A Belgic-Early Roman Site at Great Mongeham, near Deal

A BELGIC-EARL Y ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL KEITH PARFITT In 1980 Mr L. Danstead of Cherry Lane, Great Mongeham, excavated a new duck pond in the field adjoining his house (Fig. 1). During the course of this work he discovered two large sherds of pottery, subsequently reported to the writer. Examination indicated that they came from a necked jar of Belgic-early Roman date and a visit to the field resulted in the recovery of a further fifty-six Belgic and Roman sherds from the spoil-heaps around the new pond. This clearly suggested the presence of a hitherto unknown site. During the spring of 1981 members of the Dover Archaeological Group, under the direction of the writer, hand-excavated a series of forty-nine trial trenches in the field adjacent to the pond in an attempt to establish something of the nature and extent of the site implied by the pottery finds (Rankov I 982, 395). The work indicated that much of the material discovered around the pond came from a first-century AD enclosure ditch [F. I, see below]. Traces of a series of other ditches, gullies and pits were also revealed [Fs 2-17; Table 1), indicating that the enclosure formed part of a more extensive complex (Fig. 2). The area investigated during 1981 was located in a smaJI grass-field to the north-east of Mustapha, between Cherry Lane and Church Path (OS Parcel No. 7636). NGR TR 3475 5135. This lies about 200m south-east (downhill) of the parish church. The field is situated on sloping ground which falls away to the south-east and the north-east, and stands between 18-21m above OD (Fig. 1). The natural subsoil here is an orange-brown brickearth and all the archaeological features located were cut into this. They were sealed in most places by a layer of brown loamy hill wash. In 1985 a further six trenches were cut ahead of the construction of a new dwelling (Meadow View) fronting onto Cherry Lane, between Langdon House and Fieldfare, immediately to the south-west of the area previously examined (Fig. 1). Five more features were located here [Fs 18-22, Table 1], all ditches and gullies, except for a pit [F. 127 N 00 100 200 Metres Fig. 1 Map showing location of the excavated site. A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL 18]. None of these features appeared to represent continuations of those previously discovered and none produced any datable material, although there seems little doubt that they are broadly contemporary with the remains examined in 1981 (only F. 19 appears on site plan, Fig. 2). THE EXCAVATED FEATURES The excavations in 1981 and 1985 revealed a combined total of twenty-two individual archaeological features (Table 1; Fig. 2). Parts of three probable enclosure ditches [Fs 1, 2 & 12] were examined, along with sections of three other ditches [Fs 17, 19 & 22], eight gullies [Fs 3-5, 13-15, 20 & 21 J and eight pits [Fs 6-11, 16 & 18}. Just over half these features produced datable finds and there can be no doubt that they relate to a significant late Iron Age and RomanoBritish occupation site, the full extent of which was not established. The undated features are presumably also related to this settlement. Almost all the features located were sealed by a substantial deposit of hillwash. This reached a maximum thickness of Im and produced a total of 117 pot-sherds, mostly of Belgic and Roman date, together with two early Anglo-Saxon pieces (Parfitt and Brugmann 1997, 9) and a small amount of sandy medieval ware. A number of prehistoric struck flints and calcined flints were also recovered, together with two copper-alloy objects, a pottery spindle whorl, eleven Roman tile fragments, a few pieces of medieval peg tile, nine pieces of burnt daub, animal bone, four iron nails and two fragments of sandstone (see below). This mixed artefact assemblage is clearly derived from further up the hill towards the church and suggests activity there over many centuries. Enclosure Ditches [Fs 1, 2 & 12) The most extensive features examined during the excavations consisted of three ditches [Fs l , 2 & 12]. From their positioning (Fig. 2) it seems unlikely that these are contemporary but their exact sequence is not certain. Ditch [F. 1] appeared to delimit a rectangular enclosure and cut across the line of both [Fs 2 & 12} indicating that it was the latest. Ditch [F. 2] also seemed to form part of an enclosure, occupying an area rather larger than that defined by [F. 1]. The L-shaped ditch [F. 12] may have formed part of a third enclosure positioned in the same general area during another phase. Thus, parts of three successive ditched enclosures could be represented. Only enclosure ditch [F. 1) produced significant quantities of pottery. The 129 No. Type 1 Enc. ditch 2 Enc. ditch 3 Gully 4 Gully 5 Gully 6 Pit 7 Pit 8 Pit 9 Pit 10 Pit 11 Pit 12 L-ditch 13 Gully 14 Gully 15 Gully term. 16 Pit 17 Ditch 18 Pit 19 Ditch 20 Gully 21 Gully 22 Ditch TABLE 1. REGISTER OF EXCAVATED FEATURES (1-17 IN 1981; 18-22 IN 1985) * = minimum dimension; st/sl = steep/slope. Shape Length (m) Width(m) Axis Depth(m) Sides Base Rect 39.0x25.0* 0.90-2.75 - 0.50-1.14 St/sl Round Rect. 56.0*xl 0.0* 0.65-1.40 - 0.25-0.55 St/sl Round Linear 1.00* 0.65 NE-SW 0.16 Slope Round Linear 1.00* 0.85 N·S 0.18 Slope Round Curved 4.50* 0.55-0.65 N-S 0.08-0.14 Slope Round Oval 2.75 0.60* NW-SE 0.32 Slope Dished Oval 8.30 2.60 NW-SE 0.30 Slope Flat Oval 1.35* 1.20 NE-SW 0.37 Slope Dished Circ. Dia.= 2.45 - 0.48 Slope Flat Oval 0.72* 0.70* NE-SW 0.43 Slope Round ?Circ. Dia = 1.30 - 0.30 Slope Round Linear 58.00* 0.40-1.90 NE-SW 0.20-0.50 Slope Round Curved 4.50* 0.50-0.65 NE-SW 0.21-0.35 Slope Round Linear 1.00* 1.00 NW-SE 0.10 Slope Flat Linear 0.55* 0.45 N-S 0.25 Slope Round Circ. Dia. = 0.46 - 0.14 Slope Flat Linear 8.00* 1.30 NW-SE 0.65 Slope Round Oval 1.10 0.95 E•W 0.20 Slope Dished Linear 0.70* 0.90 NW-SE 0.37 Slope Flat Linear 4.10* 0.30* NE-SW 0.18 Slope Flat Linear 0.80* 0.37 NE-SW 0.18 Slope Dished Linear 0.76* 0.50* NE-SW 0.48 Slope Dished No. sherds [Assem./Fab) 706 [ Assm 2-4] 6 [ Assm I] 5 [IAl, IA3, B2] No finds l [ B2] 2 [Assm 7] 11 [Assm 5] 11 [B2, B8, R42) 15 (Assm6] 6 [B2, RS, R l 6] No finds l [R43] 4 [B2, RS0] No finds No finds No finds 15 [B2, R l /5/6.1/43] No finds No finds No finds No finds No finds A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL SB -'a􀀅- Gully F.14 􀂰 -----􀀞-:::􀀟=􀀠:i-..--􀀡􀀢%􀀣:-:_-:􀀤_,r,, ::GullyF.3 􀂱 .. .::::. 􀀑::- .. -􀀖􀀗􀀘,-\,... - .... , ... ] ---- 􀂲',PitF.10 ', '􀂳' Gully _ i===:8􀀍􀀎􀀏:􀀐,\ \􀀑' PitF.􀂴\\s3 -:􀂵􀂶uUyF.4 F 13 OW ,---j-- • •!&L -􀂷•L._:􀂸:- . 'I'-'- , ... ;;:w:-=i L-==Ft 􀀐 1•\r- .... ,,' \ \'_:PitF.8 : \ GullyF.5. ': \ Pit F.7 o· \ \ 17 , , ,, Cl!L􀀆 C)I::] ...b s11'. \ s1• 1 ' - Pit • , 1 • • : Enclosure Ditch F.11 : \ \ \ 􀀉 : 1 F1 􀀊 I I I 1 , ' I CIII? \ \ l 􀀠-- -. - 􀀡 S15\ \ D ',', --·􀀉·- 1\ I 1 \ 1 ef/1 I ', ', Enclosure ' 1 V ' ' I' \\ Ditch\\ (f ' /1, L-shaped \ \ F2 􀂹 1 l DitchF .12.\ 1 1 1 1 \ f/1 , Pond 􀂺 I \ I I 1 I \ I D \ 􀀊 I I 􀀑 \ ;\; 􀀇 \ 1􀀈\ )\ I I .. '1􀀨 I ' ' '\ 􀀋\. I.. \ 􀀙 􀀚 /4􀀛 􀀋 ' \ \ 􀀌'--=-'..=..-..::--"􀀍 􀀉 􀀊 \ \ , , S_10 1 ' S18 '' ,' ' __ i _____ ,-------;f;IJl---,􀀋 ,/ \j: °ullyF.15 I --------------- --􀀎 ------􀀏--(-- \\ c::J -' 􀀋23 I I \ : I I I ' I I I I 827 \ : /it F.16 i:::----􀀌-=􀀍.:-= DitchF .17 10 0 I I \ I \ I 11 1 I 1\ I I 11 11 S19 􀂻--, • I \,I, Ditch F.1􀂼: 􀂽 :: 10 Fig. 2 Overall site plan. 131 garden of Langdon House 20 Metres KEITH PARFITT two preceding ditches yielded much smaller amounts of material implying that they lay some distance away from any contemporary occupation areas. Perhaps they defined animal compounds or fields. The pottery dating evidence from [F. lJ indicates that this enclosure was occupied from about AD 30-100 (see below, Assemblages 2-4 ). The general lack of material from [Fs 2 & 12} makes their dating more difficult. None of the few pot-sherds recovered from the filling of [F. 2] (Assemblage 1) has to be later than AD 30 but a single piece of Roman tile found in the upper filling must be regarded as intrusive. Ditch [F. 12] yielded only a single sherd of second-century samian ware. This also must be treated as intrusive, unless, contrary to the records made in the field, the ditch is actually later than [F. 1]. Ditch [F. 2] This was traced in eleven trenches (Fig. 2) and consisted of a U-shaped ditch which appeared to form two incomplete sides of a rectangular enclosure. From the stratigraphic evidence it would seem to pre-date enclosure ditch [F. l], and must have surrounded a larger area. On the south-east side, the ditch was traced for a minimum distance of 56m. A rounded east corner was located and partially examined, although [F. lJ appeared to cut through it at this point (Fig. 4, Sect. 8). From the corner, the north-eastern arm of [F. 2] was traced for a minimum distance of 10m and clearly continued on into the adjacent field. The ditch varied from 0.65- l.40m in width and was 0.25-0.55m deep. It was filled with various clay deposits, probably representing largely natural silts, although perhaps including some deliberate backfill. These layers produced just six sherds of late Iron Age pottery (Assemblage 1) and a fragment of Roman tegula, probably intrusive. Ditch [F. 12] This ditch was located in thirteen trenches and was L-shaped in plan (Fig. 2), but varied considerably in size and profile. It seems quite probable that it formed the northern part of another ditched enclosure. On the north-western side the main arm of the ditch ran NE-SW and was traced for a minimum distance of 58m. It appeared to be cut by ditches [Fs 1, 13 & 17], and itself cut through an earlier pit [F. 11]. At the north-east end the main arm made a sharp right-angled turn to the south-east and was traced for a further minimum distance of 7 .60m. This north-eastern arm consisted of no more than a gully, 0.40-0.S0rn wide and up to 0.20m deep. The north-east end of the main arm seemed to be only slightly larger, although it had been 132 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL partly destroyed here by a later gully [F. 13). As it proceeded southwestwards the main north-western arm increased in size, reaching a maximum width of 1.90m. It was only l .00m wide where it was cut by ditch [F. 17] at the south-west end. The north-west arm was up to 0.50m deep and was mostly filled with grey and brown clays (Fig. 3, Sections 15 & 23) which were devoid of any finds, except for a Dr. 37 samian base of second-century date, found where it was cut by [F. 17]. This is likely to be intrusive and is most probably derived from [F. 17] itself (see below). Ditch [F. 1] This was located in sixteen trenches and consisted of a U-shaped ditch, forming three sides of a rectangular enclosure, with two rounded corners. It bounded an area measuring 39m (SW-NE) by at least 25m (SE-NW). The fourth, north-west, side of the ditch was not located and must lie in the adjacent field. A 17m length of the ditch on the south-east side had been destroyed during the construction of the pond in 1980 (Fig. 2). On the south-western side the ditch was about 2.65m wide and between 0. 70-1.1 Orn deep (Fig. 4, Sect. 18). It appeared to cut through two earlier ditches [Fs 2 & 12] (see above). Considerable quantities of pottery (Assemblages 3 & 4) and 'chaff-tempered' ware, representing dumped domestic rubbish, came from the upper and middle fillings of the ditch here. The south-eastern arm had been cut through by the pond and seems to have been the source of much of the pottery initially recovered from the spoil-heaps. Where sectioned on the south-west side of the new pond it was 2.75m wide and 1.14m deep (Fig. 4, Sect. 10) but to the north-east it was considerably smaller, being only 1.40m across and 0.70m deep (Fig. 4, Sect. 1). Pottery was less frequent here, the brown clay fill apparently representing largely natural silt. Near the east corner the ditch was cut by an oval pit [F. 6] which produced a sherd of late Roman pottery (see below; Fig. 3, Sect. 3). On the north-east side, the ditch was between 0. 90-1.40m wide and 0.50-0.70m deep. It was filled with a series of brown clays, which produced only five pot-sherds. The east corner of enclosure ditch [F. 2) appeared to be cut by this section of the ditch (Fig. 4, Sect. 8). Overall, the filling of the enclosure ditch produced a total of 706 pot-sherds and these constitute a very useful stratified sequence of local wares (Assemblages 2-4). More than half the material was contained within the upper filling of the ditch (Assemblage 4 ), especially on the south-western side. A significant collection of chaff-tempered ware was also recovered, together with some animal 133 NW SE SW NE nat b/earth Seetion No.3 SE NW Enclosure Ditch F .1 Section No.27 SE NW SE NW ..,, .j:>. I Section No.11 NW SE Section No.15 Section No.23 Section No.19 0 2 Metres Fig. 3 Sections across enclosure ditches, Fs 1, 2 & 12 and other features (Sect. Nos 3, 11, 15, 19, 23, & 27). NE Section No.8 SE Section No.1 SW SW Enclosure Ditch F.1 Enclosure Ditch F.1 NW Section No.18 SE Section No.10 ._._._..,....Jo'---------'--------'-2-------'3 Metres Fig. 4 Sections across enclosure ditches, Fs I and 2 (Sect. Nos 1, 8, 10 & 18). NE NW Enclosure Ditch F.1 KEITH PARFIIT bone. A few fragments of Roman tile came from the upper filling of the ditch, together with part of a cast copper-alloy bracelet and a possible clay mould (see below). A total of eighty-six sherds of pottery were recovered from the lower fillings of the ditch (Assemblage 2) and the date of these suggest that it was in use around the mid-first century AD, probably at about the time of the Roman Conquest. The 222 sherds from the middle fillings (Assemblage 3) are of a similar date and imply that sections of the ditch had been quite rapidly in-filled with dumped domestic rubbish. The upper fillings yielded about 400 sherds (Assemblage 4), which included Romano-British fabrics for the first time. There are also four pieces of Roman tile. It would seem that the ditch had been completely in-filled by the end the first century AD. Straight Ditch [F. 17] This U-shaped ditch was traced in two trenches for a minimum distance of about 8.0m. It was 1.30m wide and 0.65m deep (Fig. 3, Sect. 27), appearing to cut through ditch [F. 12] (see above) and a pit [F. 16]. The grey-brown clay filling produced a total of fifteen Roman pot-sherds, including fragments from two second-century vessels of samian ware (Dr. 3 I and Dr. 33). Too little of this ditch was seen to be certain of its purpose but it might conceivably have formed part of yet another enclosure. If so, this is likely to have been of the second century AD, post-dating enclosure ditch [F. 1]. Gullies [Fs 3-5, 13-15, 20-21] Short sections of eight separate gullies, aligned on various axes, were located (Fig. 2; Table I). Seven of these lay outside the areas delimited by [Fs 1 & 2] and are unlikely to be related to these main ditched enclosures. Gully [F. 13], situated within the area enclosed by [Fs 1 & 2), cut through the north corner of L-shaped ditch [F. 12] and so must be later. It could be contemporary with either [F. 1] or [F. 2]. From the limited sections revealed, it is impossible to determine the precise function of any of these gullies, although they clearly indicate that the main enclosure ditches formed part of a more extensive complex of features. Sufficient lengths of [Fs 5 & 13] were seen to indicate that they were curved. Only three of the gullies [Fs 3, 5 & 13] produced any pottery. Five sherds came from [F. 3]; three were flint-tempered and two grogged. A single grog-tempered sherd came from [F. 5], whilst curving gully [F. 13] produced four sherds - three grog-tempered Belgic pieces and a Roman Dressel 20 amphora fragment. 136 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL Pits [Fs 6-11, 16 & 18] Portions of eight fairly shallow pits of varying sizes were located (Fig. 2; Table 1), although most were not fully excavated. Two [Fs 11 & 16] were cut by later ditches and one [F. 6] was cut into the top of the latest enclosure ditch [F. 1] (Fig. 3, Sect. 3). Five of the pits produced small groups of pottery, all including some Roman wares (Assemblages 5-7). Of particular interest from F. 6 (Assemblage 7) was a single sherd of late third/fourth-century Oxfordshire ware found in the upper filling, together with a second century samian fragment (Dr. 33). The Oxfordshire piece is perhaps the latest Roman sherd from the site. The small quantity of finds recovered from the various pits implies that they were not extensively used for the dumping of domestic refuse, a view perhaps reinforced by their shallow depths. The clay fillings suggest that the majority silted up naturally. From the substantial area they covered, it seems possible that some, like [Fs 7 & 9), could represent small quarry pits. FINDS The project archive consists of 62 recorded deposits, 22 recorded features (Table 1), two site plans, 33 measured sections and four boxes of finds from the 1981 excavations. All the finds from the site have been placed in store at Dover Museum, together with a copy of the field records. Prehistoric Struck Flints and Calcined Flints (not illustrated) by Keith Parfitt A total of thirty-eight prehistoric struck flints was recovered from the excavations. Seventeen of these came from hillwash deposits sealing the site and a similar number were recovered as residual material from the filling of enclosure ditch [F. 1]. The flints are all in a fresh condition with little or no patina; locally collected downland flint provided the raw material. There is one core but the bulk of the pieces are waste flakes. Ten flakes show some evidence of retouching but the only recognisable tools are a scraper, a piercer and a large unstratified blade with a worked notch. In the absence of any particularly diagnostic tool-types, a broad Neolithic-Bronze Age date may be tentatively suggested. Flints of this general date occur extensively in the Deal region. At least casual occupation around the present site must be indicated. 137 KEITH PARFITT Forty-nine generally large and unabraded calcined flints were also recovered from the site. Most of these were contained within the hill wash deposits but ten came from the filling of enclosure ditch [F. 1] with two more from gully [F. 3). It seems probable that the bulk of these are residual and most are likely to be contemporary with the struck flints discovered. Objects of Iron and Copper-alloy (Fig. 5) by Keith Parfitt Seven pieces of iron were recovered, most of which were heavily corroded. The identifiable items were all simple nails. Three objects of copper-alloy were discovered:- i) Nail-cleaner with stamped decoration, in the form of a diagonal cross, on both sides. Surviving length, 32 mm (Fig. 5.1). Crummy (1983) Type 2a, datable to the mid-later first/second century AD. Lower hillwash. GM-81-5. ii) Small fragment of a plain, heavy ?ring. Lower hill wash (not illustrated). GM-81-5. iii) Fragment of cast bracelet with pear-shaped cross-section. Upper filling of enclosure ditch, F. 1 (not illustrated). 2 1 Fig. 5 Small finds from the excavations (scale 1: 1). Pottery (Figs. 6 & 7) by Malcolm Lyne The excavations produced a total of 970 sherds of pottery (11,022g). With the exception of a few Anglo-Saxon and medieval pieces recovered from the upper levels, all this material is of late Iron Age 'Belgic' and Roman date. The bulk of the material (706 sherds) came 138 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL from the filling of enclosure ditch [F. 1) with a further 117 pieces from the hill wash layers sealing the site and another fifty-six from the spoil-heaps around the new pond. All the assemblages were quantified by numbers of sherds and their weights per fabric. Fabrics were identified using a x8 magnification lens with inbuilt metric scale for determining the natures, sizes, shapes and frequencies of added filler inclusions. The numbered fabric series for the calcined-flint and related mixed grit fabrics drawn up for the nearby Green Lane, Whitfield site (Lyne 2002) is also used here with additions: 'Belgic' grog and sand tempered, Gallo-Belgic imports and Romanised wares are given the codings created by the Canterbury Archaeological Trust for pottery in east Kent (Macpherson-Grant et al 1995). Late Iron Age Fabrics IA l Irregular handmade fabric with profuse ill-sorted up to 5mm IA2 IA3 IA4 IA7 IA8 BI B2 B2.1 B3 B4 B5 B8 B9 calcined-flint filler. Similar fabric but with sparse up to 3mm calcined-flint filler. Polished black fabric with profuse silt-sized quartz and moderate to profuse up to 1mm calcined flint filler. Similar fabric but with very sparse calcined-flint filler. Handmade fabric with profuse up to 2mm calcined flint filler. Handmade underfired black fabric with sparse, ill-sorted, up to 2mm fossil shell and polished surfaces. ?Neolithic. 'Belgic' fine-grog-tempered ware. 'Belgic' coarse-grog-tempered ware. 'Belgic' coarse-grog-tempered ware with pale siltstone grog. 'Belgic' grog-tempered ware with additional sparse flint. 'Belgic' grog-tempered ware with additional rounded chalk inclusions, from the Folkestone area. 'Belgic' grog- and sand-tempered ware with occasional shell flecks. An early variant of Fabric B2. 'Belgic' fine-sanded handmade fabric from the Folkestone area. 'Belgic' soot-soaked coarse-sanded ware. Gallo-Belgic Fabrics B12 Terra Rubra Fabric lA. BER6 ELG Early Gallo-Belgic Whiteware (Rigby and Freestone 1995, Fabric B). BER7 Early Gallo-Belgic Whiteware (Rigby and Freestone 1995, Fabric IB). BERll Flagon Whiteware (Rigby and Freestone 1995, Fabric WWI). BER14 Buff ware. 139 • KEITH PARFITT Roman Fabrics R l R5 R6.l R6.3 R16 R23.3 R27 R42 R43 R47 R50 R64 R75 R91 LR7 LRl0 Native Coarse Ware, c. AD 170-300. Canterbury coarse-grey sand-tempered ware, c. AD 70-175. Canterbury coarse-orange sand-tempered ware, c. AD 70- 200+. Canterbury coarse-buff sand-tempered ware, c. AD 70-200+. Fine grey Upchurch fabric, c. AD 43-270. Pompeian Red ware: Fabric 3, c. AD 43-150. Mica-dusted ware. South Gaulish La Graufesenque Samian, c. AD 43-110. Central Gaulish Samian, c. AD 120-200. Italian Dressel 2-4 amphorae. Dressel 20 amphora fabric. Rhenish Whiteware mortarium fabric. Miscellaneous fine cream wares. Flagon Wbiteware (Rigby and Freestone 1995, Fabric WW6). Oxfordshire Parchment ware, c. AD 240-400+. Oxfordshire Red Colour-coat, c. AD 240-400+. Assemblage - Phase 1, c. 50 BC - AD 30 No. 1: From the fill of Enclosure Ditch F. 2 (GM-81-16 & 29) The eleven cuts across the line of this feature were largely sterile but a total of six sherds (72g) of pottery came from fill contexts 16 and 29. These comprise one oxidised fragment in sandy Fabric B8, four sherds in grog-tempered Fabric B2 and one very abraded piece in calcined-flint tempered Fabric IA4. There are no diagnostic sherds and all that can be said is that a late Iron Age date is probable by virtue of the feature's relationship with later, better-dated features and the presence of 'Belgic' grog-tempered ware . Assemblages - Phase 2, c. AD 30 -100 No. 2: From the lower fills of Enclosure Ditch F. 1 (GM-81-14, 28, 39,43,45, 52, 57 &60) The lowest ditch fills cut by the various sections through the feature yielded a total of eighty-six sherds (1002g) of pottery, including large fresh fragments from the following vessels:- Fig. 6.1 Bead-rim in black Fabric B8 with an external buff margin and horizontal combing on the body. Ext. rim diameter 160mm. There is a small perforation through the wall of the vessel immediately below the rim, probably for suspension. GM-81-28. 140 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL Fig. 6.2 Everted-rim jar with rippled shoulder of Thompson (1982) type B2-1 in soapy smooth black Fabric Bl. Ext. rim diameter 140mm. c. 50 BC-AD 50. GM-81-43. Fig. 6.3 Bead-rim jar of Thompson ( 1982) type Cl-1 in black Fabric B2 with combing on the body. Ext. rim diameter 160mm. There are fourteen joining sherds of the vessel from this context and another from the middle fills of the ditch (GM-81-42). c. AD 30-60. GM-81-45. Fig. 6.4 Pedestal-base in polished black Fabric B3. GM-81-52. Fig. 6.5 Bead-rim jar of Thompson (1982) type C4 in black Fabric B2 with combing on the body. c. AD 30-100. GM-81-60. The assemblage also includes fragments from the following vessel and suggests that the ditch was cut around the time of the Roman Conquest. Fig. 6.6 Fragment from small white-slipped red-brown Central Gaulish flagon or beaker (Rigby and Freestone 1995, Fabric WS 3). c. 25 BC-AD 50. GM-81-39. Further fragments came from the middle and upper fills of the ditch (GM-81- 59 and 58, respectively). No. 3: From the middle fills of Enclosure Ditch F. 1 (GM-81-27, 38, 42, 51 and 59) The 222 sherds (29 lOg +) of pottery from these fills forms a large enough assemblage for quantification by numbers of sherds per fabric: it is unfortunate that the large butt-beaker sherd in Fabric BER7 was mislaid before weighing. TABLE 2. NUMBERS OF SHERDS PER FABRIC FROM MIDDLE FILLS OF ENCLOSURE DITCH [F. 1] Fabric No. Sherds % Weight(g) IA3 2 0.9 10 BI 28 12.6 384 B2 151 68.0 1,978 B2.1 12 5.4 146 B5 7 3.1 140 B8 14 6.3 216 Bl2 1 0.5 14 BER7 1 0.5 - WS3 5 2.2 18 Misc 1 0.5 4 Total 222 (100) 2,910+ 141 KEITH PARFITT ) l􀀇JJI (, 7 8 10 l I (/,,lr\ { 9 􀀃 􀀁 J 13 I 􀀃' 1 t ) I \ 15 ) 16 I \ ) 17 I =r - -- · , l\li ) · 19 I . r, .. 18 .. cm. Fig. 6 Pottery from the excavations (Assemblages 2 - 4). 142 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL The assemblage is fairly well sorted with no large numbers of sherds from single vessels to distort the percentages. The only two calcinedflint tempered sherds are heavily abraded and obviously residual in an assemblage totally dominated by Belgic grog-tempered wares (89 per cent). Small amounts of sandy 'Belgic' pottery come from the Folkestone area (6 per cent): there are a few sherds from Gallo-Belgic imports, including a fragment from a closed form in Terra Rubra TRl(A) fabric (c. 15 BC-AD 25) and the vessel described below (Fig. 6.14). An absence of Romano-British grey and finewares suggests a date-range similar to that for the lower ditch fills of around the time of the Roman Conquest. Fig. 6.7 Jar of Thompson (1982) type B2-1 in grey Fabric B2 fired black. Ext. rim diameter 140mm. c. 50BC-AD 50. GM-81-38. Fig. 6.8 Storage-jar of similar type in patchy grey/black Fabric B2. Ext. rim diameter 280mm. GM-81-42. Fig. 6.9 Jar of similar type with everted rim and two narrow cordons on the upper shoulder in hard pale-grey Fabric B5 fired brown with black surface patches. Ext. rim diameter 180mm. GM-81-59. Fig. 6.10 Jar of Thompson (1982) type B3-1 in black Fabric B2. Ext. rim diameter 120 mm. c. 50 BC-AD 100. GM-81-38. Fig. 6.11 Bead-rim jar of Thompson ( 1982) type C 1-2 in black Fabric B2 with body combing. Ext. rim diameter 180mm. c. 50 BC-AD 100. GM-81-38. Fig. 6.12 Platter of Thompson ( 1982) type G 1-1 copying GalloBelgic CAM 1 form in polished black Fabric BI. c. 20 BC-AD 50. GM-81-38. Fig. 6.13 Butt-beaker of Thompson ( 1982) type 05-6 in black Fabric Bl fired polished brown. Ext. rim diameter 140mm. c . AD 1-50. GM-81-59. Fig. 6.14 Gallo-Belgic butt-beaker of Stead and Rigby (1989) Type 2B2 in pale-grey Fabric BER7 with cream-buff surfaces (Rigby and Freestone 1995, Fabric IB). Ext. rim diameter, 117mm. c. AD 10-40. GM-81-38. Fig. 6.15 Narrow-necked flask in polished orange/black Fabric B l . Ext. rim diameter 85mm. First century AD. GM-81-59. No. 4: From the upper fills of Enclosure Ditch F. 1 (GM-81-3, 20, 26, 36, 37, 40, 41, 44, 50, 56 & 58). The 398 sherds (4181g) of pottery from these contexts is large enough for quantification by numbers of sherds and their weights per fabric. There is a far larger range of fabrics in this assemblage than in that from the middle fills of the feature. These fabrics include Romano- 143 KEITH PARFITI TABLE 3. NUMBERS OF SHERDS PER FABRIC FROM UPPER FILLS OF ENCLOSURE DITCH [F. 1]. Fabric No. Sherds % Weight(g) % IAl 2 0.5 36 0.9 IA2 8 2.0 112 2.6 IA7 I 0.2 28 0.7 BI 13 3.3 124 3.0 B2 261 65.6 3032 72.5 B2.1 6 1.5 82 2.0 B3 7 1.8 120 2.8 B4 5 1.3 78 1.9 B8 48 12.1 316 7.5 B8+chalk 2 0.5 28 0.7 B9 1 0.2 14 0.3 BER6 1 0.2 4 0.1 BER9 1 0.2 2 0.1 R5 10 2.5 48 1.1 R6.1 3 0.8 4 0.1 Rl6 22 5.5 116 2.7 R27 4 1.0 16 0.4 R91 2 0.5 19 0.5 WS3 1 0.2 2 0.1 Total 398 (100) 4181 (100) British ones for the first time; indicating that rubbish was still being dumped in the top of the ditch until at least the end of the first century. Small quantities of pottery were supplied from both the Upchurch and Canterbury kilns and there is an increase in the percentage of, now Romanised, sandy black wares from the Folkestone area (13 per cent ). Grog-tempered 'Belgic' wares remain by far the most significant single component of the assemblage. Fig. 6.16 Jar of Thompson ( 1982) type B 1-1 with everted rim and broad shallow groove above the shoulder in black Fabric B4 fired brown. Ext. rim diameter 180mm. GM-81-56. Fig. 6.17 Necked jar of Thompson (1982) type Bl-4 in hard grey Fabric B2fired black. Ext. rim diameter 160mm. GM-81-50. Fig. 6.18 Small jar of Thompson (1982) type B2-3 in black Fabric B2 variant with sparse grog. Exterior surface grey-brown with burnished chevron decoration on the shoulder. A similar example comes from Mill Hill, Deal only a short distance to the south -east (Thompson 1982, 692, no. 817). Ext. rim diameter 130mm. GM-81-40. 144 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL 21=0 ====\􀀈, re r-- 􀀆iii#\\ \ 23 21 ) 1====_=r7 , _ .,/ 24 22 ) i ........... f 25 ) 26 J '>, 27 28 ' 32 30 ( 31 J 34 \ ) 0 10 cm. 33 Fig. 7 Pottery from the excavations (Assemblages 4 - 6). Fig. 6.19 Jar with everted rim and single neck cordon in orange Fabric B8 variant with sparse vesicles from leached-out chalk, fired polished black. Ext. rim diameter 160mm. GM-81-58. Fig. 7.20 Upper part of barrel-shaped bead-rim jar of Thompson (1982) type B5-3 in black Fabric B2 with external polish. Ext. rim diameter 120mm. c. 50 BC-AD 50. GM-81-58. Fig. 7.21 Bead-rim jar of Thompson (1982) type Cl -2 in black Fabric B2 with body furrowing. GM-81-41. Fig. 7.22 Bead-rim jar of type C4 in grey Fabric B2 fired black with 145 1 L KEITH P ARFITf body combing. Ext. rim diameter 130mm. c. AD 30-100. GM-81-50. Fig. 7.23 Platter of Thompson (1982) type G l -6 imitating CAM.7 in grey Fabric B2 fired polished black. Ext. rim diameter 160mm. c. AD 43-70. GM-81-58. Fig. 7.24 Another example in polished brown/black Fabric B2. Ext. rim diameter 220mm. c. AD 43-70. GM-81-50. Fig. 7 .25 Base from pedestal urn or bowl in brown/black Fabric B2. The upper part of the vessel has been deliberately cut away above the pedestal with the remaining portion inverted for use as small cup. GM-81-41. Fig. 7.26 Butt-beaker copy in dark-grey Fabric B8 with pink margins and grey-brown surfaces. Ext. rim diameter 90mm. c. AD 70-100+. GM-81-56. Fig. 7 .27 Small lid-seated jar in orange cored Canterbury greyware Fabric R5. Ext. rim diameter 110mm. c. AD 70-175. GM- 81-50. Fig. 7.28 Lid in Canterbury greyware Fabric R5. Ext. rim diameter 200mm. c. AD 70-175. GM-81-50. Fig. 7.29 Small bowl or cup in grey Upchurch fineware Fabric Rl6, ?copying South Gaulish samian Dr. 24. Ext. rim diameter 130mm. A similar vessel was found in Canterbury (Jenkins 1953, fig. 4.17). c. AD 43-70. GM-81-41. Fig. 7.30 Necked-bowl of Monaghan (1987) type 4Jl .3 in similar fabric, with elaborate body moulding. Ext. rim diameter 130mm. c. AD 43-120. GM-81-41. Fig. 7 .3 I Raised pedestal base from vessel of uncertain type in cream/pink fabric with very sparse up-to 0.3mm quartz and crushed orange grog filler. GM-81-40. Assemblages - Phase 3, c. AD 100-300 No 5: From the fill of Pit F. 7 (GM-81-31) This quite large pit produced a mere six sherds (88g) of pottery, comprising three residual fragments in calcined-flint tempered Fabric IA 7, two fragments from a jar with corrugated neck in grogtempered Fabric B2.1 and the following piece:- Fig. 7 .32 Rim from flagon in pale grey Fabric R6. l fired rough pink. Ext. rim diameter 110mm. A second-century AD Canterbury kilns product. No. 6: From the fill of Pit F. 9 (GM-81-33) The fifteen sherds (86g) of pottery from this feature are heavily 146 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL broken up but include a fragment of Gallo-Belgic Whiteware and the following:- Fig. 7.33 Platter copying CAM 12 form of Thompson (1982) type G 1-7 in black Fabric B 1 with polished surfaces. Ext. rim diameter 200mm. c. AD 50-100. Fig. 7.34 Beaker of Monaghan (1987) type 2A3 in black Fabric R16 with buff margins. Ext. rim diameter 100mm. c. AD 100-130. The last piece indicates an early-second-century date for the filling of the feature. No. 7: From the fill of Pit F. 6 cut into the fills of Enclosure Ditch F. 1 (GM-81-12). This feature produced just two sherds (34g) of pottery, comprising a fragment from a Central Gaulish samian Dr. 33 cup (c. AD 120-200) and a burnt fragment of an Oxfordshire Red Colour-coat wall-sided mortarium (Young 1977, type C97, dated c. AD 240-400). The mortarium fragment is one of the latest Roman sherds from the site and indicates that the enclosure formed by F. 1 had been abandoned and forgotten by the mid-third-century AD. Pottery from the hillwash (GM-81- 2, 4, 5, 8, 9, 10, 11, 15, 18, 19, 23, 34 & 35) The 117 sherds from the hillwash deposits covering the site are for the most part late Iron Age and Roman in date but include little Roman material which need be later than c. AD 200. There are, however, two sherds in what appears to be c. AD 450-650 dated chaff-tempered Early Saxon fabric from contexts 5 and 10. These are indicative of occupation in the vicinity of the site during that period (see below). Chaff-Tempered Ware (not illustrated) by Keith Parfitt and Geoff Halliwell Some 300 fragments (1kg) derived from small, crude ceramic vessels of chaff-tempered ware (Macpherson-Grant 1980) were recovered from the site. Apart from one fragment in the hillwash, all this material came from the south-western side of enclosure ditch [F. I]. The date-range of this quite distinctive ceramic-type seems to be entirely confined within the first century BC to first century AD and as here, it is consistently associated with 'Belgic' grog-tempered pottery. When the Great Mongeham site was first excavated this assemblage represented one of the largest groups so far recovered (Barford 1982) but the material is now well-known across east Kent, 147 KEITHPARFfIT with a larger assemblage recently being published from Green Lane at Whitfield (Parfitt 2002). Typically, the great bulk of the fragments from the present site are small. The overall size, shape and function of the vessels represented thus continues to be problematical. However, the crudely applied collars with protrusions or lugs first recognised at Green Lane (Parfitt 2002, 391) also occur in this assemblage. Almost one third of the pieces from Great Mongeham appear to represent fragments of collar or lug. There are four such fragments which are still attached to the rim. One of these includes a substantial projecting lug with a surviving length of 55mm. This piece is of some particular interest because the angle of attachment to the rim suggests that the lug cannot have projected horizontally, as inferred at Green Lane (Parfitt 2002, fig. 7). Although the exact angle of the rim is difficult to gauge, the lug would appear to have extended downwards, at an angle of roughly 45° to the body of the vessel. On this evidence it would now seem that these irregular projecting lugs can come off the rim at differing angles, but still more examples are required before the overall form can be determined. Objects of Fired Clay (Fig. 5.2) by Keith Parfitt 1) Roughly shaped spindle-whorl made from a re-used sherd of orange-red grog-tempered pottery (Fig. 5.2). Possibly unfinished. Upper hillwash. GM-81-4. 2) Two pieces of a small ?container with thick walls and base. Possibly a mould used in casting metal. Overall diameter, 41mm (not illustrated). Upper filling of enclosure ditch [F. I}. GM-81-58. Tile and Sandstone (not illustrated) by Keith Parfitt Sixteen pieces of Roman tile were found, most being small fragments. The hillwash layer produced eleven of the pieces, whilst the upper filling of enclosure ditch [F. 1] yielded four more. The upper filling of enclosure ditch [F. 2] contained a single piece of tegula. A few fragments of medieval peg-tile were also recovered from the hillwash. Most Roman tile fragments are of orange-red fabrics but four pieces in the hillwash and one from [F. l] are of a distinctive cream-pink sandy fabric. These can be paralleled exactly by material recovered from the small Roman villa at Sandwich, some 6.5km to the north-west (tile fabric a; Parfitt 1980, 241). Tiles of this colour and fabric were being manufactured at Eccles in the Medway 148 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL valley during the second/third century (Detsicas 1967, 170-17 4) and this appears to be the most likely source of the present material (identification kindly provided by Louise Harrison, CAT). Too few tile fragments were recovered to suggest the presence of any Roman buildings roofed with this material in the immediate vicinity of the excavated site (but see below). Seven fragments of sandstone were recovered from the excavations. The largest was a piece of Folkestone Greensand recovered from the middle filling of enclosure ditch [F. I]. This appears to be part of a quern and bears traces of a central, vertical hole. Its surface had been partially vitrified by intense heating after breakage. Two other small pieces of Folkestone Greensand came from the hill wash. The remaining fragments, all but one from enclosure ditch [F. lJ, are of a much softer sandstone, probably derived from the Thanet Beds. These pieces show no signs of working and seem too soft to be derived from quernstones or rubbers. Animal bone by Jill Bowers The excavations produced a total of sixty-nine pieces of animal bone (1.1kg), mostly teeth and unidentifiable fragments. Just over half the bone came from hillwash deposits. The filling of enclosure ditch [F. 1) produced thirty pieces, including the jaw of a young pig and teeth from pig and cow. The filling of enclosure ditch [F. 2] yielded a fragment of leg bone of uncertain type. DATING AND DISCUSSION It seems clear that the remains excavated off Cherry Lane formed part of a larger site. The stratigraphic and pottery dating evidence suggest that occupation began here just before the Roman Conquest and several subsequent phases of activity must be represented. The full extent of the site has yet to be revealed and it seems possible that the main focus of settlement lay on higher ground to the north-west (Fig. 1). The present area appears to comprise successive elements of the ditched fields and enclosures associated with a Belgic-early Roman farmstead, of which a number of examples have now been located in east Kent, at sites such as Faversham (Philp 1968, 62), Barham (Philp and Philp 1974) and Green Lane, Whitfield (Parfitt 2002). At Faversham the Belgic settlement there was subsequently replaced by a simple villa in the Roman style (Philp 1968, 67) and this general sequence of development is becoming increasingly familiar on sites 149 KEITH PARFITT throughout the South-East (Black 1987). Whether this was the case at Great Mongeham presently remains unclear but there is other evidence for late Iron Age and Roman occupation in the region. In 1939, during the digging of a cesspit for a new house about 250m north of the present site (at NGR TR 3477 5156; Fig. 1), evidence of a hearth associated with Belgic and early Roman pottery was recorded (Stebbing 1939; Phillips 1964a). This could suggest that either the present site continues this far north, or another more or less contemporary site lay immediately adjacent. Contained within the fabric of the nearby parish church of St Martin (Fig. 1) are about a dozen fragments of Roman tile (Jones 1992, 234) and a surface scatter of Roman roof-tile with some pottery, has been noted in a field north of the church (Frere 1990; Parfitt and Brugmann 1997, 9). A metal-detecting rally conducted across a large block of ground to the north and east of the church in 1993 allowed the recording of an extensive scatter of Roman coins here, mostly of late third- and fourth-century date. A slight concentration of these was noted northeast of the church (Esmonde Cleary 1994, 291). A single late thirdcentury coin of Carausius had been previously discovered in a garden off Northbourne Road, at NGR TR 3474 5154 (Fig. l; Phillips 1964b). The available evidence thus seems to suggest that there was extensive Romano-British occupation around the present-day village of Great Mongeham. The region subsequently became an area of early Anglo-Saxon settlement (Parfitt and Brugmann 1997, 8-9) and part of the reason for this may well lie in its Romano-British antecedents. Although a good start has been made, there is clearly much more archaeological investigation needed in and around the village before any detailed analysis can be attempted. ACKNOWLEDGEMENTS The writer would like to thank the respective landowners, Mr Leo Danstead and the late Mr H. Harrop, both of whom allowed unconditional access to their ground. Of the Group members who carried out the work, mention must be made of: Mrs Shirley Creasy, Mrs Joan Briggs, Miss Tina Avann (now Mrs Parfitt) and Messrs Ben Stocker, Geoff Halliwell, Mike Halliwell, Pat Mungovan, Howard Jones, David Holman, the late John Bray and the late Charles Burch, for their sustained efforts over a number of weekends. The keen interest of Mrs D. Jull is also gratefully acknowledged. Mr Barry Corke has prepared the site drawings for publication and Dr Malcolm Lyne has described and drawn the pottery. 150 A BELGIC-EARLY ROMAN SITE AT GREAT MONGEHAM, NEAR DEAL BIBLIOGRAPHY Barford, P.M., I 982, 'A New Type of Kentish Briquetage', Kent Arch. Rev., 69, 204-205. Black, E.W., 1987, The Roman Villas of South-East England (B .A.R. 171 ). Crummy, N., 1983, The Roman Small Finds from Excavations in Colchester 1971-9, Colchester Arch. Rep. 2. Detsicas, A.P., 1967, 'Excavations at Eccles 1966', Archaeologia Cantiana, LXXXII, 162-178. Esmonde Cleary, A.S., 1994, 'Roman Britain in 1993: Kent, Great Mongeham', Britannia, XXV, 291. Frere, S.S., 1990, 'Roman Britain in I 989: Kent, Northbourne', Britannia, XXI, 364. Jenkins, F., 1953, Roman Canterbury: An Account of the Excavations at No. 5 Watling Street, June-December 1947 (Canterbury Excavation Committee Report No. 6). Jones, H.A., 1992, 'A Survey of Roman Fragments in Churches in S. E. Kent', Kent Arch. Rev., 110, 230-235. Lyne, M., 2002, 'The Pottery', in Parfitt, K., 2002. Macpherson Grant, N., 1980, 'Chaff Tempered Ware', Kent Arch. Rev., 61, 2-4. Macpherson-Grant, N., Savage, A., Davey, M., Riddler, I. and Cotter J., 1995, Canterbury Ceramics 2. The Processing and Study of Excavated Pottery. Monaghan, J., 1987, Upchurch and Thameside Roman Pottery. A ceramic typology for northern Kent, first to third centuries AD (B.A.R. 173). Parfitt, K., 1980, 'A Probable Roman Villa on the Sandwich By-pass', Kent Arch. Rev., 60, 232-248. Parfitt, K., 2002, 'A Prehistoric Site off Green Lane, Whitfield, near Dover', Archaeologia Cantiana, CXXll, 373-396. Parfitt, K. and Brugmann, B., 1997, The Anglo-Saxon Cemetery on Mill Hill, Deal, Kent (Soc. for Med. Arch. Mono. 14). Phillips, A.S., 1964a O.S. Record card TR 35 S W 12. Phillips, A.S., 1964b O.S. Record card TR 35 SW 8. Philp, B., 1968, Excavations at Faversham, 1965 (First Res. Rep., Kent Arch. Res. Groups Council). Philp, B. and Philp, E., 1974, 'A Romano-British Site near Halfway House, Barham', Kent Arch. Rev. 37, 206-210. Rankov, N.B., 1982, 'Roman Britain in I 981: Kent, Great Mongeham', Britannia, XIII, 395. Rigby, V. and Freestone, I., 1995, '5. Early Gaulish and Rhenish Imports, 5.2 Discussion', in Elder, J. (ed.) Excavations in the Marlowe Car Park and Surrounding Areas (The Archaeology of Canterbury Vol. V), 640- 652. Stead, J.M. and Rigby, V., 1989, Verulamium: the King Harry Lane Site (English Heritage Arch. Rep. 12). Stebbing, W.P.D., 1939, Note, Archaeologia Cantiana, LI, 205-6. 151 KEITH PARFITT Thompson, I., 1982, Grog-tempered 'Belgic' Pottery of South-eastern England (B.A.R. 108). Young, C.J., 1977, Oxfordshire Roman Pottery (B.A.R. 43). 152