The Roman Archaeology of the Isle of Thanet

THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET D. R. J. PERKINS When the Thanet Archaeological Society, then 'Thanet Archaeological Unit', came into being in 1976, very little was known about settlement and activity on the Island during the Roman period. A number of discoveries had been made during the northern and southern urban extensions to Ramsgate in the 1870s (Hicks 1878). Thereafter, Roman remains were occasionally encountered and indifferently recorded, perhaps the best known being the villa site at Margate discovered and sampled by Dr Arthur Rowe in 1923 . 1 This paucity of properly investigated and published sites created a general impression among archaeologists that Thanet had been sparsely populated during the Roman period, merely a granary for the shore forts at Richborough and Reculver. The latter proposition has now been overthrown, largely as a result of a quarter century of fieldwork by the Thanet Archaeological Society, particularly in the field of aerial survey, and the creation of a Thanet Sites and Monuments Register. Added to this has been the implementation of PPG 16, bringing contract archaeology to Thanet with the activities of the Trust for Thanet Archaeology and others. A new picture of Roman Thanet is emerging, that of a populous island landscape liberally sprinkled with villas and farming settlements. The object of this paper is to summarise the evidence, presenting it in the form of a gazetteer of the eighty-three sites so far recorded (see Appendix 1). The pattern of settlement and activity in Roman Thanet now discernible is discussed below. SETTLEMENT AND ACTIVITY IN ROMAN THANET Villas and/or substantial buildings: possible religious sites To date twenty-three sites in this category have been identified, see Appendix and Fig. 1. Only four, numbers 1 - 4 have been investigated 43 22 • 58 29 10 38a Sb 59 • Margate 0 • * 64 78 ,J.2 6 460 71 0 43 * 61 3km 77 Broads1aixs • Major Building 0 Settlement site * CremBlion * !nhumation • Mixed Rite Fig. l Roman Thanet: Distribution of Sites by Type. Roads are shown by solid line (good evidence) and broken line (speculative). THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET to any degree, and only Site l at Abbey Farm has been fully excavated. The Abbey Farm evidence is still being processed, but at the time of writing it is safe to say that this was a large well appointed villa with out-buildings. Settlements and Occupation Deposit Sites Twenty-six sites are described in the Appendix and their distribution shown in Fig. 1. Of these, two, Sites 33 and 34, are in the inter-tidal zone, and another, Site 37, appears in eroding cliffs. Four sites - 29, 32, 37 and 38 - can be called large in that they extend over several hundred metres, presumably as single units. Others are of unknown extent, knowledge of them being limited to discoveries of occupation deposits. Within the broad description 'Settlement', Sites 26 and 30 may well be military, Site 49 was considered by its investigator to have been a pharos, and Site 40 consists of a ditched field system with a few features. While Site 29 has provided ample evidence of occupation, it may more properly belong in the next category. Industrial Sites Site 29 has yielded evidence of iron working over a large area, while at the same time constituting a rather inhospitable location for settlement (see discussion below). Cemeteries Twenty-two cemeteries have been recorded in Thanet as well as thirteen 'single finds' (Fig. l ). The latter cannot safely be described as 'single burials', as in most cases no effort was made at time of discovery to explore further. Of the twenty-two cemeteries, seven appear to contain cremations only, ten have only inhumations, and five are mixed rite. Where reliable dating evidence exists, the funeral rites follow the expected chronological sequence, with cremation being carried out through the first and second centuries, giving way in the late second century to mixed rite, and thereafter inhumation burial grounds. The distribution of the cemeteries, especially where they may indicate or confirm the presence of a Roman road is dealt with below and in discussion. Roads Site 83 consists of a short length of Roman road, apparently aligned on Site 22, the Belgic/Roman complex at Westbrook, Margate. The obvious presence of another ?prehistoric - Roman road, Dunstrete is dealt with in discussion. 45 D.R. J. PERKINS DISCUSSION In attempting reconstructs of various aspects of Roman Thanet, difficulties arise from the quality of the data. Firstly, for various reasons, there are no dates for sixty-two of the eighty-three sites listed. Of the remaining twenty-one, some dating information in the form of coin and/or ceramic evidence exists (see Table 1) but the datable material may not be typical of the whole site. Only the date bracket for the Abbey Farm villa (Site 1) which has been subject to total excavation can be accepted. The dates given in Table 1, whatever their reliability, seem to indicate occupation peaking in the second century. TABLE l. LIST OF SITES WITH DATABLE COIN AND/OR CERAMIC FINDS First/second-century Second-century Second/third-century Third-century Third/fourth-century Fourth-century Sites: 4, 11, 27, 35, 36, 71 Sites: 3, 8, 24, 28, 53, 63, 64, 65 Site: I Sites: 66, 67, 70, 80 Site: 68 Site: 81 A second problem is that the degree of investigation varies greatly from site to site. Villas and other buildings, whether recorded as structures or explicit cropmarks, present no difficulty. Evidence for 'settlement sites' however, can range from extensive groups of remains, to chance encounters with occupation deposits, perhaps within a single feature. While bearing in mind the strictures imposed by the foregoing, the writer herein adopts a heuristic approach, attempting the construction of a paradigm that may have to be drastically revised as the passing years bring further discoveries. Roads It seems to be accepted that a 'Harbour Road' extended from the Watling Street terminus at Canterbury to the Wantsum Channel shore at Upstreet (Fig. 1 ). In the controversy over the course of the road connecting Richborough to the main Roman road system of Britain (Panton 1994) one putative route (Knox 1941) has it joining the 46 THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET Harbour Road at Upstreet, on the narrows of the Wantsum just opposite Sarre in Thanet, the natural ferry point to the Island. While no trace of metalling has ever come to light, there can be little doubt that ancient Dunstrete (now in part the A253) running from Sarre via lord of the Manor to Ramsgate, was a Roman road, or more probably, a prehistoric track adopted by the Romans. Evidence for this lies in the distribution along its length of cemeteries and occupation deposit sites (Fig. I). The postulated extension running in to the Ramsgate valley is shown in broken line. Similar evidence, though more attenuated, can be drawn upon to postulate a road joining the cluster of sites north and south of the Ramsgate valley with those at Margate. A length of road plotted at Hartsdown, Margate (Site 83) is so aligned as to be part of such a road, serving the Ti voli villa (Site 2) and a settlement/temple site at Westbrook (Site 22). The linear distribution of sites in some other areas is also suggestive of roads or tracks. One such appears to commence on the Ebbsfleet peninsular, a Prehistoric and Dark Age entrepot for Thanet in the mouth of the Wantsum. From the Ebbsfleet villa and settlement (Sites 3 and 4 7) a line of sites curves north-east parallel to the shoreline, then east towards the Ramsgate site cluster (see broken line in Fig. 1). Another possibility, less well defined, is a road from Sarre to Margate, hugging the coastline from Sarre to Brooksend (Site 9) and then following the ridge of the escarpment north of the Acol and Shottendane valleys. Also suggestive is the linear grouping of major buildings running from Site 16 on Dunstrete north-east to the Margate area (see below). Had cemeteries been discovered along this line it would be very convincing as a road. Recent work by the writer has made a case for Thanet as an important 'Gateway Community' (Hirth 1978) during much of the last two millennia of prehistory.2 Such communities depended for their being on inter-regional control of trade and communication. In the case of Thanet, an island commanding the passage of the vital Wantsum Channel sea-way, internal communication would also have been important. The Romans may well have inherited a complex system of tracks linking sheltered natural harbours at Ebbsfleet and Sarre with extensive Bronze Age and Early Iron Age settlements on the higher ground, in particular the oppida at Dumpton and North Foreland.3 Villas and major buildings In examining the distribution of villas and other major buildings it has been thought profitable to examine the spatial relationships as 47 D.R. J. PERKINS revealed by Thiessen (or Voronoi) Polygons, but with an illustrative range set around each building at a distance of 1km (Fig. 2). The truncated edges of each polygon mark the line that is equidistant between each building. Unrestricted edges serve to show that there is nothing known in that direction for at least 1km. The lack of dating evidence means that we can have no idea how many buildings listed in the Appendix coexisted at any particular period. However, the use of space throughout the island during the Roman period is illustrated. Of particular interest is the relatively regular spacing between major buildings in distinct spatial groups, usually about 800m, or 1.2km, or 1.6km. For this see groups composed of Sites 11, 16, 10; and 1, 13, 3 and 18. A linear relationship emerges between Sites 16, 19-21 (probably the same complex) I 2, 17, and 2. Is this the line of a road linking Sarre with Margate? If not, and if the buildings did not exist at the same time, what mechanism dictated the regular spacing? The more Figs l and 2 are examined, the more possibilities spring to light. It is all too easy, however, to assemble 'ley-line' constructs from fortuitous relationships. Another aspect demonstrated by Fig. 2 is the density nearby of List 2 sites in relation to the List 1 buildings (see Appendix 1). Where the List I and 2 sites are very close, within say I 00m, such densities could be indicative of the relative importance of buildings. For the most part though, the presence of List 2 sites within the polygons should be assumed as accidental. A valuable function of Fig. 2 is that of drawing attention to gaps on the plan where no villas or major buildings have been discovered. The most significant of these covers all of Thanet east of a line from Cliftonville south to Ramsgate. It is hardly likely that this large area held no buildings, and remains may have been destroyed or gone unremarked during the Victorian expansion of Ramsgate and Broadstairs. Minor clues exist. For example, during the excavation of an Iron Age site at North Foreland, Broadstairs in 1999 fragments of flue and roof tile were found.4 Foundations of a 'huge wall' were dis covered nearby in 1876 (Dowker 1878). Another gap appears in the Manston area, possibly the result of afforestation, and those at Sarre and Birchington may eventually be filled by discovery. An interesting pattern can be seen in Fig. 1 on the southern flank of the island between Cliffsend and Monkton. Is this the relationship of villa sites to the spring line on the down-slope of the Thanet Beds? Do they relate directly to the cemeteries on the chalk downland escarpment above and to the north of them. Perhaps the same topography to the west holds similar groupings. From the little evidence so far to hand, the Thanet villas can be seen 48 A 0 Fig. 2 Roman Thanet: Postulated Settlement Patterns. 0 40 0 82 * 61 3km a 41 • 0 * * 77 Broadstairs Major Building Settlement sile Cmnatlon Inbumaliou Mixed Rite D.R.J.PERKJNS to vary considerably in size and status. At the lowest end of the scale, small and constructed of inferior materials, are the villas at Ebbsfleet (Site 3) and Lord of the Manor (Site 4). From what was observed of those sites, the villas at Tivoli (Site 2) and Drapers Mills School (Site 6) appear to have been larger and much better appointed buildings, possibly on a par with the ideally situated and rather grand Abbey Farm villa (Site I). Of the remaining List 1 sites little can be said that does not stray into the realm of guesswork, although it is worth mentioning that from cropmark evidence alone, the building complex at Plumstone Farm (Site 16) is of impressive size. Settlement and occupation deposit sites Only a few of the List 2 sites have been confirmed by the nature of their discovery to have been settlements of some extent, and only one of them (Site 30) has been excavated beyond the level of sampling. The rest can only be described as occupation sites, as the limited evidence gives no indication as to size and status. To further frustrate interpretation, there are no notes or finds for most of the sites and so no dating evidence. Despite the foregoing, the distribution shown in Fig. I can be informative in some ways. It is seen that just over 30 per cent of all the sites are situated on, or close to, the modern shoreline. While it must be borne in mind that a coastal strip up to 600m wide has been lost by cliff erosion since the first century, a preference for settlement near the shore seems well established. Advantages to be gained from this would include food gathering on the beaches and wave-cut shelf, communication, either by boat or using the shore as a road, and security. Any human threat posed to the islanders would come from the sea, and living near or on the shore would greatly reduce the risk of surprise. Other than those sites closely strung on the line of a road or track, a tight cluster of three sites can be seen near Stone Gap, Broadstairs (Sites 24, 63, and 70). ls this a village, or nucleated settlement centred on an undiscovered major building? Cemeteries It can be safely assumed that each of these indicates the presence at reasonably close proximity of an occupation site. Where Fig. l demonstrates no such presence, there is scope for future fieldwork. The most significant evidence in terms of size afforded by any of the burial grounds is that of the Mount Pleasant cemetery at Minster (Site 58). Wherever sampled over an area of up to six hectares, it has revealed a high density of burials. The number of graves that can be 50 THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET extrapolated from this is far greater than what might be expected from the nearby Abbey Farm villa (Site 1), and more indicative of a Roman Minster in the form of an extensive shoreline settlement. Another cluster of four burial sites is on the high ground west of the Ramsgate valley (Sites 31, 52, 60 and 62). It is possible that they are elements of an extensive linear cemetery aligned on a road, perhaps the end of Dunstrete. The convergence of postulated roads on Ramsgate, with the distribution of cemetery and settlement remains is indicative of a high density of settlement. A local focus of importance? Industry The only evidence for this in Thanet is Site 29, an apparently extensive iron working site at Mount Pleasant, Minster. The location of the site, on the escarpment ridge and exposed to the prevailing westerly winds, would be ideal for furnaces blown by natural draught, although this technology is thought not to have been employed in Britain (Cleere 1976). Thanet is not an obvious venue for ironworking activity, as the nearest sources of ore would have been far off in the Weald. What the island may have possessed though, was timber for fuel, manpower, and a nearby centre of demand for ironwork, the fort and naval anchorage at Richborough. SUMMARY AND THE FUTURE POTENTIAL FOR RESEARCH Generally, the foregoing shows Roman Thanet as an area with a high potential for study, both in terms of fieldwork, and in the location and reappraisal of material surviving in local and more distant collections. Examination and careful consideration of the distribution shown in Fig. 1 should generate many questions and suggest programmes of fieldwalking and aerial survey. Particular objectives could include a search for apparently 'missing' villas in the eastern quarter of the island, north-west in the Birchington area, and on the Thanet Beds spring-line towards Monkton. The existing data also provides scope for purely academic work - for example, on the 'ritual' well shafts discovered around the island. Best known of these belongs to Site 27 at Rams gate (Hillier 1889) but in 1938 square shafts containing, human, equine, and bovine remains were exposed by cliff erosion at Birchington (Site 27, Worsfold 1946) and another in Sea View Avenue was excavated by British Museum staff in 194 7. These shafts may be compared with seventy or so distributed throughout Britain (Ross 1968). Similar shafts, trun- 51 D.R. J. PERKINS cated by cliff erosion and wave action can be seen exposed in the wave-cut shelf at Minnis Bay (Site 34) and Epple Bay, so that fieldwork, albeit wet and muddy, is also possible. The writer hopes that this paper will lay a foundation for research in Thanet during the opening century, especially by amateurs, whose activity remains the backbone of archaeology. ACKNOWLEDGEMENTS Thanks are due to Ges Moody for his comments, preparing the G.I.S. database, and for generating the distribution maps (Figures I and 2). Also to the late Len Jay, Chairman of Thanet Archaeological Society, whose notes collected over many years contributed in no small way to the lists below. APPENDIX I GAZETTEER OF RECORDED ROMAN SITES IN THANET The Gazetteer is divided into List 1 (Villas and/or other substantial buildings) and List 2 (all other settlement/occupation sites). Other than full or partial excavation, the criteria for the identification of List 1 sites here applied are: a classic villa type cropmark, and/or abundant evidence of a substantial building with villa attributes such as roof and flue tiles, pilae, tesserae, painted wall plaster, etc. Numbers in brackets are Thanet Sites and Monuments Register files (TSMR). Many will have Kent SMR counterparts. Six figure numbers are OS grid square co-ordinates. Most of the sites in this gazetteer are unpublished. T.A.S.A. = Thanet Archaeological Society Archive Notes. List I: Villas and/or other Substantial Buildings Abbey Farm (316) TR 315646. Large winged villa with external bath house. Excavated 1996- 1999. Substantial remains nearby (Building 4) may represent another villa or a temple. 2 Tivoli, Margate (7) TR 352702. Part excavated 1923 by Dr Arthur Rowe. Walls, frescos, tessellated floors, hypocaust. Developed, part may have survived. Iron Age - Belgic remains, pottery kiln. 3 Ebbs/Leet (3 I 4) TR 334634. Sampled by trenching during evaluation. Wall foundations and mortar floors, coins, ceramics. 4 Lord of the Manor ( 160) TR 359653. A rescue excavation ahead of urban development in 1982 revealed wall foundations, rammed chalk floors, and a cellar lined with flints in mortar. Dated by ceramic and coin evidence to the mid-first century. 52 THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET 5 Potten Street, St Nicholas (582) TR 261674. Walls and wall foundations with an area of pebble floor were found during a highways watching brief (Thanet Way Improvement Scheme 1991). Extensive infrastructure with pits and ditches extending out 200m to the south-east. 6 Drapers Mills School, Margate (49) TR 363699. Discovered in the 1930s when the school was being built. Research sampling in the l 960s and 1980. Tessellated floor, walls, and building debris recorded. Much of the site will have been destroyed during the building of the school and houses. 7 St Mildred's Bay, Westgate (263) TR 377704. Tessellated floors were found in the 1870s when they were destroyed by hotel construction. A few midden pits remained under the hotel gardens until a second phase of development in the I 980s. About IO0m north of the site in the inter-tidal zone there is a feature that appears to be the remains of a revetment forming a quay on the bank of the ancient creek bed. 8 Belle Isle ( 126) TR 253667. A scatter of building debris and pottery throughout an area of about Iha was fieldwalked in 1979. Materials included roof tile and pilae fragments, painted wall plaster, and tes• serae. A pot containing a silver coin hoard (Marcus Aurelius) was found nearby in 1630 (Lewis 1723). 9 Brooksend (68) TR 286683. Trenching in l 980 next to an explicit small winged villa cropmark revealed boundary ditches with building materials and artefacts. 10 Way Hill, Minster (245) TR 324655. Wall foundations amid an extensive system of ditches and pits were sectioned by a pipeline in 1985. Iron-working features were found nearby. 11 Seamark Hill, Monkton (221) TR 286659. Rectilinear cropmark. A pit about 50m away was sectioned during pipeline work in 1987, it yielded coins and building materials c. AD 75 - 200 (Perkins 1989). 12 Woodchurch, Birchington (I 16 515) TR 323678. A rectilinear cropmark interpreted as a large villa in a ditched enclosure. An attempt was made to protect this site as an Ancient Monument (SAM 365) but only about 8 per cent of the cropmark falls within the scheduled area. Fieldwalking over the site in 1999 - 2000 produced building materials, sherds, and Roman small finds (Perkins 2000). 13 Grinsel Hill, Minster (310) TR 324647. The flat top of a low hill near the ancient shoreline has a dense scatter of building debris and pot sherds, coins, etc. Prehistoric settlement remains beneath. 14 Lydden, Ramsgate (a) (71) TR 353675. Rectilinear cropmarks and many metal detector finds. IS Lydden, Ramsgate (b) (556) TR 359675. Rectilinear cropmarks appear to show a multi-roomed building; many metal detector finds. Situated about 250m from Lydden (a). 16 Plumstone Farm, Monkton (579) TR 301658. Extensive rectilinear crop- 53 D. R. J. PERKINS mark complex appears to show a large multi-roomed building or several buildings. A family who farmed the land in the mid twentieth century remember massive foundations being hit by the plough, and an 'authoritative archaeologist' of the day identifying the site as a 'large Roman villa'.5 An adjacent Roman settlement (100m south) was excavated in 1994 (see 30). 17 Garlinge, Margate (303) TR 339693. A rectilinear cropmark appears to represent a winged villa; tile fragments have been found on the margins of the field. Speculative Sites; possible villas 18 Cottington Lane, Cliffsend (255) TR 338641. Settlement remains found in the 1970s; waterlogged deposits yielded samian and other fine wares, and building materials. (Macpherson-Grant pers. comm. 1976) 19 Acol, Birchington (113) TR 314675. A complex rectilinear cropmark appears to represent a building and out-buildings within a doubleditched rectangular enclosure. (It could be a medieval manor site, since the cropmark is reminiscent of that produced by the remains of Upper Court, Mansion, Ramsgate. There is however, no record or oral tradition of a lost manor site at Acol.) 20 Somali Farm, Birchington (a - 465) (b - 466) and (c - 467) all within square TR 3268. Three ill-defined cropmarks that are nevertheless suggestive of villa remains. Some trenching work was carried out on one cropmark by Quex Park Museum staff in the 1960s with negative results. Possible Shrine or Temple Sites 21 Sparrow Castle, Birchington (115) TR 319657. A cropmark shows a large rectangular enclosure with causeway entrance at one end, and an elongated rectangular (?sanctuary) building at the other. The enclosure ditch was sectioned by a pipeline in 1988, and Belgic and Roman sherds were found. This site is roughly equidistant between possible villa sites 12 and 18. 22 Westbrook, Margate (2 - 229) TR 335706. This site is under the clifftop lawns, with pits containing midden materials eroding from the cliff-face onto the promenade car park. Remains were found in 1931 during excavations that formed the 'Sunken Garden', and sample trenching was carried out in the 1980s. Wall foundations and ditches were recorded with finds of tesserae and tile. Belgic and Roman ceramics included small 'votive' cups. The spread of the site is large enough to represent several buildings, a villa and temple? 23 Abbey Farm, Minster, Building 4 (316) TR 31446456. At the time of writing, part-excavated 'Building 4' at the Abbey Farm villa complex seems likely, from what has so far been recorded of its layout, to have been a shrine. 54 THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET List 2: Settlement Remains, Cemeteries, and Miscellaneous Sites 24 Stone house School, Broadstairs (62) TR 396691. Various small-scale rescue excavation between 1907 and 1971 indicated a large settlement site, Belgic - second-century Roman. 25 Seven Stones Estate, Broadstairs (606) TR 39256635. Roman settlement remains were recorded during building development in the 1960s (Minter and Herbert I 973). 26 'Cheesemans Camp', Mans ton (132) TR 324668. An earthwork consisting of a rectangular double-ditched enclosure with opposed entrance causeways was \eveled to make a pony paddock in 1988. It was generally thought to be a medieval moated enclosure, but the potsherd surface scatter and sherds from ditch fills were Roman, and a cremation burial was found in the bank between the ditches (T.A.S.A.). 27 Junction of South Eastern and Park Roads, Ramsgate ( 183) TR 375644. A settlement site with pits and ditches was found here in 1889, also a well 115ft (35m) deep. Occupation period was Be\gic to second century (Hillier 1889). 28 Whitehall Road, Ramsgate (616) TR 37356600. Settlement remains discovered in 1993 (T.A.S.A.). 29 Mount Pleasant, Minster, both sides of the A253 ( 152 ) TR 3 I 2658. An extensive Roman settlement and iron working site was found during defence work on Manston Airfield in 1944 (Grimes unpublished). A pipeline cut made just south of the adjoining road in 1984 revealed Roman features that may have been bloomeries (Perkins 1985). 30 North of the A253, Minster (258) TR 303659. Part of what appears to be an extensive area of settlement was excavated ahead of road construction in 1994. Features included a number of sunken-floored buildings (Archaeologia Cantiana, cxvi, 1996); see also Site 16. Inhumation burials were found a few metres to the south in the l 960s.(T.A.S.A.). 31 St. Mildred' s Road, Rams gate ( 181) TR 373646. Building work in the area in about 1870 revealed cremation burials, but also many pits with midden materials (Hicks I 878). 32 Clifftop near 'Hug in', Cliffsend, Ramsgate (203) TR 35306440. Settlement remains throughout an area of at least Iha. Large pits have inhumation burials in midden layers. Many metal detector finds (T.A.S.A.; Hearne et al. 1995). 33 Minnis Bay (west), Birchington (223) TR 273694. Inter-tidal zone site, pits with potsherds (T.A.S.A.). 34 Minnis Bay ( east) Birchington ( 19) TR 285697. Inter-tidal zone site, pits in the chalk of the wave-cut shelf contain potsherds, woodwork, and well preserved organics (T.A.S.A.). 55 D.R. J. PERKINS 35 Shuart Farm, St Nicholas (257) TR 296677. Settlement remains found included a rammed chalk floor and wall foundation, with first/ second-century ceramics, and a small millstone (T.A.S.A.). 36 Monkton Roundabout, A253 (144-145) TR 285655. Settlement remains found in 1982 included an enclosure ditch, pits yielding much ceramic material (second half of first century) and one grave (Perkins 1984). 37 Clifftops between Minnis and Grenham Bays, Birchington (202,225,226) TR 288700. Settlement(?) features sectioned by the chalk cliffs include well shafts and pits or ditches. 38 Beside the A253, Minster (243 - 247) TR 324656 - 326656. Settlement remains consisting of pits and ditches were encountered for over 400m during a pipeline watching brief in 1984 (Perkins 1985). Similar remains had been found 1 00m north on the airfield in I 971 (Macpherson-Grant, pers. comm.). (See also 10 in List I.) 39 Fitzroy A venue, Kingstgate (15) TR 391706. Settlement remains found in 1912 (T.A.S.A.). 40 East Northdown, Margate (13, 14) TR 382705. Field system and features investigated in 1984 (Smith I 987). 41 Foreness Point, Margate (366) TR 385714. Settlement remains excavated 1977-78 (Wilson 1977). 42 College R oad, Margate (9) TR 36370 I. Settlement remains found in 1925 (T.A.S.A.). 43 Sacketts Hill, St. Peters, Broadstairs (198) TR 370691. Settlement remains discovered and sampled during roadworks in 1985 (T.A.S.A.). 44 Fleet Court Farm, Manston (273) TR 344672. Pipeline wayleave exposed a large scatter of sherds and building material fragments in 1987 (Perkins 1989). 45 Sprattling Court Farm, Mansion (135) TR 354665. Occupation deposits exposed by deep ploughing. Trenching revealed an occupation layer rich in potsherds and midden debris (T.A.S.A.). 46 Stonar Hill, Ramsgate (587) TR 3566. Ploughing on the hilltop in I 985 brought potsherds and building materials to the surface (T.A.S.A.). 47 Ebbsfleet Farm, Minster (311) TR 332632. Occupation of the hilltop is indicated by a scatter of potsherds and many finds of coins and bronzes (T.A.S.A.). The site, which has been subject to extreme plough attrition, is 300m south-west of a villa site (3 in List I). 48 Chester Road, Westgate (525) TR 32726993. Settlement remains recorded as being discovered in 1950 (T.A.S.A.). 49 Boxlees Hill, Minster (I 85) TR 309631. Cremations and foundations were uncovered during construction work in the 1960s and examined by J • Coy (T.A.S.A.). In Roman times this site would have been on an island in the Wantsum sea channel, about 900m south of the Roman 56 THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANET shoreline at Minster, and 3.3km north of the Richborough shore fort. The foundations are possibly those of a pharos. 50 Junction of Station Approach and Margate Roads, Ramsgate (165) TR 377655. Cremation cemetery (Journ. Brit. Archaeo. Assn., 1847, II, 281). 51 Ozengell, Ramsgate (161) TR 360655. Cremation burials eroded into the railway cutting in 1972 (T.A.S.A.) and a Roman lead coffin was discovered nearby in 1846 (Wright 1856). 52 Junction of Westcliff and Grange Roads, Ramsgate (183) TR 375644. Inhumation cemetery discovered in the 1870s. Saxon burials also present? (Hicks 1878, 17; Meaney 1964). 53 Twenties Brickfield, Shottendane, Margate (625) c. TR 344693. A second-century cremation cemetery was investigated by a Captain Mark Bullen sometime between 1860 - 1865. The urns are in the Margate Museum as part of the 'Rowe Bequest' (T.A.S.A.). 54 Little Cliffsend Farm, Ramsgate (175) TR 359644. A cremation burial found in 1876 (Cotton 1895). Monitored cliff erosion in the period 1976-1990 recorded Roman material from pits and a truncated foundation (T.A.S.A.). 55 Epple Bay, Birchington (24) TR 307696. Cremation burial found 1904 (Ant. Journ., IV, 2, 158). 56 Westgate Bay, Westgate (3) TR 320704. Inhumation cemetery discovered in 1922. 6 Graves truncated by erosion can be seen in the cliff-face, and pottery falls onto the promenade (T.A.S.A.). 57 Junction of Queens and Thanet Roads, Ramsgate ( 169) TR 389654. Two burials found in the l 870s (Hicks 1878). 58 Mount Pleasant, Minster, junction of Tothill Street with the A253 (189) TR 310655. Mixed rite cemetery covers several hectares. It has been sampled when encountered by gravediggers and during road works (T.A.S.A.). 59 Thorne Farm, Minster (187) TR 335654. A mixed rite cemetery just south of the A253 was discovered during pipeline work in 1984 (Perkins 1985). 60 Willson 's Road, Ramsgate (182) TR 375646. A single burial with a poppy head beaker and sherds (T.A.S.A.). 61 Rumfields Brickfield, Broadstairs (123) TR 378677. A cremation cemetery was discovered there in the 1900s (Minter & Herbert 1973). 62 Southwood, Ramsgate (178) TR 369674. A single inhumation with an amphora and other vessels including remains of a bronze bucket (Payne 1902). 63 Stone Road, Broadstairs (62) TR 396691. Two inhumations found in 1905. Dated from potter's stamps to the second century (Minter and Herbert 1973 & Biggs I 972). 57 D.R. J. PERKINS 64 Trinity Square, Margate (1) TR 355712. Mixed rite cemetery found in 1939, early second century (T.A.S.A.). 65 Shuart Farm, St Nicholas (104) TR 271657. Single cremation brought to the surface by the plough, second-century (T.A.S.A.). 66 St. John's, Margate (196) TR 356703. Mixed rite cemetery found in 1791. Coin dated to the third century (T.A.S.A.). 67 Minnis Bay, Birchington ( 194) TR 27086934. Single inhumation found in 1986, third-century ceramics (T.A.S.A.). 68 St. Lawrence College, Ramsgate (164) TR 376658. Inhumation cemetery found in the late nineteenth century, third/fourth-century coins (Kent, VCH, iii, 164). 69 Whitehall Road, Ramsgate (616, b) TR 37406595. Cremation burial discovered in 1993 (T.A.S.A.). 70 Lanthorne House, Broadstairs (102) TR 393689. Inhumation cemetery found in 1896. The skeletons were in six trenches 'huddled together, as though buried in great haste'. Coin dated to the third century (Payne 1895). 71 Avenue Gardens, Palm Bay, Margate (20 I) TR 372712. A first-century cremation cemetery was discovered during road construction in the (?)1920s (T.A.S.A.). 72 St George's Church, Rams gate (168) TR 381652. Gravediggers found a cremation in a B82 urn (T.A.S.A.). 73 Clifftop at Grenham Bay, Birchington (2, 20, 522). TR 29456995 - 29647005. Large quarry-type pits with Belgic-Roman materials found in 1947; inhumation burial found in 1961 (T.A.S.A.). 74 Alpha Road, Birchington (22) TR 301694. Inhumation cemetery discovered in the nineteenth century (Payne 1897). 75 The Retreat, Epple Bay Avenue, Birchington (23) TR 306698. Inhumation cemetery discovered sometime in the first half of the twentieth century (T.A.S.A.). 76 Ursuline Convent, Westgate (30) TR 316695. Roman burials found in the nineteenth century (O.S. Records). 77 Close to Bleak House, Broadstairs (261) TR 39956758. Inhumation cemetery found and part-excavated in the l 960s (T.A.S.A.). 78 Arthur Road, Cliftonville, Margate (615) TR 362712. Cremation burial inserted into the fill of a large pit backfilled as a midden. Found during building work in 1996 (T.A.S.A.). 79 Cliffsend, Ramsgate (376) TR 344642. Cremation assemblage comprised of urn and glass bottle contained in large amphora found c. 1870 (Scott-Robertson 1878). 58 THE ROMAN ARCHAEOLOGY OF THE ISLE OF THANE'I' Finds and features of interest 80 Pysons Road, Ramsgate (193) TR 37306695. A hoard of mostly thirdcentury coins found in 1970 (Cullen, Merrifield 1970). 81 Main Slipway, Ramsgate Harbour (184) TR 386646. Construction work in 1838 uncovered remains of Rams gate's medieval pier. Beneath this, sunk in the chalk bedrock, were found fragments of wooden piles and brass coins of Constantine.7 82 Northdown Hill, Margate (59) TR 381696. During brickearth extraction in 1876 a vertical shaft was found cut into the chalk. A gallery leading from this contained a quern, Roman sherds, midden material, and a second-century coin (Dowker 1878). In 1982 the writer and friends entered another nearby larger and deeper man-made cavern. Iron nails of Roman type were recovered, and a Chi-Rho monogram was observed carved on a smooth area of chalk (T.A.S.A.). 83 Hartsdown Road, Margate (513) TR 34866990. Evaluation trenching in 1995 and rescue excavation in 2000 exposed a length of metalled road and sectioned it in five places (Perkins 1996; 2000; T.A.S.A.). NOTES 1 Dr Rowe's notes and Ant. Journ., 5 (I 925), 164. Journ. Roman Studies, 14 (1924), 240. Kent Messenger, 30 May 1925. 2 Thesis submitted by the writer for the award of Ph.D from the Institute of Archaeology, University College London, November 2000. Title: A Gateway Island. An exploration of the evidence for a cultural focus in the form of a 'Gateway Community' in Thanet during the Bronze Age and Early and Middle Iron Ages. 3 Since 1994 the Trust for Thanet Archaeology has carried out excavations on two large multi-phase sites, at Dumpton and North Foreland, Broadstairs. At both sites Bronze Age occupation had been followed by the establishment of Iron Age oppida developing as hil!forts or promontory forts. 4 See note 3. 5 Informant a Mr Goodhew of Minster in Thanet. When he was a boy his father had Plumstone Farm. They encountered foundations during ploughing and investigated. A partner of Goodhew Snr (described as an authoritative archaeologist of the day) said that they represented remains of a large Roman villa. 6 Dr Arthur Rowe's notes and correspondence with Howard Hurd, Local Studies collection at Margate Public Library. 1 Arch. Joum., 1845, vol. I, 69. Numismatic Chronicle, 1847. REFERENCES Biggs, H., 1972, Kent Arch. Rev., 273. Canterbury & Thanet Trusts (anon.), I 996, 'Interim report on excavations in advance of dualling the A253', Archaeologia Cantiana, 116, 305. 59 D. R. J. PERKINS Cotton, C., 1895, The History and Antiquities of the Church and Parish of St. Laurence, 262. Cleere, H., 1976, 'lronmaking', in Roman Crafts, Strong, D. and Brown, D. (eds), Duckworth, London, 127. Cullen, P., 1970, Kent. Arch. Rev. (Spring) 23. Dowker, G., 1878, 'Abstract of Proceedings', Archaeologia Cantiana, 12, xxxiv. Grimes, W. (unpublished but ?pending), Excavations on Defence Sites 1939-45, vol. 2, English Heritage. Hearne, C., Perkins, D., and Andrews, P., 1995, 'The Sandwich Bay wastewater treatment scheme archaeological project, 1992-94', Archaeologia Cantiana, 115, 265. Hicks, R., 1878, 'Roman Remains found at Ramsgate', Archaeologia Cantiana, 12, 14. Hillier, J., 1889, 'A British Village at Ramsgate', Archaeologia Cantiana, 18, I. Hirth, K. G., 1978, 'Interregional Trade and the formation of prehistoric gateway communities', Am. Ant., 43. Knox, C., 1941, 'St. Margaret's Bay and the Roman roads from Richborough to Dover and Canterbury, Arch. Cant., 54, 38 and fig.6. Lewis, J., 1736, History ofThanet, 2nd ed., London, 27. Meaney, A., 1964, Gazetteer of Early Anglo-Saxon Burial Sites, 132. Merrifield, R., 1970, Kent Arch. Rev. (Autumn), 2. Minter, A. and Herbert, B. (eds), 1973, Archaeological Discoveries in Broadstairs and St. Peters, 9. Payne, G., 1895, 'Researches and Discoveries in Kent', Arch. Cant., 21, xlvii. Payne, G., 1897, 'Researches and Discoveries in Kent', Arch. Cant., 22, !xii. Payne, G., 1902, 'Researches and Discoveries in Kent', Arch. Cant., 25, lxvii. Perkins, D., 1984, 'The Thanet Gas Pipeline, Phases I and II', Archaeologia Cantiana, 101, 86. Perkins, D., 1985, 'The Monkton Gas Pipeline', Archaeologia Cantiana, 102, 43. Perkins, D., 1989, 'The Selling - Thanet Water Main, 1987', Archaeologia Cantiana, 107, 276. Perkins, D., 1996, 'Hartsdown Community Woodland Scheme', Archaeologia Cantiana, 116, 268. Perkins, D., 2000, 'Roman finds at site of medieval village: Woodchurch, Birchington ', Archaeologia Cantiana, 120, 379-381; 'Roman Road metalling and Prehistoric Sherds: Hartsdown, Margate', ibid., 378-9. Ross, A., 1968, 'Shafts, pits, wells - sanctuaries of the Belgic Britons?', in Coles, J. and Simpson, D. (eds), Studies in Ancient Europe - Essays presented to Stuart Piggott, 255-85. Scott-Robertson, W ., 1878, 'Archaeological Notes on Thanet', Archaeologia Cantiana, 12, 331. Smith, G. H., 1987, 'A beaker(?) burial monument ..... from East Northdown, Margate', Archaeologia Cantiana, 104, 237. Wilson, J., 1977, Kent Arch. Rev., 49. Worsfold, F. H., 1946, Ramsgate Public Library Archives 2/4. Typed notes and parish magazine article. Wright, T., 1856, 'Celt, Roman, and Saxon', Coll. Ant. 60