Excavation at Syndale Park, Ospringe

EXCAVATION AT SYNDALE PARK, OSPRINGE LUCY SIBUN with contributions by Malcolm Lyne, David Rudling, Luke Barber and John Shepherd Archaeology South-East were commissioned to undertake archaeological investigations, comprising an evaluation, watching brief and small-scale excavation, on the route of a new gas pipeline from Sittingbourne to Faversham. Little of archaeological interest was observed during most of the watching brief and evaluation (details of four of the trenches TTl to TT4 are housed with the archive). 1 However, two evaluation trenches, A and B, located in Syndale Park to the south of the A2 (Fig. 1) contained concentrations of Romano-British features and were subjected to small-scale excavation in August 1994. The current article represents a summary of these excavations: the full Fig. l Site and Trench Location Plan. 171 LUCY SIBUN report, including all specialist analysis of the excavated artefacts and ecofacts, is housed with the archive which has been deposited at the Fleur de Lis centre, Faversham. 2 Previous discoveries in the vicinity of Syndale Park indicate a substantial Romano-British settlement in the area. These include Claudian cinerary urns, Roman deposits containing coins and pottery, first to second century cremations and a substantial Romano-British cemetery, containing 387 burials (both inhumations and cremations). To the north of the A2 are the remains of Stone church. The church, which was in use until the sixteenth century, originated as a mausoleum in the Roman period. Excavations in the same field uncovered a Romano-British occupation layer, small pits, hearths and domestic rubbish. To the south of the road stratified Romano-British occupation layers were uncovered, producing second to third century pottery. 3 EXCAVATION RESULTS - TRENCH A This was 76m long and 2m wide and contained numerous features such as rubbish pits, post-holes and ditches (Figs 2 and 3). Full details are housed with the archive but the main features are discussed below. First and Second centuries AD. The first phase of occupation is represented by rubbish pits and ditches. Pits 10, 36, 52, 58, and 92 all date to the first century. Contexts 22, 32 and 90 date from the first to second centuries. The fill of pit 32 (33), produced over 4.5kg of pottery strongly suggesting that rubbish disposal was its primary function. Context 88 was an east-west ditch (Fig. 4: SS). The lower fills of this ditch, Contexts 89 and 77, were apparently deliberate rubbish deposits. Context 72 was a second-century layer of cobbled flint laid down in cut 98 (Fig. 4: S3). Above this was a thin second to fourth-century deposit (Context 71 ). Three post-holes, Contexts 12 (Fig. 4: SI) 65 and 84 date to this phase as does occupation layer 67. Second - Third centuries. The only context in Trench A that dates to this phase is Context 76, a fill of Ditch 88 (Fig. 4: SS). Third and Fourth centuries. Two pits date to the third to fourth centuries, Contexts 2 and 73 (Fig. 2 and Fig. 4: SS respectively). The remainder dating to the fourth century. Contexts 14 and 24 are rubbish pits and both contained large amounts of pottery and shell. 172 EXCAVATION AT SYNDALEPARK, OSPRINGE Trench A r·-·- -·-·---·-----·-· ,.-·􀀊---•-a i - -N- (11/ 2 - 􀀩 􀀪\ s\ .. • 1, 3 .... ,9􀀔 L __________________ \ 􀀫 ·-----·--·-----'·--·--·-b Fig. 2 Plan of Trench A (south). Trench A -N- 9-- . --· --, __ · '-----r· --. 􀀷,· -- . -- . I ,,, :,,' / I I 60 1, I I I // : 110 54 􀀄, = " :!- I -..\.- 1 ' I -• - _, ----·--· ·--'-h 1,􀀥-°;7'""􀀦􀀧·--·--·--·--·, :· .... 􀀨0 6:· \.• •-: .. 'a 􀀸.. -,• •88 • ===;;:---;:::::::::::;:::􀀹 􀀺-Flint 70 0 2m 90 Fig. 3 Plan of Trench A (north). 173 SI 􀀁 s N LUCYSIBUN S2 S3 SS N 0 1m S6 43 18 .. 46 • 45 s lfil Flint li!.il Chall< Wi) Shell 0 Im Fig. 4 Selected Section Drawings (Trench A). s Ditch 18 (Fig. 4: S6) contained eight distinct fills (Contexts 19, 20, 21, 40, 41, 42, 43, 44 ), but of these, only Context 20 produced a large quantity of artefacts, which dated to the fourth century. It would seem that the ditch had been filled, and gone out of use by the fourth century. The ditch was cut by Context 45, a fourth-century rubbish pit. A third linear feature, Context 38, also dates to the fourth century (Fig. 4: S2). Context 16 appeared to be part of a curving ditch but this could not be ascertained from the small section exposed. Layer 70 was a layer of flint cobbling, above Context 71 in Cut 98. This layer dates to the fourth century and could represent the resurfacing of a cobbled road, in use since the earliest phase of occupation (Context 72). Layer 55 was situated just to the south of Cut 98 and also resembled cobbling, again dating to the fourth century. Undatable features include Post-holes 8, 28, 56, 109, Pits 26, 30, 50 (Fig. 4: S4), 79 (not illustrated), gully 68 and feature 110. Contexts 4 and 6 were identified as field drains. EXCAVATION RESULTS - TRENCH B Trench B, measuring 62m in length was adjacent to the A2 on its south side (Fig. 1 ). All features dated to the first to third centuries. (See Fig. 5) 174 EXCAVATION AT SYNDALB PARK, OSPRINGB r---,,;1,,.-=-·-;􀀖· -·-·, 􀀗 􀀘-􀀙---·-·-·-·-•-•8 I 120"'1»f􀀳 1 i -,·• "4< . •• I. ,::: 􀀅 t '116 • 􀀴 ;rr,--, ----.'f>¾._ 10s ·-IJ,JJ.J '--kl ', N . 􀀬 107 • 122 ____, I L·-·-·-·-·-·-·-·-·-·-·-·-·-·- ·-·b •-r-•-·1·􀀞5-·-·-·-·-·T·11-y-· ,'T􀀅c f= ==}-123 ,01 ··---W .. J l,,\Y i : --------rt-,-,. 99,,;;;-􀀌96 94 b---·-· I ·-·-·-·-·-·L·-· l.......:.􀀠.L,.i., __ d r .---·-·-·-r-·,-r,-·-·-·-·-·-·-·-•-·e t I I I I : : : 113􀀵 0-112 I # I I I I I I I I I I -·-·-·-·-·'􀀚-!.t.. __________________ , 8---·-· -·-· ··-·-·-·-·-·-·-· -·-·-r 4metr8s to Trench B O 2rn end o1 trench ,,-- r>r-100 '-·-·-·-·-·-·-·-·-Lli._. ______ _ Fig. 5 Plan of Trench B. First and Second centuries. Pits 101, 105, 107 and 114 date to this period. There were several shallow linear features within the trench, (Contexts 112,113,122). Their precise function remains unclear and most were insubstantial. Context 112 and 113 did not produce artefacts but are likely to be of this early period. Context 122 is dated to the second century. Second and Third centuries. This phase is represented by rubbish pits (Contexts 96, 103, 120) and an oven or hearth (Context 118). The heat from 118 had burnt the surrounding natural Brickearth, hardening it and turning it a dull red-orange colour. The pottery recovered from the fill was of second and third century date. A shallow ditch of this period was approximately 0.33m deep and 2.50m wide (Context 123). Pits 94, 99 and 116 did not contain any dating material. A single post-hole was uncovered in the trench (Context 125). The post-hole did not contain any dating material. The features on the site are representative of domestic occupation. The ditches seem to have been in use during most phases of occupation. Jn the fourth century they seem to have been finally filled, the ground then reused for rubbish disposal. The small area exposed makes it difficult to interpret the ditches as the stretches available for study were too 175 LUCY SlBUN short. They may form part of enclosures, drainage or boundary features. It is possible that the shallower ditch, Context 123 (Trench B) was for drainage, but the more substantial ditches, Contexts 18, 38, and 88 (Trench A) could have been boundary features. The only other feature in the trenches which seems to have been in constant use is the possible cobbled flint road, Contexts 98, 71, 72, and 73. This road or metalled track was re-surfaced at least once during its life span. THE FINDS POTTERY by Malcolm Lyne The pottery found during the recent excavations spans virtually all of the period of Roman occupation, although late second- and early-third century assemblages are few in number and rather small in size. However, the current assemblage is important in providing the opportunity to compare with those from the nearby cemeteries. The following report emphasises the main points/attributes of the assemblage: a full report, including detailed fabric descriptions is housed with the archive. Fabrics - the coarsewares C. I A Early hand-made coarse, grog-tempered ware. The early East Kent grog-tempered ware industry spans the years from preConquest to late first-century and continued to be associated with large furrowed and combed storage jars well into the second century. C. I B Late, hand-made, grog-tempered ware with ferrous inclusions, usually fired black. A production centre may have been operating at Richborough during the late fourth to early fifth centuries. 4 The evidence from Canterbury, suggests that the industry commenced between 350 and 370. C. IC Late, hand-made, grog-tempered ware with profuse, coarse siltstone inclusions. Vessels in this fabric appeared in East Kent around 270. A large quantity at Lympne5 suggests a production source nearby. Manufacture continued through the entire fourth century. C. ID Soapy-orange-brown, hand-made 'Patchgrove' ware. West Kent fabric, used for storage jars late first to early third centuries but rare at Faversham. C.2 Hand-made fine sanded fabric with crushed calcined-flint inclusions. Almost entirely associated with bead-rimmed jars of pre-Flavian date. 6 C.3 Hand-made shell-tempered ware. An early fabric represented by a few jar body sherds from first-century features. 176 C.4 C.5 C.6 C.6A C.7 C.8 C.9 C.10 C.11 C.! IA C.12 C.12A C.13 C.14 EXCAVATION AT SYNDALE PARK, OSPRINGE Wheel-turned, fine sanded black fabric with white quartz and mica inclusions. Probable first-century date. Variable sandy brown to brick-red fabric fired brownish-grey to black externally, with or without mica-dusting. Second-century date. Two wasters amongst the assemblage indicate a probable production site at Faversham. Some forms are identical to those produced by the Canterbury sandy grey-ware industry; possible that outlying kilns of that industry were sited at Faversham.7 Brickearth digging in 1910 destroyed 'early brick or pottery kilns' east of the Roman settlement at Faversham.8 Wheel-turned, coarse-quartz-gritted grey-ware. From first century, but only significant during third century. An East Kent fabric. Similar to C.5. They may have a common origin. Very fine sanded version of Fabric C.6, probably same source. BB I. Vessels in this Dorset fabric are very rare in the assemblage. Coarse-quartz sanded pimply high-fired fabric with occasional flecks of calcite and red-ferrous inclusions. A late third/fourthcentury fabric of Kentish origin. The coarseness and occasional presence of body riling suggests that this fabric is a Kentish equivalent of the Overwey one, manufactured on the HampshireSurrey border and elsewhere in South-East England 3 30-400+. Very fine sanded grey-ware with reddened core and occasional black slipped surfaces. Finer than C.8 but coloration suggests the same source. Some forms similar to later fourth-century Alice-Holt examples. Very fine sanded to sand-free Hadham grey-wares. Imported into Kent from the late third century, more significantly in the late fourth century. Similar to fabrics C.11 and 13. Very fine sanded Alice Holt grey-ware.9 Coarse sanded Alice Holt grey-ware. 10 Coarse quartz-sand tempered buff to cream ware. Typical 'Overwey' fabric. A very fine version of fabric C.12. Very fine quartz-sanded, wheel-turned grey or black-ware. Characteristic of Thames-side products, Second-to early Fourthcentury date, 11 probably from several production centres. A wheel-turned grey-ware 'dump' category. Mayen ware.12 Distinctive, pimply stoneware from the Rhineland. Fabrics - the finewares: F.lA F.IB F.2 F.3A F.3B F.3C South Gaulish samian. Central Gaulish samian Wheel-turned, sand-free Upchurch ware.13 Large quantities brought in between 70 and 150. Oxfordshire red and brown colour-coated wares.14 Oxfordshire white-slipped wares.15 Oxfordshire whitewares.16 177 LUCY SIBUN F.4A Lower Nene Valley orange fabric with brown-black colour-coat. 17 F.4B Lower Nene Valley whitewares with brown-black colour-coat.18 F.5 Miscellaneous sand-free whitewares. F.6 Very fine sanded orange fabric with matt grey-black colour-coat. Of probable local or Colchester origin. F.7 Very fine to sand free red fabric with white slip. Mainly used for flagons and butt-beakers manufactured at Hoo and on the Upchurch marshes.19 F.8 Similar fabric but without white slip. F.9 Cream to buff fabric with rouletted decoration. Used for buttbeakers and other forms imported from North Eastern Gaul before AD 70. F.10 Very fine buff fabric with soft, red ferrous inclusions and a highly polished exterior with rouletted decoration. F.! I Hadham red oxidised wares. F.12 German marbled wares. Pale buff to cream fabric fired pinkishbrown with patchy 'sponged', brown colour-coat. A flagon associated with an AD 313-3 I 5 coin was recovered from Ospringe cemeteries. 20 F.13 Pompeiian Red ware. 21 Manufactured in coastal regions of Belgium, mid-first century to c. 250. Pottery Groups - Pre-Flavian Virtually all pre-Flavian contexts contained sherds of Fabric C.2, accompanied by furrowed jar fragments in fabric C. IA. No large single group of early material are present. The more diagnostic pieces include: (Fig. 6) Bead-rimmed jar in orange-brown fabric C.2 with carinated shoulder and diagonal combing on the body. Type 303-2 (AD 50-70). (All types quoted without reference are from Monaghan 1987).22 Context 11. 2 Cordoned jar rim in similar fabric. Context 67. 3 Bead-rimmed jar in brown fabric C. IA. Context 59. 4 Imitation butt-beaker rim in grey fabric C.13 Context 11. 5 Carinated bowl in sandy, black fabric C.5 with rough, micaceous surfaces. Similar to type I !BI-I (AD 40-70). Contextl3. 6 Small, bead-rimmed beaker in grey fabric F.2. Context 59. Pottery Groups - AD 70-150 There are many more assemblages belonging to this phase than pre-Flavian ones. The largest is from Pit 32 (fill 33). This contains pottery with a date range of AD 43-200, although most is before AD 150. The large percentages of hand-made Fabrics C. IA and 2 (67% of all the coarse pottery) indicate that the bulk of the pottery is pre-Flavian and FI avian 178 EXCAVATION AT SYNDALE PARK, OSPRINGE /---- 􀀂z􀀃􀀄􀀅 \. 􀀏---􀀐I ;' ,, 1K, ,􀀄􀀅 􀀈I -----􀀉· 􀀊 .,-- ,I ( 􀀄,,C) \JW / ,ft------_=-;\ \ ',--,"I /) ' ' 1------,---, \ l .I \.'\ .f 7 )- :' ,/ t~ ,,I - -_ 􀀉 .f- J '" "\ ·,· -----,;::=::=::\· r---- .1 􀀉􀀊 . --. -􀀋 , i ' _' } f -(1 ,,I . ( 􀀈====􀀉1--3 L=====􀀈I J .) .. I ()-- 20 • 0 5cm , , Fig. 6 Syndale Park: Pottery Finds 1-2 J. 179 LUCY SIBUN in date, although the fabric F.2 and F.7 vessel forms are Flavian and secondcentury. Hand-made grog-tempered vessels in the latter fabric continued to dominate the coarseware market at Faversham until the mid-second century. The following diagnostic sherds came from Context 33: (Fig. 6) 7 Bead-rimmed jar in grey-brown fabric C.2 fired black externally, with horizontal shoulder combing. 8 Bead-rimmed jar in black fabric C.1 A Type 3F4 (70-150) 9 Large, everted rim storage jar in smooth orange-brown Patchgrove type fabric C. ID with burnishing on the shoulder. l O Rim from one of several everted-rim storage jars in black fabric C. 1 A with combed or furrowed bodies. 11 Straight-sided dish in similar fabric. 12 Rim from cordoned jar in grey-brown fabric C. I 3. 13 Dish with beaded-rim in grey-black fabric C. I 3. 14 Pie-dish in grey fabric C. I 3. Type 5D5 (110-170). 15 Several conjoining fragments from fine-cordoned bowl in similar fabric. Type 4Jl-3 (43-120). 16 Straight-sided dish with knife-trimmed base, in grey fabric C.6A fired reddish-grey. A cracked and warped kiln waster. Two other extreme waster sherds in a red cored Fabric C.5 variant are also known from this feature. 17 Flagon rim in red fabric F7 with white slip. Type IE2-6 (60-120) 18 Flagon rim in red fabric F7 with white slip. Type IE6-1 ( I 00-200) This pit also contained fragments from two South-Gaulish Samian Dr.27s, a Dr.18 and an I 8R of pre-Flavian and FI avian date. Other vessels of note from this period include the following: (Figs 6 and 7) 19 A jar in brick-red fabric C. I A with black patches and band of incised crosses on the shoulder. Pit 105 (Fill 106). 20 Everted-rim jar in similar fabric with a diagonal slashing on the shoulder and underside of rim. Pit I 20 (Fill 121). 21 Bowl with reeded rim in gritty-black fabric C.5 fired dirty brown. Pit 120 (Fill 121). This form was produced at Canterbury from c. 90 onwards and remained as part of that industry's range of products until c. 180.23 22 Biconical in fabric F2 with applied blue-grey wash. Type 202-2 (43-100) Ditch 88 (fill 89). 23 Flat-rimmed bowl in highly burnished black fabric F.2. Type SBS (60-130). Post-hole 90 (fill 9 I). 24 Dish in grey-fabric F.2 Type 7 A2 (43-140). Post-hole 90 (fill 91). Pottery Groups - AD 150-270 There are considerably fewer Antonine to early third-century pottery assemblages from the excavations. These come mainly from two pits in Trench B. All assemblages are too small for meaningful quantification. Pit 103 con- 180 EXCAVATION AT SYNDALEPARK, OSPRINGE ., J -\ 7􀀄· :1 r\ ] _) ,, J ,/ \ ,.I /' \' J--J "=\ _) --==- -3s1-(􀀇 .. 􀀇 I /·􀀈- J 7 ' 26 7 -􀀋 \ \ 21! - , ( ' 381 < 􀀈-- 􀀉I 􀀉I /' 􀀊- J _/" , I \J '] J ·\ 30 ,, ' " --- .l--J -􀀅.􀀆-, -40 􀀄1 􀀅􀀆-, \.-- - 31 I 1 ) .t --------r- 42 --1-,,􀀅\ O 5cm Fig. 7 Syndale Park: Pottery Finds 22-42. 181 LUCY SIBUN tained sherds of BB 1 (Fabric C7) as well as fragments from a Pompeiian Red platter (F.14 ). This pit also contained a fragment from an Oxfordshire brown colour-coated beaker with rouletting and applied white barbotine scrolling. Pit 114 contained the following diagnostic sherds: (Fig. 7) 25 Cavetto-rimmed jar in brown fabric C. 1 A. 26 Hand-made fabric C. JA cordoned-bowl fired grey-brown with exterior burnishing. 27 Lid-seated bead-rim jar in a black-cored reddish-brown gritty fabric C.5 variant. 28 Pie-dish in similar fabric with highly polished surfaces. 29 One of two everted-rimmed jars in brown sandy fabric C.5 fired dirty grey. 30 Lid-seated jar rim in sandy grey-brown fabric C.5 fired dirty reddish- brown. Paralleled in form at Canterbury in a late secondcentury context. 24 31 Flat-rimmed bowl in similar fabric with a slightly micaceous surface. Pottery Groups - AD 270-400+ Group 5: Ditch 18 had small amounts of first- to second-century pottery in its lower fills, but the upper one (fill 20) had a large assemblage of fourthcentury sherds (5.530 kg.). Some of the material is of late third- to early fourth-century date, but other pieces are later than 370. The assemblage is the largest from the entire site and its mainly mid to late fourth-century date is indicated by the high percentage of late East Kent grog-tempered wares. Fabric C. I B accounts for almost a quarter of the coarse-wares and C. IC for a further 16%. Compared with the products of the early Belgic grog-tempered ware industry (Fabric C.1 A) there is a considerably greater emphasis on bowls and dishes in both of the late industries. Another characteristic of late grog-tempered ware is a marked lack of dry-storage vessels. Fourth-century forms in unattributed grey sandy wares (Fabric C. I 3) account for nearly a quarter of the coarse pottery. Pollard is of the opinion that grey, wheel-turned wares ceased being produced on any scale in East Kent after the mid-fourth century, with very little other than local grog-tempered wares and imported Alice Holt products circulating there after 370.25 Some jars in high-fired pimply grey fabric C.8 may be fourth-century products from the ill-understood Preston kiln near Wingham.26 This appears to have produced copies of fourth- century Alice Holt forms, although all but one of the pots found are now lost. The red-cored copies of Alice Holt forms 5B.8 and 6A.12 in fabric C.9 may well have a similar origin. (Figs 7 - 8) 32-33 Everted-rim jars in pale grey fabric C. IB fired patchy-black/ grey/brown. 34 Beaded and flanged bowl in similar fabric. 35 Cavetto-rimmed jar in grey-black fabric C. I C. 182 EXCAVATION AT SYNDALE PARK, OSPRINGE 36 Bowl with rudimentary flange, in similar fabric. 37 Dish in similar fabric. 38 Everted-rim jar in grey fabric C.13. 39 Convex-sided beaded and flanged bowl in medium grey fabric C. I 3 with fine burnishing. 40 Straight-sided dish in black fabric C.13 with external grooving just below the rim. 41-42 Beaker rims in grey fabric C.13 43 Cavetto-rimmed jar with rim-edge moulding, in reddish, sandy fabric C.8 fired grey. 44 Straight-sided dish with rim-edge beading, in similar fabric. 45 Flagon in soft, grey fabric C l 3. 46 Hook-rimmed jar in buff-cream fabric C.12 with rim-edge blackening. 27 47 Beaker rim in grey-cored orange Oxfordshire fabric F.3A with brown colour-coat. 0 5cm \. ml 7 \. J / Fig. 8 Syndale Park: Pottery Finds 43-55. 183 LUCY SIBUN 48 Mortarium in white slipped fabric F.3B. This deposit also contained a fragment from another Oxfordshire mortarium in fabric F.3C. 49 Straight-sided dish in white Nene Valley fabric F.4B with brown colour coat.28 50 Fragment from indented beaker in very fine orange fabric F.6 with reddish-brown colour-coat. There is the left leg and part of a pleated tunic from a moulded figure occupying one of the indentations. Indented beakers with moulded figures of deities within the indentations have been discussed by Webster.29 Several examples from a variety of sources are known from SouthEast Britain and mainly of third to fourth-century date. The best known example from Kent is the one from Richborough.30 The Richborough fragments are in a very different fabric from the Faversham piece, but two other fragments from Aylesford are in a similar orange ware with dark-red colour-coat. The fragment from the Faversham excavations has its pleated tunic paralleled on a fragment in Peterborough museum. Webster suggests that the Peterborough figure's tunic is a stylised representation of Vulcan's apron.31 That deity is, however, usually depicted with heavy boots, absent from the Faversham fragment. We cannot be sure of the deity represented, but Mars or Apollo are most likely (Fig. 8). Pit I 6: the pottery from Pit 16 appears to be of exclusively late fourthcentury date. Nearly three-quarters of all the coarse pottery consists of hand-made grog-tempered ware, with a small number of mainly late fourth-century forms in other fabrics making up the rest of the assemblage. The following sherds were of note: 51 One of five everted-rim jars in grey-black fabric C. IB. 52 One of two beaded and flanged bowls in similar fabric to 103. 53 One of three deep convex-sided dish/bowl in similar fabric. 54 Hook-rimmed jar in coarse-grey Alice Holt/Overwey fabric C.l lA.32 55 Convex-sided dish in buff fabric C. I 2A with black patches. The changing pattern of pottery supply The first-century settlement at Faversham acquired wares from both East and West Kent. Up to half of all the coarse pottery of pre-AD 70 date consists of grog-tempered furrowed jars from the Canterbury area and smaller, but significant, quantities of handmade bead-rimmed jars in fabric C.2 thought to have been manufactured on the Upchurch marshes to the west.33 Other wares of native type in fabric C.4 may also have a western origin, as similar vessels are known from early levels at the Eccles villa in the valley of the Medway. Sandy grey jars and lid-seated bowls of North Gaulish type were produced at Canterbury before AD 70. Another Continental potter seems to have es- 184 EXCAVATION AT SYNDALE PARK, OSPRJNGE tablished a kiln at Faversham at the same time. A carinated and mica-dusted bowl in underfired sandy black fabric C.5 is present in Post-hole 13 and more obvious wasters are known from Flavian deposits at Syndale Park. This indicates that the settlement at Faversham had attained enough importance by AD 70 to attract a potter producing high quality wares. Pre-Flavian finewares are limited in range but include small quantities of Gallo-Belgic whiteware butt-beakers from the Amiens region and South Gaulish samian. The Flavian and early second-century pottery assemblages from Syndale Park show slow changes in coarse pottery supply. Sherds in the flinttempered fabric C.2 are almost certainly residual, but large scale supply of grog-tempered and furrowed jars from the Canterbury region seems to have continued. Some sandy wheel-turned greywares from both West and East Kent are also present in relatively small quantities and in the former case include early-to-mid second-century pie-dishes of Thames-side types. The local potteries producing wares in fabric C.5 continued to function during this period and the early-mid second-century transitional Belgic Grogtempered/ Native Coarse Ware jars from the site may also have been made locally. The presence of such vessels indicates the continuing strength of the grog-tempered ware tradition in the Faversham area and elsewhere in East Kent at a time when the production of such wares in West Kent was largely at an end. Flavian finewares include South Gaulish samian imports, but the most significant supplier was now the Upchurch kilns. Upchurch wares had made their first appearance at Syndale Park just before AD 70, in the form of beadrimmed beakers, but now accounted for nearly a quarter of all the pottery. There is some disagreement as to when sand-free grey Upchurch ware was first made. Pollard considers that these wares first appeared after AD 70, whereas Monaghan opts for a Neronian origin.34 • 35 The evidence from both Faversham and Eccles strongly supports the latter author's views. The Syndale Park assemblages of this date also include 'Hoo' flagon fragments in a white-slipped oxidised version of the grey Upchurch fabric. The Antonine and early third-century pottery assemblages were small, but the large number of Antonine and early third-century pots from the nearby Ospringe cemeteries help to fill in some of the details of pottery supply to Faversham during the period. Local kilns continued producing wares in fabric C.5, but the bulk of the coarse pottery coming into Faversham now consisted of sandy grey wares from the Thames-side potteries and hook-rimmed and roll-over rimmed jars in the coarse sanded fabric C.6 of similar or more local origin. Grog-tempered storage jars continued to be made locally at first, but were then replaced by similar vessels in Pollard's higher fired Native Coarse ware. This industry is an Antonine and third-century phenomenon in East Kent. Small amounts of Dorset BB l began to appear at Faversham during the early third century. Quantities at Faversham are so small as to make it difficult to believe that the settlement was a profitable market for the distant Durotrigian potters and their negotiatores. It may be, however, that these wares were not the main commodity being marketed but were a small part of ship-borne mixed cargoes of salt, Kimmeridge shale jewellery, Purbeck 185 LUCY SIBUN stone and other items known to have been produced around Poole Harbour in Dorset. Antonine and early third-century finewares from Syndale Park and the Ospringe cemeteries include Pentice beakers from the Amiens region, Central and East Gaulish samian, Moselkeramik from Trier, Lower Nene Valley and Colchester finewares and some early examples of Oxfordshire red-colour-coated vessels. The Pompeian Red ware platter rim from Pit 103 is unusual and indicates trade links with the Belgian coast during this period. The Dressel 20 amphora is the most common type in Britain, but it is clear that the inhabitants of Faversham had access to unusually large numbers of such containers. The settlement may have used harbour facilities on the Faversham creek for olive oil importation from southern Spain, alternatively a local merchant may have been trading in some commodity which was packaged in re-used Dressel 20 containers. The second hypothesis is supported by the fact that one of the cemetery amphorae had a small hole plugged by a lead rivet and another had the rim cut down. The late third- to fourth-century pottery assemblages from Syndale Park reveal considerable changes in supply during the period. During the late third and early fourth centuries Faversham continued to acquire most of its coarse pottery in the form of sandy grey wares from the Late Thames-side industry pottery producers of West Kent and similar industries in the east of the civitas. Native Coarse wares at Faversham were replaced by the products of another grog-tempered ware producing industry, operating near the shore fortress at Lympne and distributing its handmade products in significant quantities across East Kent.36 Most of the fine wares and mortaria reaching Faversham now emanated from the Oxfordshire kilns on the Middle Thames and included vessels in red-colour-coated, white-slipped, white and parchment wares.37 The few continental pottery imports on British sites after the mid third century tend to be concentrated in the south-east, particularly Kent. The Syndale Park pottery includes fragments of German marbled ware and these indicate late trading connections with the Rhineland, either direct or via other ports, such as Richborough where other marbled ware fragments have been found. Radical changes in pottery supply took place during the late fourth century. The Thames-side grey ware industry ceased production and a new handmade grog-tempered ware producing industry sprang up in East Kent, with production centres at Richborough and possibly Canterbury.38 The Lympne grog-tempered ware industry continued making its handmade wares and together these two centres saturated the region with their crude handmade products. At least one grey sandy ware producer did survive, at Preston near Wingham to the north-east of Canterbury. Its high fired and semi-vitrified wares include blue-grey riled jars and other forms clearly influenced by contemporary Alice Holt and Overwey products, now entering the region from a source well over 200km away on the Hampshire-Surrey border. All of these wares are present in Ditch 18 and other late assemblages from Syndale Park, together with Hadham pottery from north of London. Oxfordshire colour-coated wares continued to be significant here and elsewhere in East Kent. 186 EXCAVATION AT SYNDALE PARK, OSPRINGE Late fourth-century Continental imports at Faversham are few and far between but include a coarse pimply Mayen ware bowl from the Rhineland. The impression is given of a shortage of good quality coarse pottery at Faversham and elsewhere in East Kent after 370. The local inhabitants were content to largely make do with crude grog-tempered kitchen wares of Iron Age appearance, with any good quality wheel-turned coarse and fine wares having to be brought in small quantities from far away and well beyond the limits of the normal distribution zones of the potteries in question. This phenomenon is encountered in late fourth-century pottery assemblages from sites all over south-east Britain and suggests declining output and demand for pottery. The collapse of a monetary economy during the early fifth century would have closed down most remaining pottery producers and there is every indication that Kent became almost totally aceramic between c. AD 420 and the arrival of Jutish potters in the middle years of the fifth century. THE COINS by David Rudling A total of 24 Roman coins were recovered from the Faversham excavations. Of these, three date to the third century AD with the remaining 21 dating to the fourth century. Full details and a catalogue are housed in the archive. One possible unstratified coin flan was recovered during the excavations (Fig. 9: l 0). This flan, with a diameter of 13mm and weight of 2.16g was a circular copper alloy mould-casting, with part of the sprue remaining. The upper surface is rough and flat, the underside smooth and domed. If this is a coin flan then it might date to the late Iron Age rather than Roman period. Flan production is discussed by Van Arsdell.39 METALLURGICAL REMAINS AND METAL WORK by Luke Barber Nineteen contexts of all periods produced a total of 193 pieces of slag and hearth lining, weighing 10.3kg. This material is fully listed, along with samples of the slag varieties, in the archive. All the slag is the result of secondary working of iron and no tap slag is present. The quantity of iron forging slag strongly suggests iron-working was being undertaken close to the excavation area. Pit 22 (Fill 23) contained the single largest group, totalling 91 slag and five hearth lining fragments, weighing 4.1 kg. Within this group were the remains of three hemispherical forge bottoms ranging in diameter from a very small example at 90mm to a more usual size of 134mm. Iron The majority of the iron, which totalled 78 pieces, consisted of nails or nail fragments. By far the most common type of nail is a round flat, or very low 187 LUCY SIBUN domed, headed variety with square-sectional shank. This general purpose nail type, with head diameters ranging from 10-24mm, appears in contexts spanning the first to fourth centuries. Two headless nails are also present (Contexts 1 and 102) as well as a single hobnail (Context 72). The remaining ironwork, totalling 22 pieces, and consists of sheet fragments and unidentified pieces from larger objects. Only five pieces are diagnostic of form. These include a double-spiked loop from Context 20,40 two wall hooks from Contexts 74 and 126,41 and an iron knife blade with a small loop at the end of the handle, similar to one from the Lullingstone villa.42 The knife from the present excavations came from an undated post-hole (Context 125/126) the Lullingstone example was from a fourth-century gully. Non-ferrous metalwork The thirty-seven non-ferrous metalwork items, recovered from 12 different contexts, consisted almost entirely of copper alloy objects; only four pieces of lead are present. All this material is fully listed in the archive. A selection of the more diagnostic pieces are illustrated in Fig. 9 and catalogued below: Fragment of a copper alloy armlet of rectangular sectioned-sheet (1 x 3mm) decorated with punched dots within marginal grooves. A similar, but complete example was found at Colchester.43 Context I. 2 Broken copper alloy hair-pin with ovoid knob head with incised line decoration and double incised line around collar. The decoration on the knob head consists of paired incised lines arranged in a cross pattern and conforms to Cool' s Group 12.44 A virtually identical, though slightly larger example, is illustrated. 45 These pin types have been found to concentrate in North Kent. Context 1. 3 A virtually complete, but bent, copper alloy hair-pin with slightly flattened ovoid knob head decorated with diagonal incised lines. Three further incised lines around collar. Although the Faversham example bears some resemblance to both Cool's groups 13 and 19, the latter group appears to be the more likely as the knob head does not have a concave base. Examples similar to the Faversham one have been found at Richborough but remain undated.46 Context 1. 4 A faceted-headed copper alloy pin decorated with ring and dot motifs on all but the top and two lower facets. Similar pins have been found at Canterbury,47 and Thetford,48 where they are ascribed a middle or late Saxon date. Context 1. 5 A simple copper alloy pin formed from a tapering wire rod with a slightly conical finial. Possibly a hair-pin. This example would fit best with Group 24,49 although the Faversham example comes from a late context whereas Group 24 is considered to be predominantly of second-century date. The pin may, however, be residual. Cut 45, fill 46: late fourth century. 188 -1 -- . .• 1 EXCAVATION AT SYNDALE PARK, OSPRINGE -· 2 I I 0 ,, - I II - I I -• 3 5 5cms 8 Fig. 9 Syndale Park: Metalwork Finds 1-10. 189 6 • LUCY SIBUN 6 A copper alloy needle with elongated threading eye. First-early second-century. Pit 36, Fill 47. 7 The remains of a tinned copper alloy brooch of Hod Hill type with heavy cross-moulding on the bow. The hinge, pin and part of the catch plate are missing. A similar example without side-lugs is published elsewhere.50 Firstcentury AD. Pit 36, Fill 47. 8 A fragment of an arm from a pair of copper alloy tweezers with two gradually diverging incised lines on the outside edge. A similar example has been found from Colchester. 51 Pit 52, Fil! 53. 9 A copper alloy brooch of Dolphin Type with missing pin and heavily corroded spring.52 The foot is plain but the large wings are ribbed. The catch plate has a triangular perforation. First century AD. Pit 114, Fill 115 (Trench B). GLASS by John Shepherd An hexagonal-sectioned bottle was recovered from context 33 (fill of pit 32, Fig. 10) The vessel belongs to the well known group of containers dating from the mid-first to early third centuries which included the ubiquitous square-sectioned bottle. The hexagonal form, which was not as common as its square counterpart, was a feature of the late first and early third century repertoires in, especially, the north-western provinces.53 These bottles were used primarily as transit or storage containers of liquid comestibles and cosmetics. The design of the vessel was intentional to make it a sturdy container. A single circle with a central dot design in found on the base. This is not common on hexagonal bottles. It is possible that it was used as some form of identification mark either by the glass house or any individuals concerned with the bottling and distribution of the vessel and its contents. Fig. 10 Syndale Park: Glass vessel. 190 EXCAVATION AT SYNDALE PARK, OSPRINGE DISCUSSION The results of the recent excavations and previous discoveries in the area indicates the existence of a substantial Romano-British settlement concentrating around Ospringe and Syndale Park. It has been suggested that this could be the Roman town of Durolevum, which is recorded as being 16 Roman miles from Rochester and 12 miles from Canterbury. 54 This suggests that the site is situated along Watling Street, the Roman road that runs between these towns, along the line of the present A2. As there is little evidence for any settlement at the allocated point, and the settlement at Syndale Park is only 2 miles further east, it is likely that this site is actually part of Durolevum. The interpretation of the recently excavated features will add little to our knowledge of the settlement's infrastructure as a whole, given the limited extent of the trenches. However, a large occupation area is suggested by the wide area of densely packed features within the excavation. Dense intercutting features are present along the length of Trench A. To the north, in Trench B, and closer to Watling Street, features are located along the length of the trench but are more sparsely spaced with little intercutting. The absence of features within the assessment trench TT4 (Fig. 1) indicates that the area containing a high concentration of features is limited to Trench A. It is therefore possible that the southern edge of the settlement is located just to the south of Trench A, or TT4 represents an open space within the settlement. However, the lack of Romano-British features located in trench TT3 would make the former seem more probable. The lack of material noted o n the south side of the A2 to the west of Trench B is odd considering discoveries to the north. It is possible an open space existed here or the settlement did not extend much further to the west at this point. The latter theory would explain the presence of the cemetery (Fig.1) which is likely to have been placed just outside the settlement area. Similarly the cemetery noted to the east may mark the settlement's eastern extent although further excavation would be needed to test this. The settlement appears to be concentrating around the junction of the A2 and the B2045 and it is possible a road from Faversham or Oare Creek joined Watling Street at this point. Context 98, interpreted as a possible road, was approximately 35m from the line of Watling Street and from the small area visible in the trench, it appears to run parallel with it. It is possible that the road originally joined Watling Street or was an internal thoroughfare through the settlement. However, only area excavation or a systematic trial-trench sample could trace the extent to which the settlement displayed any planned layout or internal organisation. 191 LUCY SIBUN Previous excavations in the area produced material dated to the second and third centuries AD, and generally lacked material from the first and last centuries of Roman occupation. 55 The pottery from the recent excavations bas identified two main phases of occupation, the first from the beginning of the Roman period through to the middle of the second century and the later phase from AD 270 onwards, with a particular concentration of material from the middle and late fourth century. There is clearly a gap in the pottery sequence from the recent excavations from between AD 150 and 270. Only a few sherds dating to this period were recovered and these are mostly from features in Trench B. These results are in sharp contrast to those of the previous work which concluded that there is evidence for a minor settlement on both sides of Watling Street during the second and third centuries AD. Perhaps this is evidence of a localised shift in settlement after the first century, further to the west and spreading to the north side of the road, returning to the south and further east for the fourth-century occupation. The possible reasons for the shifts in the concentration of settlement, if they did occur, remain unknown. No definite structural remains could be identified during the excavations but roof, floor and box-flue tiles were all recovered. Previous work undertaken to the north of the A2 concluded that buildings would have been of timber construction and the recent excavations would seem to support this. However, the presence of roof tile fragments would suggest that at least some of the buildings had tile hangings and the hypercaust tiles suggest that there were some high status buildings within the settlement, some of which at least are likely to have been of masonry construction. The diversity and quantity of artefacts recovered from the excavations has provided a greater insight into some of the activities which may have been taking place on, or in close proximity to the site. This allows for a better understanding of the status and economy of the settlement. Most of the activities that have been suggested or identified from the finds seem to have been taking place during the first and second centuries. The pottery wasters suggest that there was a local production site, a fact that is further suggested by Whiting, 56 mentioning the destruction of Roman brick or tile pottery kilns to the east of Faversham. The forging of iron during the same period is also suggested by the slag recovered during the excavations, but the narrow trenches did not allow for any suggestion to be made about the precise location of this activity. The presence of these industries indicates that the site was in a favourable position not only for raw materials but also for marketing goods. It is quite possible other trades such as bone and 192 EXCAVATION AT SYNDALE PARK, OSPRINGE leather working and textile production were also practised. The overall impression is of a fairly large community surviving in a semiurban settlement which contains both reliance on the land and industry and trade for its existence. The site was on the edge of two ceramic market areas and as such benefited from both. However, continental imports never made up a large percentage of the present excavated assemblage suggesting that the settlement, or at least parts of it, was not of a very high status. Despite this, continental connections are always present, particularly with Gaul and the Rhineland, suggesting the settlement did have continuing long-distance contact. How much of this was direct or via Canterbury cannot yet be ascertained without further excavations. The faunal remains consisted primarily of cattle and sheep (totalling 90%) with a much small assemblage of pig (5%). These would have formed the main part of the diet, with a greater emphasis on pig during later centuries. Their presence suggests that animal husbandry was taking place near to the settlement and butchery marks suggests that much of the processing was also undertaken on site. The agricultural economy appears primarily pastoral. Only two charred seeds were recovered during environmental processing, both from grassland species suggesting the presence of pastoral land nearby. However, three fragments of quern and millstone recovered imply that crop processing was undertaken. The fragment of millstone recovered from the fill of Cut 30 suggests that this processing was taking place on a relatively large scale. Excess flour would easily have been transported by road or sea (via Faversham Creek) to markets at Canterbury or Rochester. More environmental evidence is therefore needed in order to establish the importance and diversity of any arable cultivation connected with the site. The large quantities of oyster shell recovered indicate that this was a relatively important source of food, not surprising given the proximity of the site to the Swale. Relatively small quantities of mussel, whelk and carpet shell were also recovered. The site at Syndale Park is obviously of great importance to the archaeology of the area. More information has been added by the pipeline excavations to the already existing body of data from the site. However, until the full extent of the settlement is established and large areas of it exposed little can be said of its position within the settlement hierarchy of the area and indeed its own internal morphology. Much more environmental work will need to be undertaken in the future to assess the economy of the site. 193 LUCY SIBUN ACKNOWLEDGEMENTS The author would like to thank British Gas (South Eastern) for funding and co-operation throughout the project; Lis Dyson of Kent County Council for monitoring the work at all stages; Mark Gardiner and Luke Barber for managing the project; the excavation team and all the specialists for their contributions to the report. Elizabeth Somerville would also like to thank Helen Crudgington and David Dunkin who did the initial identification and measuring of the shells. NOTES 1 Priestley-Bell, G., I 994, An Archaeological Watching Brief on part of the Faversham to Sittingbourne Gas Pipeline. SEAS Project No: 1994/46. Trenches TT! and TT2 were located to the south of the church of SS Peter and Paul centred on TR 0003/ 6022 and TQ 9986/6022 respectively. 2 Kirk, L., I 996 (with contributions from Malcolm Lyne, David Rudling, Luke Barber, John Shepherd, Elizabeth Somerville and Pat Hinton), An Archaeological Excavation at Syndale Park, Ospringe, Faversham, Kent. SEAS Project No; 160. Archive Report. May. 3 Rose, J. A., 1967, 'Report - Faversham Archaeological Research Group', Kent Arch. Review, 9, 6. 4 Wasters of this fabric were found by the author amongst pottery from Chalk House, Richborough. J. P. Bushe-Fox, 1949, Fourth Report on the Excavations at the Roman Fort at Richborough, Kent, Rep. of Res. Cte of Soc. of Antiq. of London, I 6, Oxford; Lyne, M. A. B. (1994), Late Roman Handmade Wares in South East Britai11, unpublished PhD Thesis, Reading University. 5 Young, C. J., 1980, 'The Pottery', in Cunliffe, B. W., 'Excavations at the Roman Fort at Lympne, Kent I 976-1978', Brita11nia, 11, 227-288; (Reduced-ware 8). 6 Pollard, R. J., 1988, The Roman Pottery of Kent, KAS Monograph V, p. 46. 1 Ibid., fig. 28, 68-70. 8 Whiting, W., Hawley, W. and May, T., 1931, Report on the excavations of the Roman Cemetery at Ospril!ge, Kent, Rep. of Res. Cte of Soc. of Antiq. of London, 8, Oxford. 9 Lyne, M.A. B. and R. S. Jefferies, 1979, The Alice Holt/Farnham Roman Pottery lndus1ry, CBA Research Report No. 30, Fabric CID. ,o Ibid., Fabric CID. 11 Monaghan, J., 1987, Upchurch and Thameside Roman Pottery, BAR (B) 173. 12 Fulford, M. and J. Bird, 1975, 'lmported Pottery from Germany', Britannia, 6, 171-181. 13 Monaghan 1987, op. cit. (see note 11), Fabrics Nl and 2. 14 Young, C. J., I 977, Oxfordshire Roma11 Potlery, BAR (B) 43, 1230. 15Jbid.,117. 16 Ibid., 56. 17 Howe, M. D., Perrin, J. R. and D. F. Mackreth, 1980, Roman Pottery from the Nene Valley: A Guide, Peterborough City Museum, 0cc. Paper No. 2. 18 Ibid. 194 EXCAVATION AT SYNDALE PARK, OSPRINGE 19 Pollard 1988, op. cit. (see note 6), 61. 20 Whiting e1 al. 1931, op. cit. (see note 8). 21 De Lael, S. J. and Thoen, H., 1969, 'Etudes sur la Ceramique de la Necropole Gallo-Romaine de Blicquy (Hainault)', Studia Historica Gandensia, 114. 22 Monaghan I 987, op. cit. (see note 11 ). 23 Pollard 1988, op. cit. (see note 6), 77. 24 Wilson, M., I 987, 'IIIB The Pottery', in S. S. Frere et al., Canterbury Excavations, Intra and Extra Mural Sites 1949-55 and 1980-84, The Archaeology of Canterbury VIII, 195-219, fig. 85:20. 25 Pollard I 988, op. cit. (see note 6). 26 Dowker, G., 'Roman Remains at Preston near Wingham', Archaeologia Cantiana, 12 (1878), 47-8. 27 Lyne and Jefferies, op. cit. (see note 9). Class 3C, dated 330-420. 28 Howe et al. I 980, op. cit. (see note I 7), Form 87. 29 Webster, G., 1989, 'Deities and religious scenes on Romano-British pottery', J.R.P.S., Vol. 2, 1-28. 30 Bushe-Fox, J. P., 1932, Third Report on the Excavations of the Roman Fort at Richborough, Kem, Rep. of Res. Cte of Soc. of Antiq. of London, I 0, Oxford, 185-6, Pl. XLIII. 31 Webster 1989, op. cit. (see note 29). 32 Lyne and Jefferies, op. cit. (see note 9), type 3C.l l, dated 300-420. 33 Pollard I 988, op. cit. (see note 6), 46. 34 Ibid., 59. 35 Monaghan 1987, op. cit. (see note 11 ), 217. 36 Lyne 1994, op. cit. (see note 4 ), 419. 37 Young 1977, op. cit. (see note 14). 38 Lyne I 994, op. cit. (see note 4), 435. 39 Van Arsdell, R. D., 1989, Celtic Coinage in Britain, London. 4° Crummy, N., 1983, 'The Roman Small finds from excavations on Colchester 1971- 9', Colchester Archaeological Report, 2, Colchester Archaeological Trust, No. 4065. 41 Ibid., No. 4076. 42 Meates, G. W ., 1987, The Lullingstone Roman villa, Vol. ll - The Wall Paintings and Finds, KAS Monograph III, Fig. 43 No. 243. 43 Crummy 1983, op. cit. (see note 40), No. 1709. 44 Cool, H. E. M., 1990, 'Roman Metal Hair Pins from Southern Britain', Archaeological Journal, vol. 147, 148-182. 45 Ibid., Fig. 8, No. 3. 46 Ibid., 168. 47 Blockley, K. et. al., 1995, Excavations in the Marlowe Car Park and Surrounding Areas, The Archaeology of Canterbury V, fig. 444 No. 480. 48 Rogerson, A. and C. Dallas, 1984, 'Excavations in Thetford 1948-59 and 1973-80', East Anglian Archaeology Report, 22, Norfolk Archaeological Unit, Norfolk Museums Service, Fig.112 No. 45. 49 Cool 1990, op. cit. (see note 44). 50 Hattatt, R., 1989, Ancient Brooches and other Artefacts , Oxbow, Fig. 189, No. 83. 51 Crummy 1983, op. cit. (see note 40), No. 1883. 52 Hattatt 1989, op. cit. (see note 50), Fig. 157. 195 LUCY SIBUN 53 Shepherd, J. D., 1982, 'The glass fragments from Southfield Road', in Clarke, G. 'The Roman villa at Woodchester', Britannia, 13, 197-228. 54 Philp, B. J., 1976, 'The Probable Site of Durolevum', Kent Arch. Review, 43, 62-64. 55 Ibid.; Rose 1967, op. cit. (see note 3). 56 Whiting et al. 1931, op. cir. (see note 8). 196