically 0.20m wide, before sloping down more gradually towards a slightly concave base. There were three post-holes related to this feature [118, 138 and 140 not illustrated]. Post-hole 118 was at the northern end of 166, 138 and 140 were in the south-eastern corner. The main cut for this feature had four fills. The earliest was 117, a light greyish/yellowish brown, plastic/soft, slightly silty clay with no inclusions and no finds. Fill 105 was mid to dark greyish brown, firm but slightly friable silty clay with rare small, irregular flints, burnt flint and charcoal fragments. It also contained eleven sherds of flint-tempered and one sherd of flintand- grog-tempered pottery. Fill 103 was light to mid reddish/ yellowish brown, firm but slightly plastic slightly silty clay that appeared to be burnt. It contained rare, small to medium rounded flints and rare charcoal fragments as well as four sherds of flinttempered, one sberd (2g) of flint-and-grog-tempered pottery. Fill 116 was probably a much later occurrence than the other fills, representing material that accumulated after the other fills had slumped into the cut. It was a light greyish/yellowish brown, firm, slightly silty clay with common small rounded flint inclusions. Again, this fill contained prehistoric pottery, consisting of six sherds of flinttempered, and one flint-and-grog-tempered sherd. Roughly 12.Sm to the west of 106 was cut 122. It was of a similar shape and size and orientation to the other four features but lacked any internal structure or post-holes. It had an uncertain relationship with the terminus of ditch 5. The cut was an irregular oval 3.40m long and 1.80m wide and was only 0.25m deep, suggesting truncation by ploughing (which may explain the lack of internal structure). The cut contained a single fill, which was mid greyish brown, compact, slightly silty clay with occasional small, rounded flints. There were no finds from this feature. The final of these five features [156] was located towards the western end of the site, close to the haul road area. It contained four post-holes [158, 160, 177 and 179] and a shallow feature [181] cut into its base (these are not illustrated). The main cut was aligned N-S and was 3.50m long. The shape in plan was roughly rectangular, but with the south-eastern corner foreshortened to create a narrower end 272 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY at the south. It was 2.70m wide at the northern end and 1.70m wide at the southern end. Cut into the base of the southern end was an ovalshaped feature [181]. The fill of this was mid to dark grey compact silty clay with rare small to medium rounded and angular flint and this contained eleven sherds of flint-tempered and five sherds of flint-and-grog-tempered pottery. The cut was 1.40m on its long access (E-W) and 0.7m wide. It had a fairly flat base that was 0.16m below the base of the main cut. This oval feature has been interpreted as a possible hearth, though there was no evidence for any burning. Also cutting the base of 156 were four post-holes. The largest of these was 0.45 x 0.30m; the smallest was 0.25m in diameter. They ranged from 0.15-0. l 8m in depth. Eight sherds of flint- and two sherds of flint-and-grog-tempered pottery came from these post-holes. The backfill of the main cut of the feature was a dark to mid grey, compact clay with rare small to large irregular flints and contained four sherds of flint-tempered and one sherd of flint-and-grog-tempered pottery. The dating of this phase is problematic. At the time of excavation, it was thought that the five features in this phase were Anglo-Saxon sunken-featured buildings. Subsequent analysis of the pottery found it to be Mid Bronze Age to Late lron Age. The only Anglo-Saxon pottery came from the ditch 7 and this was a single sherd. This means that a Mid Bronze Age/Early Iron Age date is more likely (although residual pottery cannot be completely discounted). They have been phased separately from the other Mid Bronze Age/Early Iron Age features due to the alignment of ditch 7 which matches that of these features as well as to the uncertain relationship between cut 122 and ditch 7 (122 may have cut this ditch). Though it seems likely that these features were later than the Phase I features, it has not proved possible to say by how much, perhaps there was only a short time span involved and this was why it did not prove possible to establish the chronology by excavation. The contents of the soil samples (animal and fish bone, marine mollusc shells, burnt flint fragments, charred grain, etc.) are suggestive of occupation and this together with the internal structures and post-holes suggests that they represent small huts. This would be consistent with the sort of small farmstead suggested above. MEDIEVAL AN D POST MEDIEVAL OCCUPATION Phase 3 This phase comprised a ditch, a rectangular feature and a pit, 􀁰II 􀁱ocated at the southern end of the site and extending beyond the hm1t of excavation. 273 MICKDIACK Ditch 8 had been re-cut, this later cut removing much of the profile of the earlier ditch. Three slots were excavated through this ditch. The original ditch was steep-sided and a maximum of 1.1 Orn deep. The full width is not known as the re-cut removed the upper edges of the original ditch, so it was evidently narrower. One ditch fill was noted; this was a light yellowish brown, firm, slightly silty clay containing rare, small chalk fragments. There were no finds. The re-cut of ditch 8 had a maximum width of 1.30m and was 1.10m deep. It varied in profile but could broadly be described as having steep sides, becoming near vertical towards the base, the base being flat. The ditch had three fills; the primary fil! contained a sherd of organictempered pottery dated to c.575- 750 and a sherd of Canterbury-type sandy ware dated c.1050-1225. The intermediate fill contained three sherds of shell-tempered pottery dated to c.1050-1225. The upper fill contained a sherd of Mid-Late Iron Age pottery, a sherd of Upchurch fineware (late first to third century), a fragment of Roman brick and thirteen sherds of shell-tempered pottery c.1050-1225. This re-cut ditch was in turn cut by a rectilinear feature [19]. This consisted of a cut 2.90m long and 1.50m wide, orientated E-W. It was fairly straight-sided, with a flat base and was 0.20m deep. It had a single fill which contained chalk and charcoal fragments as well as animal bone, oyster shell, one fragment of post-medieval roof tile (presumed residual), one fragment of slate and eleven sherds of early medieval shelly ware of c. l 050-1225 date. The final feature ascribed to this phase of activity was a smalJ subcircular pit [11]. This was 1.18 x 0.76m and 0.53m deep. The fill contained many fragments of non-local stone packed mainly into the upper part of the fill. The stone included quern, rubbing and hammer stones. This context also contained twenty-three sherds of early medieval shelly ware. The features in this phase are all considered to belong to the early medieval period (c. 1050-1225). A small amount of Roman material was recovered from the re-cut of ditch 8. This consisted of a fragment of Roman brick and a sherd of Upchurch ware. Due to the greater quantity of medieval pottery from all three fills of the re-cut ditch, it is considered that the Roman material is residual. As these features were located at the southern limit of excavation, it seems likely that medieval activity would have extended to the south. The ditch was fairly deep, but presumably it was part of a medieval field system. The other two features were presumably part of this system, but had been used for the disposal of domestic rubbish. Phase 4 Post medieval activity comprised of scatter of nine features, eight of which were in the car park area and one in the northern area. The features contained early nineteenth- to twentieth-century material. 274 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY There were five pits [13, 15, 28, 54 and 95] and three post-holes [26, 30 and 91]. The features in this phase do not seem to form any coherent pattern but presumably relate to agricultural activity. Nine pits [3, 5, 7, 9 (access road area, not illustrated) 42, 52, 82, 154 and 194] and four post-holes (80, 93, 149 and 152) remained. THE PREHISTORIC POTTERY by Peter Couldrey A total of 488 sherds, weighing 1.5kg, was recovered from stratified contexts, with the exception of two vessels, one of which is Early Bronze Age (c.2000-1500 BC) and the other Middle-Late Iron Age (c.350-1 BC), all of these are likely to date to the Late Bronze Age/Early Iron Age (c.850-600 BC). No complete profiles were recovered and most of the pottery was very fragmentary and worn, with 78 per cent of the vessels being represented by sherds weighing less than l0g. The few recognisable sherds provide the evidence for a broad indication of chronology. The following fabric groups were identified: Fabric I: flint-tempered. Most of the sherds were flint-tempered in a glauconitic sandy matrix, which also contained quartz, iron oxide and organic inclusions in varying proportions and rare fragments of calcareous sandstone. The flint inclusions vary in size and density from being sparse to moderate at less than 0.5mm to being common up to 3mm and moderate to sparse up to 5mm. Fabric 2: flint-and-grog-tempered. This fabric contained sparse to moderate flint tempering with generally moderate to common grog inclusions up to 2mm in size, in a sandy matrix. characterised by common inclusions of black glauconitic grains. Fabric 3: fine glauconitic sand-and-grog-tempered. One sherd ( l I 9/1) had grog inclusions up to 2mm in size in a fine glauconitic sandy matrix. It is distinguished here from the previous group because it contained no flint. However, it weighs just lg and its small size suggests that this distinction could be more apparent than real. Fabric 4: quartz in a glauconitic sandy matrix tempered. One sherd (33/1) had moderate quartz inclusions up to 0.5mm in size, in a glauconitic sandy matrix. A table showing the distribution of fabrics throughout the contexts is held in the archive. Except where stated otherwise, all the vessels referred to below are flint tempered, in Fabric l . The earliest recognisable vessel is a beaker (Fig. 2/1) in flint-and-grogtempered fabric (Fabric 2), from pit S 19, phase l . Two sherds of the same vessel were also recovered from the fill of post-hole 168, phase l. This cut pit S19 and the sherds are probably derived from this context. The precise form of the complete vessel is uncertain, and the relative position of the rim and shoulder need not be as close as shown in the drawing. The fine everted rim is similar to those associated with Clarke's Shapes I, II and III (Clarke I 970. 275 MlCKDIACK i- ' 2 \-􀀂 ,, 􀀌,.· 5 I' 􀀦 ,,, £1\ -1:9 -uA 8 ' I 9 ,_. •-􀀃 /II . 6 '\_ 10 Prehistoric Pottery from Barton Hill Drive, Minster I',:· .r. ., .. ',.- , . (', ' ·.- ·.7''\4 11 ' ' ,- " 14 ,=15 16 1μ:o/\:t: :·:.-::}·:.􀀧.:\:.:·:􀀨-TC􀀩:-?􀀪t 􀀅􀀆s7:􀀇 ✓ I I \ ,- " 19 17 Post Roman Pottery from Barton Hill Drive, Minster 7 Fig. 2 Pottery from Barton Hill Drive, Minster-in-Sheppey: Scale I :4. 276 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY App. 1.2). The sherds are worn but were evidently decorated with twisted cord impressions. These are set in horizontal rows and separated by a zone of panels of vertical grooves apparently alternating with plain areas. This overall design conforms to Clarke's Basic European Motif Group I, numbers 1 and 9 (Clarke 1970, 424-5). This vessel from Barton Hill Drive is likely to fall within the first half of the second millennium BC. The surviving fragments of this vessel weighed just 30g and the sherds are extremely worn. It is possible that it was residual, redeposited from elsewhere on the site. The remaining ten sherds from the pit were all flint-tempered (Fabric 1 ), plain and indistinguishable from most of the other pottery on the site. Nor is there any certain evidence for other vessels dating to this period. The other sherds in fabric 2 are always small, almost al ways in the minority within any context, and exhibit no datable traits. While they could be contemporary with the beaker, there is equally no reason why they should not be contemporary with the bulk of the flint-tempered sherds which are attributable to the Late Bronze Age or Early Iron Age. The fill of feature 145 was cut by a post-hole 14 7 (both phase I), which contained the remains of two vessels with worn surfaces but both with walls between 3-5mm thick. One was extremely fragmentary and is not illustrated; the other is shown as Fig. 2/2. Such thin walls are often associated with fineware bowls of the eighth or seventh centuries BC, as at Monkton Court Farm (Macpherson-Grant I 994, 259 and fig. 20, M24 and M26). Of the remaining features, the largest quantity of pottery came from pit 98 (phase 2). This produced one rim and two shoulders, which are illustrated in Fig. 2/3, 2/4 and 2/5 and fragments of two flat bases. The slack shoulder with finger-tip impressions (2/3), the shouldered vessel (2/4) and the rim decorated with fine diagonal grooves (2/5) can be paralleled in decorated Post Deverel-Rimbury (PDR) assemblages of the ninth and eighth centuries BC (Barrett 1980). The light parallel grooves on a thin-walled vessel (Fig. 2/6, from feature 156, phase 2) can also be paralleled at Monkton Court Farm (MacphersonGrant 1994, 265, fig. 9). Internally bevelled rims are also known from Middle and Late Bronze Age contexts, but the fragment of such a rim (Fig. 2/7) also from 156, is likely to be contemporary with the PDR material. A common PDR form, the plain inturned rim (Fig. 2/8), was recovered from the upper fill of ditch 4. This form is found from the Middle Bronze Age to the fifth century BC. Similarly, the body sherd with a single groove, 3mm wide (Fig. 2/9) from the lowest fill of cut 181 (phase 2), could belong to the Late Bronze Age or even to the fifth to fourth centuries. Here they are both assumed to be contemporary with the bulk of the Late Bronze Age/Early Iron Age material. The fragmen t of a flat base (Fig. 2/IO) with plain smooth surfaces, in a fine glauconitic sandy fabric with quartz inclusions (fabric 4), is probably Middle-Late Iron Age. This fabric is commonly associated with footring jars of the Middle Iron Age. It is the only example of this fabric from the site. It is evidently residual in the recut of ditch 8. The pottery from the site is extremely fragmentary and dating the sherds on the basis of individual traits is problematical. Some could belong to anytime from the Mid Bronze Age to the Early Iron Age. The beaker can be placed 277 MICKDIACK broadly within the first half of the second millennium BC. With the exception of one sherd, all of the remaining identifiable vessels, which are all in fabric 1, are likely to fall between c.850-600 BC. It is also assumed that the bulk of the remaining body sherds, which share the same fabric, also fall within the same period. The exception is the flat base in fabric 4, from the recut of ditch 8, which is likely to be Mid-Late Iron Age in date. While the pottery can be taken as evidence of phases of activity on the site, using it to date the features is less certain. The small size of the sherds suggests that they could all be residual, incorporated in the fill of features well after the pots themselves had been discarded. THE POST ROMAN POTTERY by John Cotter A total of 163 sherds (] 018g) of post-Roman pottery was recovered. The bulk of the post-Roman pottery derives from rubbish pits and ditch fills situated at the southern end of the site. A much smaller scatter of pottery occurs across the entire site. Though very fragmentary, the early medieval pottery is, in the main, in a reasonably fresh condition, though some sherds are fairly worn. On the whole, the early medieval material (81 per cent of the post Roman assemblage) appears to represent contemporary early medieval rubbish dumping suggesting occupation on or near the site during this period. The much smaller amounts of Anglo-Saxon and medieval (thirteenth- and fourteenthcentury) pottery are generally very small and worn. These are demonstrably residual in their contexts or, in one or two cases, intrusive into prehistoric contexts. The latest pottery represents 'Victorian' rubbish dumping. Anglo-Saxon Pottery EMSJD Fine sandy ware c.450-700: a single small sherd in a reduced grey fabric, probably from a smallish thin-walled globular jar. The external surface is smooth, possibly burnished, and bears traces of decoration. This consists of at least three lightly incised parallel lines with a possible trace of another line at right angles to these. These could possibly represent part of a chevron design. EMS4 Organic-tempered ware c.575-750: two small worn sherds including a plain everted rim from a small jar (Fig. 2/11) with orange-brown surfaces and grey core, from part of the recut ditch 8 (phase 3). Locally this type could be more long-lived than its estimated date-range at Canterbury as it occurs on other (unpublished) sites in the Minster area in association with Ipswich ware, suggesting a currency perhaps as late as 850. MLS7B Ipswich pimply ware c.720-850: large worn body sherd (max. length 90mm). Pale grey with coarse rounded to sub-angular inclusions up to 3mm across, predominantly quartz, with some flint. Characteristic external girth grooves. From a large jar or spouted pitcher. From ditch 7 (phase 2). Ipswich ware is the commonest (English) imported type found on Mid-Saxon sites at Minster and Canterbury. 278 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY MLSJ0O Unidentified, possibly Mid-Saxon ware c. 650-850: a very worn base sherd (uncertain flat or sagging) in coarse reduced sandy ware, possibly Mid-Late Saxon Canterbury sandy ware (MLS2) or possibly 'Belgic'/Roman (possibly wheel-turned). Occurs only with prehistoric and 'Belgic' pottery in ditch 4 (phase 1 ). Early Medieval Pottery EMJ Early medieval Canterbury sandy ware c. !050-1225: four body sherds from a minimum of three vessels; those from ditch 8 (phase 3) and ditch 5 (phase 2) are possibly from the same vessel. The usual form, from which these sherds almost certainly derive, is the large sagging-based cooking pot/jar. Made in the Canterbury area, including (after c.1140) Tyler Hill, the commonest early medieval ware on east Kent sites. These examples probably date as the majority shelly ware (see EM2), although they are not in themselves very diagnostic. EM4 North or West Kent fine sandy ware c.1125-1250: three small joining body sherds from a single vessel in a soft fine brown sandy ware with a well-defined light grey core containing sparse burnt-out organic inclusions. Traces of decoration externally comprising a single incised wavy line were noted. In Kent, this sort of decoration is usually found on the shoulder of jar forms and on spouted pitchers in particular. A source in the Medway estuary or further west seems probable. Small amounts of this ware, probably distributed coastally, occur at Canterbury, Chestfield (near Herne Bay) and particularly Dover. The sherd here was associated with two sherds of EM2 shelly ware ditch 2 (phase 1) and is most probably of similar date (see below). EM2 Early medieval shelly ware c.1050-1225: the predominant post-Roman pottery type from the site (120 sherds). On the basis of rim sherds alone a minimum of eleven vessels is represented (10 cooking pots/jars and I jug). All these vessels appear to be hand-made but in most cases, the rims may have been finished-off on a turntable. This type of shelly ware (with little or no sand) appears to have a longer currency in the Swale area (probably to c.1250) than it has further east at Canterbury where all types of shelly ware appear to die-out c.1225, probably because of competition from the local sandy ware industry at Tyler Hill about a mile north of the city. On the basis of parallels (see below), the range of rim and vessel forms present from Barton Hill Drive, however, indicate a currency within the period c.1150- 1225. The only other post-Roman pottery associated in context with EMZ here are very small amounts of other mainly twelfth-century wares (EM!, EM3 and EM4). At nearby Iwade (see below) developed EM2 forms occur in association with Tyler Hill ware and other glazed jugs suggesting a date of c.1225/50-1300 for the main phase of activity there. At Barton Hill Drive, perhaps significantly, such associations do not exist and the rim forms present seem less developed and therefore perhaps earlier than those at Iwade. They are, furthermore, better paralleled among the large assemblage of shelly wares from Townwall Street, Dover in contexts dated c.1150-1225 (Cotter 279 MICKDIACK forthcoming a), but it quite possible that the Barton Hill Drive material belongs entirely to the second half of the twelfth century. The EM2 shelly ware assemblage, though relatively small, offers a useful comparison to other shelly ware assemblages from Kent and is therefore worth describing in some detail. Shell-tempered pottery was a very widespread tradition in early medieval Kent and shelly wares were doubtless produced at many locations throughout the county. Predominantly, it would seem, along the coast, using contemporary crushed marine shell, but also at inland locations, such as Ashford, using shell derived from fossil sources (Cotter 2002). By studying assemblages such as this, we can come to a better understanding of the local variations within this ubiquitous and superficially homogeneous tradition. Fabric description: the code EM2 refers to shelly wares containing little or no sand as opposed to those with moderate to abundant sand (see EM3). The predominant firing trend here is for weakly oxidised orange-brown, fawn, or chocolate-brown surfaces and margins, and light grey cores. Some examples, however, are reduced brownish-grey to dark grey, either on one or both surfaces. Several examples, in addition, are soot-blackened externally from their function as cooking vessels. Hardness ranges from soft to fairly hard; and surfaces feel from fairly rough to almost soapy. The matrix is smooth with abundant (microscopic) silt-sized quartz and abundant very fine to medium mica; in rare instances up to l mm across. There is considerable variation in the amount of visible quartz sand present which is rare, sparse or sparse-moderate, though always present to some degree. Quartz grains are generally 0.25-0. 75mm across, mostly rounded, occasionally sub-angular, mostly clear with occasional orange or brownish grains. Rare rounded grains of milky quartz or chert up to 1 mm have been noted. There is also sparse rounded to sub-rounded black and red flint up to 1 mm across and either red iron oxide or soft red-brown iron-rich clay pellets up to 3mm across. Sparse very coarse black organic inclusions occur in most examples. Moderate to abundant coarsely crushed marine shell is the predominant visible inclusion. This has generally survived well in the sense that only in a small number of sherds has it been completely or partially dissolved-out. Shell inclusions up to a maximum of 8mm have been noted but the majority are under 3-4mm. The shell consists of a mixture of fresh and worn platy fragments and lumps, white, grey or banded. Platy bivalve shell appears to be the main constituent. The predominant species recognisable appears to be mussel (Mytilus edulis), followed by cockle (Cerastoderma edule), tiny scallops (Chlamys sp.), and probably clam-like Baltic Tellin (Macoma balthica), occasional tiny gastropods (probably Hydrobia sp.), rare or occasional honey-structured shell (probably barnacle) and microscopic spiral-chambered forams. Of the minimum eleven vessels represented, one is a jug and the rest are large cooking pots/jars. The jug is represented only by a rim, which is 􀂙f collared form (Fig. 2/12, ditch 1, phase 1). There is a slight possibility that tt could be a spouted pitcher but a jug seems more likely. Shelly ware jugs are very rare at Canterbury and only slightly more common from other, primarily coastal, sites in Kent. The jug form, along with the collared rim, appears in Kent c.1140/50 but is probably not common until the end of the century. 280 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY Several shelly ware jugs are known from Dover (Cotter forthcoming a) but the percentage they form of the large shelly ware assemblage there is probably no greater than that seen at Barton Hill Drive. Nearer to home, sbelly ware jugs are known from north Kent at Teynham, Seasalter near W hitstable (Dunning 1956, fig. 6.A5) and from findspots near Canterbury at Sturry and Blean (Thornden Wood), as well as from Canterbury itself. Cooking pots/jars are overwhelmingly the predominant vessel form represented (minimum ten vessels). The rim diameter range is 210-360mm with five vessels forming a cluster in the 340-60mm range, indicating a preference for large vessels most probably used for communal cooking. There is a tendency for the smaller-diameter cooking pots to have simpler rim forms, generally beaded or thickened/flat-topped (Fig. 2/13 and 2/14), while the larger examples have more developed, externally projecting, sub-flanged or subsquared rims (Fig. 2/15, 2/ I 6, 2/ l 7 and 2/ 18), possibly to facilitate lifting or the tying-down of cloth covers. There are, however, exceptions which have large diameters but relatively simple rim forms (Fig. 2/ 19 and 2/20). Some examples exhibit external sooting and one example appears to have been perforated though the neck/shoulder area (Fig. 2/18) possibly indicating a specialised function. Basal fragments indicate typical medieval sagging bases. The apparent lack of bowls is noteworthy. Comparison with other shelly ware assemblages: in terms of fabric the closest parallels are with the shelly ware assemblages from Ferry Road, I wade (site code FR I 99; Cotter forthcoming b) and Teynham (Cotter 200 l a), both of which lie a few miles south of Minster on the opposite (mainland) side of the Swale. The EM2 fabric from these locations is more or less identical in character and almost certainly part of the same local industry or tradition using contemporary marine shell as the main tempering agent. This was added to a fairly pure, virtually sand-free clay (possibly estuarine?) con• taining abundant fine mica. Superficially, this is very similar to the EM2 fabric at Canterbury and Dover. The Canterbury type is the more similar and could share, at least in the earlier period (c. 1050-1150), a common origin with the Minster and !wade types somewhere in Swale/Sheppey area (Cotter 2001 b, 231-2). It differs, however, in that the shell content is generally fresher and predominantly cockle or scallop rather than mussel; gastropod, barnacle, echinoid spines and bryozoans are also generally commoner - particularly in later examples (c. I 150-1225, as with EM3). The later EM2 type at Canterbury, mainly because of the increased gastropod content, probably shares a common origin with the shelly-sandy ware EM3, which, it has been suggested, was in north-east Kent along the now-vanished banks of the Stour/ Wantsum estuary that formerly separated the Isle of Thanet from the mainland (ibid. Cotter 2002; and forthcoming a). The Dover types of EM2 and EM3 also contain abundant gastropod and probably have the same origin as the latter. These differences, however, are not easily recognisable without the aid of a microscope and the examination of a reasonable number of sherds. With regard to Canterbury, furthermore, it is important to bear in mind that the sand-free EM2 shelly ware fabric was largely replaced during the second half of the twelfth century by the shelly-sandy ware EM3 which survived until c.1225. Further west, in the Swale area, the predominance of EM2 as late as c.1250, resulted in the adoption of more developed late-looking forms 281 MICK DIACK than exist in BM2 at Canterbury. Among these the jug form and cooking pots and bowls with markedly flanged rim forms probably influenced by glazed Tyler Hill ware. Despite fabric similarities, however, there are significant differences in the EM2 shelly ware vessel types represented at both Minster and Iwade. At Minster, the EM2 assemblage comprised almost exclusively cooking pots and a single jug. In contrast, at Iwade the assemblage comprised almost entirely bowls and one cooking pot. This could, at first glance, suggest differences in either the type of food consumed at each site, or in the way it was prepared. Although there may be something in this idea, the explanation preferred here is a chronological one. The Minster assemblage is the earlier of the two, dating to c.l 150-1225, whereas the Iwade assemblage is later dating to c .1225/50-1300. At Minster, furthermore, EM2 shelly ware dominates the assemblage whereas at !wade it forms only 20 per cent of the assemblage, Tyler Hill ware being the dominant type (71 per cent). Most of the cooking pots and jugs there were therefore in Tyler Hill ware, with local shelly wares, now in serious decline, just about meeting the demand for large bowls. These were cooking bowls, a form that appears to have become increasingly popular in the late twelfth and thirteenth centuries. EM3 Early medieval shelly-sandy ware c.1075-1225: this comprises only five smallish sherds including a base sherd. These could represent three vessels, probably cooking pots/jars. There appears to be no significant difference on this site between the EM3 fabric and the majority EM2 fabric except that the former contains abundant medium-coarse quartz sand. One sherd, however, contains a very coarse rounded inclusion of fine-grained calcareous sandstone. Medieval Pottery MS London-type ware c.1140-1375: Fine orange-brown sandy ware produced in the London area, a small worn sherd from a jug base with traces of clear glaze, possibly later twelfth to thirteenth century. This is a fairly common fine ware type from sites in north Kent, and almost exclusively jugs. It came from a modern context (cut 13, phase 4). Ml Tyler Hill ware (Canterbury sandy ware) c.1225-1350. Two small worn scraps (one glazed), probably from jugs. These probably represent casual loss or manuring of the site rather than medieval occupation. Somehow, these were found intruded into otherwise prehistoric contexts (the fill of 145, phase 1 and of 181, phase 2), possibly through animal action or shrinkage cracks in the underlying clay. Despite the identification of possible Anglo-Saxon sunken-featured buildings, the pottery indicates only slight evidence of human activity on, or in the vicinity of, the site during this period. The small amount of Anglo-Saxon pottery present consists of types well known from other (mostly unpublished) Anglo-Saxon sites in the Minster area. It is clear, however, that the main period of post-Roman activity on the site, at least in terms of pottery use, 282 EXCAVATIONS AT BARTON HILL ORNE, MINSTER-IN-SHEPPEY occurred c.1150-1225. The pottery used consists almost entirely of locally produced shelly wares with a much smaller number of sandy ware vessels from the Canterbury area and at least one sandy ware vessel with a probable north-west Kent origin. This pottery appears to be entirely domestic in character and probably represents occupation on, or very near, the site. A single, residual, jug sherd in London-type ware may date from this period, or from slightly later in the thirteenth century. Thereafter there is no convincing ceramic evidence for occupation until the modern period. Two very small worn jug sherds in thirteenth- and fourteenthcentury Tyler Hill ware probably represent field-manuring rather than occupation. There is no pottery from the late medieval (c.1350-1550) or postmedieval periods (c.1550-1775), although there is a fragment of glazed peg tile of probable sixteenth- or seventeenth-century date as well as some (mainly late) post-medieval small-finds. A small number of later sherds, mostly datable c.1825-1900, apparently represent Victorian rubbish dumping. THE LITHIC ARTEFACTS by Beccy Scott Ten struck flint artefacts were recovered from Barton Hill Drive, Minster; the condition of these (most exhibiting some degree of patinated damage) suggests that they are not derived from a primary context. They are probably reworked from elsewhere, perhaps previously having been spread over the surrounding land surface. The technological and typological characteristics of the material suggest that it is later Bronze Age in date. The retouched forms are modified in an ad hoe manner in response to the blank form, rather than the imposition of a preconceived tool form. Mode of percussion is predominantly hard, with little platform preparation; incipient cones are common on several artefacts, reflecting poor control of flaking. This is typical of Later Bronze Age (and Iron Age) lithic reduction. The raw material used is poor and internally flawed, and seems largely derived from a fluvial context. Whilst the material itself has inherently imposed limitations upon the extent to which it could be reduced, in earlier prehistoric industries better quality material is often imported to overcome these difficulties. As such, the material is likely to be Later Bronze Age in date. Such material is widespread across the Kentish landscape and in this instance, seems to be reworked from primary context. As such, this material probably reflects the selective incorporation of the now deflated land surface into subsurface features. No real interpretation of activity on the site can be advanced on the basis of this assemblage. THE SMALL FINDS by Ian Riddler A small quantity of small finds were recovered from the site, most were unstratified or from late post-medieval contexts. In terms of the overall assemblage the only objects not of a late post-medieval date were two quern fragments, one from a small pit (11) and one from the fill of ditch 2. The 283 MICKDIACK quern fragment from pit 11 weighed 1,428g and was composed of Lower Greensand. The quern fragment from ditch 2 weighed 18g and was composed of basalt Java stone THE ANIMAL BONE by Robin Bendrey Archaeological excavation produced 107 fragments of animal bone, weighing 2956.5g, of which 46 per cent are identified by number and 91 per cent by weight. The following taxa were positively identified in the hand-recovered bone assemblage: cattle, Bos sp. domestic; sheep, Ovis sp. domestic; horse, Equus sp. domestic; and pig, Sus sp. domestic. The state of preservation of the animal bone assemblage is generally good. Root-etching on the surfaces of the bones is recorded from all contexts, and there is occasional weathering. Carnivore gnawing damage is reasonably common, being recorded on eight fragments from Phase I (9.8 per cent) and four fragments from Phase 2 (32.5 per cent). Phase 1 Eight sheep/goat bones are recovered from five contexts, and ten cattle bones from four contexts, six horse bones from four contexts and five pig bones from three contexts. Toothwear data available from one cattle mandible (M l = I, M2 = k, M3 = j (Grant 1982)) indicates that the animal was kept into old age, probably for secondary products such as milk, breeding or traction. A cattle deciduous upper second premolar (dP2) exhibits abnormal wear. The specimen has very heavy wear (in the form of a deep groove) across the posterior half of the tooth down to the base of the crown. The anterior edge of the tooth goes to a point, on which it is only slightly worn. The distribution of wear on the tooth suggests that the deciduous lower second premolar (dP2) was absent and that all the wear occurred against the deciduous lower third premolar (dP3). There are a number of reasons why this tooth might be absent (see Andrews and Noddle, 1975); however the slight wear on the anterior edge implies that the dP2 had been present but was lost prematurely. A horse metatarsal exhibits slight porous bone growth on most of the anterior surface of the shaft, suggestive of a mild case of osteoperiostitis (Baker and Brothwell 1980, 63-70). Phase 2 Nine sheep/goat bones are recovered from four contexts, and three cattle bones from three contexts and two horse bones from two contexts. Sheep bones are positively identified from this phase. A sheep/goat metatarsal has frequent heavy chopping around the distal diaphysis. This is excessive for normal butchery and probably represents a stage of bone working. A fr?gment of horse pelvis bears a number of cut marks across the pubis suggestive of dismemberment. Also, the ilium bears some (not all) characteristics suggestive of the bone having been fractured while fresh (Outram 2001), possibly for the marrow. 284 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY The small size of the assemblage severely limits the conclusions available. The fact that all the material is hand-recovered means that the smaller animals, sheep and pig, are likely to be under-represented in relation to the larger animals (Payne 1975), and the limited number of bone-producing contexts makes it extremely difficult to estimate the effect of context variability on the relative importance of the different taxa. The assemblage does, however, give an indication of the taxa exploited. Those present; sheep, cattle, horse and pig; are what would be expected as the most common animals from most sites of these periods, although it is probably not the full range of taxa originally exploited. The butchered horse pelvis from Phase 2 is an important record for the post-mortem use of horses; in this case indicating dismemberment and possible marrow extraction. Maltby (1994, 89), for Roman sites in Wessex, has shown that horses contributed more to the diet at rural sites than urban sites, possibly a continuation of Iron Age custom. A similar dichotomy has also been noted for some sites in Roman Kent (Bendrey, forthcoming); and this direct evidence from Barton Hill Drive builds up the picture of horse utili· zation on small rural sites. THE ENVIRONMENTAL RECORD by Enid Allison Eleven bulk samples were taken from Late Bronze/Early Iron Age and early medieval features on the site. Features sampled were pits, ditches, a possible sunken-featured building, and a post-hole. All samples had a high clay content and were therefore soaked in a I per cent solution of hydrogen peroxide before processing. Bucket flotation was carried out and a washover was collected on 0.5mm mesh. The residue from this process was sieved, using I mm mesh for the samples from the possible sunken-featured building, and 2mm mesh for the rest. Dried residues were examined for biological material and artefacts. The dried washovers were examined under a lowpower (x 10) binocular microscope. Pot sherds were common in several samples. The majority appeared to be prehistoric, with a few Roman and a modern sherd. It should be noted that in addition to the modern pot sherd, small fragments of brick or tile, coal and clinker were present in most of the samples indicating that small amounts of recent material had entered the archaeological deposits, probably via worm burrows or due to cracking of the soil in dry periods. The range of remains was similar in all samples and is typical of refuse from human occupation. Charred plant material, including cereal grains and chaff, was recovered from several samples. The chaff was generally well preserved, the grains generally less so. Chaff is often more useful for iden• tification of cereal species than the grains themselves. Bone was generally highly fragmented and present in small amounts. Domestic mammal and fish bone was recovered from several samples and bird bone from a single sample. Some of the mammal bone had been burnt. Shellfish remains were highly fragmented. Species identified were oyster, mussel, cockle, whelk and win 􀂲le. Fragments of a small gastropod and a bivalve were also present. All identified species are regarded as edible. 285 MICKDIACK Land snails were present in some samples but in too low numbers to be of use for environmental interpretation. Two small fragments of eggshell were recovered from one sample. The fragments have not been definitely identified, but one is very similar in both thickness and appearance to domestic fowl shell. The other fragment is considerably thinner with a much smoother texture, and appears to be of another species. GENERAL CONCLUSIONS Though there were problems with dating many of the features on the site, and much of the archaeology had been truncated by ploughing, the site nevertheless remains important. There has been little large scale excavation carried out on the Isle of Sheppey and this excavation has enhanced our knowledge of the area. Cultural material ranging from the Late Bronze Age/Early Iron Age transitional period through to late post modern, and all stages in between was recovered, demonstrating that, even if the dating of individual features was tenuous, the site has been in use throughout all that time. The excavation broadly confirmed the findings of the evaluation trenching carried out in 1998, which had suggested the presence of a Late Bronze Age/Early Iron Age farmstead with other, more ephemeral activity dating from the Late Iron Age/Roman, the Anglo-Saxon and the Norman periods. Four main areas of potential had been identified in the evaluation. The area with the maximum potential was only partly topsoil stripped, the subsoil remaining intact (this work was monitored) and thus the most promising part of the site (closest to the putative settlement) remained unexcavated. The features from phase 1, in particular the various ditches, clearly represent nearby occupation, even if the exact field system is unclear. Four hundred and eighty-eight sherds of prehistoric pottery represent a sizeable body of material, which, even if the dating of individual features was unclear, demonstrates activity on a reasonably large scale dating from this period. The location of the site, towards the summit of the gentle rise of Rape Hill would be ideal for a prehistoric farmstead, offering good views of the surrounding countryside, though actual habitation may well have been located further down the slope, possibly to the south-east, as has been suggested, as the wind would have been bitter during the winter months and Thistle Hill would provide some shelter from it (Pratt 1998a). The second phase of activity is perhaps the most contentious. The alignment of the ditches and their phasing is not clear and there is little dating evidence available from the finds. The dating material found in the possible sunken-featured buildings is all prehistoric but 286 EXCAVATIONS AT BARTON HILL DRIVE, MINSTER-IN-SHEPPEY this sort of feature is typically Anglo-Saxon in date. Feature 156 had four clear post-holes, 118 had three but the distribution of these was unusual, with one to the north and two in the south-eastern corner. Features 98 and 166 did not have evidence for post-holes but, as well as being a similar shape and size to the other features, had a 'ledge' or 'step' in them. These two features may have had a less substantial structure over them, or perhaps had been more heavily truncated by ploughing. Feature 122 was much less clear, having no internal structure to it at all, but it is on the same alignment as three out of the other four features and is roughly equidistant between those it shares an alignment with. Though Roman cultural material was scant, comprising two sherds of Upchurch fineware, two fragments of Roman brick and a small amount of 'Belgic' pottery, which is considered pre-conquest, but is difficult to date, it is sufficient to suggest that some activity took place nearby during this period. The medieval features are less difficult to date. Despite the presence of a sherd of Upchurch ware and fragments of Roman brick in the fills of ditch 8, the presence of early medieval pottery in all three of the fills of the re-cut makes a Roman date unlikely. The rectangular feature [ 19] was demonstrated to cut through the upper fill of the re-cut of ditch 8, but contained early medieval shelly ware of a c.1050-1225 date, making the two contemporary. It is considered that there are no firm conclusions that can be made about the stratigraphic relationship between the two, and thus they have been phased together. The fill of the small pit [cut 11) contained twenty-three sherds of early medieval shelly ware as well as the fragments of lava quern stone. The presence of elements of an early medieval field system and quern stone imply that a farmstead must have been somewhere nearby. As these features were located at the limit of excavation, it is likely that further features are present in the field to the south, and this is perhaps the focus of activity in the early medieval period. The small quantity of later cultural material is very much later in date (c.1825-1900) and is probably the result of manuring or dumping of rubbish. ACKNOWLEDGEMENTS The work was carried out on behalf of, and funded by, Wallis, a division of Kier Regional Limited. Thanks are extended to them and in particular to the site agent Mr Lewis Howard. The work was carried out to a specification compiled by Kent County Council. 287 MICKDIACK Thanks to Liz Dyson who prepared the specification and monitored the work. The writer would also like to thank Damien Boden, Steve Rawling, Peter Seary and Grant Shand for their very able assistance on site. Thanks to Louise Harrison for her finds work, Peter Couldrey for the prehistoric pottery analysis, John Cotter and Mark Davey for the post Roman pottery analysis, Enid Allison for her analysis of the environmental samples, Robin Bendrey for the animal bone and to Ian Riddler for the analysis of the small finds. Thanks also to Dominique Bacon for the pottery illustrations and to Will Foster for scanning and preparing them. BIBLIOGRAPHY Andrews, A.H. and Noddle, B.A., 1975, 'Absence of Premolar Teeth from Ruminant Mandibles found at Archaeological Sites', Journal of Archaeological Science, 2, 137-44. Bagwell, M., 2001, 'Phased summary and assessment document of an archaeological excavation at I wade Site A and Site B (west), I wade, Kent', Pre-Construct Archaeology Client Report. Baker, J. and Brothwell, D.R., 1980, Animal diseases in archaeology, London. Barrett, J. 1980, 'The Pottery of the Later Bronze Age in Southern England' Proceedings of the Prehistoric Society, 46, 297-319. Bendrey, R., 2001, 'Animal Bone from Glebe Land, Harrietsham', unpublished report for CAT. Bendrey, R., 2002, 'The animal bones from excavations at the Channel Tunnel Terminal at Cheriton, Folkestone', unpublished report for CAT. Bendrey R. forthcoming, 'The Animal Bone', in J. Rady, A. Hicks, I. Riddler and S. Pratt, Roads to the Past. The Monkton Landscape. Prehistoric, Roman, Anglo-Saxon and Medieval sites along the duelling of the A253, CAT Occasional Paper. Bennett, P., 1994, 'Work by the Canterbury Archaeological Trust', Arc/iaeologia Cantiana, CXIV, 463-464. Boessneck, J. 1969, 'Osteological Differences between Sheep (Ovis aries Linne) and Goat (Copra hircus Linne)'in D . Brothwell and E. Higgs (eds), Science in Archaeology, 331-58. Clarke, D.L., 1970, Beaker Pottery of Great Britain and Ireland. Cotter, J.P. 2001a, 'The Pottery', in M. Hicks and A. Hicks (eds), St Gregory's Priory, Northgate, Canterbury Excavations 1988-1991, The Archaeology of Canterbury New Series II, 231-66. Cotter, J.P. 2001b, 'Pottery Report' in P. Wilkinson, 'Further Investigations at Teynham Palace', Practical Archaeology 4, Kent Archaeological Field School, 16-17. Cotter, J.P. 2002, 'Medieval shelly wares in Kent: a summary of recent research', Canterbury's Archaeology 1999-2000, CAT, 56-60. Cotter, J.P. forthcoming a, 'The Pottery', in K. Parfitt, B. Corke and J. Cotter, Excavations at Townwall Street, Dover, 1995-6. The Archaeology of Medieval Dover I, The Archaeology of Canterbury New Series HI. 288 EXCAVATIONS AT BARTON HILL DRIVE, MJNSTER-Il-1-SHEPPEY Cotter, J.P. forthcoming b, 'The pottery from Iwade: The Street (IST-97) and Ferry Road (FRI-99)'. Unpublished CAT Client Report. Dobney, K. and Rielly, K., 1988, 'A method for recording archaeological animal bones: the use of diagnostic zones'. Circaea 5.1, 79-96. von den Driesch, A., 1976, A Guide to the Measurement of Animal Bones from Archaeological Sites. Dunning, G.C., I 956, 'Report on the pottery from Seas alter', in M. W. Thompson, 'A group of mounds on Seasalter Level, near Whitstable, and the medieval imbanking in this area', Archaeologia Cantiana, LXX, 59-63. Dyson, E., Shand, G. and Stephens, S., 2000, 'Causewayed Enclosures', in Current Archaeology 168, 471-472. Gardiner, M., 1999, Excavations at Minster in Sheppey, Kent, 1999, by Queen's University Belfast, unpublished typescript. Grant, A., 1982, 'The use of tooth wear as a guide to the age of domestic ungulates', in B. Wilson, C. Grigson and S. Payne (eds), Ageing and Sexing Animal Bones from Archaeological Sites. B.A.R. British Series 109, 91-108. Halstead, P. 1985, 'A study of mandibular teeth from Romano-British contexts at Maxey'. in F. Pryor, C. French, D. Crowther, D. Gurney, G. Simpson and M. Taylor (eds), Archaeology and Environment in the lower Welland Valley, East Anglian Archaeology Report 27, 219-24. Institute of Geological Sciences 1977: I :50,000 Geological Survey, Sheet 272 (Drift). Leach, P., 1969, 'Sheppey Romano-British Burials', Kent Archaeological Review 18, 7. Macpherson-Grant, N., 1994, 'The Pottery', in D.R.J. Perkins, N. Macpherson- Grant and E. Healey, 'Monkton Court Farm Evaluation, 1992'. Archaeologia Cantiana, CXIV, 248-288. Maltby, M., 1994, 'The meat supply in Roman Dorchester and Winchester', in, Hall, A. and Ken ward, H (eds), Urban-Rural Connexions: Perspectives from Environmental Archaeology, Oxford. Oxbow Monograph 47, 85-102. Outram, A.K., 2001, • A New Approach to Identifying Bone Marrow and Grease Exploitation: Why the "Indeterminate" Fragments should not be Ignored', Journal of Archaeological Science, 28, 401-10. Payne, S., 1975, 'Partial recovery and sample bias', in Clason A.T. (ed.) Archaeozoological studies, 7-17. Pine, J ., 1999, 'Shrubsoles Hill, Brambledown, Isle of Sheppey, Kent: an Archaeological Evaluation for Hanson Waste Management', Thames Valley Archaeological Services Client Report. Pine, J., 2000, 'Cell 3, Shrubsoles Hill, Brambledown, Isle of Sheppey, Kent: interim Report', Thames Valley Archaeological Services Client Report. Pine, J. 2001, 'Shrubsoles Hill, Brambledown, Isle of Sheppey, Kent: a post-excavation assessment for Hanson Waste Management', Thames Valley Archaeological Services Client Report. Pratt, S., 1995, 'Minster-in-Sheppey', Canterbury's Archaeology 1993-94, 26-28. Pratt, S., 1998a, Barton Hill Drive, Minster-in-Sheppey: Archaeological Evaluation (unpublished CAT Client Report). 289 MICKDIACK Pratt, S., 1998b, Harps A venue, Minster-in-Sheppey: Archaeological Watching Brief (unpublished CAT Client Report). Pratt, S., 2002, • A249, !wade to Queensborough Improvement Scheme, Kent: Archaeological Survey, Stage 3a: Updated evaluation and preliminary fieldwork proposals. (Unpublished CAT Client Report). Prummel, W. and Frisch, H-J ., 1986, • A guide for the distinction of species, sex and body side in bones of sheep and goat', Journal of Archaeological Science 13, 567-77. Willson, J., 200 l , 'Kemsley Fields, Kemsley, near Sittingbourne', in Canterbury's Archaeology 1998-1999, 32-33. 290