The Stourmouth-Adisham Water-main Trench

THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH J. D. OGILVIE, F.S.A. With Contributions by G. C. Dunning, B.Sc., D.Lit., F.S.A.; I. F. Smith, B.A., Ph.D., F.S.A.; T. C. Champion, M.A., D.Phil.; D. J. R. Ogilvie, M.A., and others In the summer of 1960, the Thanet Water Board contracted with Messrs. Whittaker Ellis and Co. to lay a 24-in. water-main from Stourmouth to their pumping station at Adisham. This necessitated the digging of a trench, 3 ft. wide and from 4-12 ft. deep, for a distance of four miles. The work lasted without interruption for a year, during which time close observation was maintained on the trench, as it soon appeared that it would be traversing land of considerable archaeological importance. The route of the trench is shown on the map (Fig. I). The conjectural maps, Fig. 2 of north-east Kent, and Fig. 3 covering the same area as Fig. l, show a coast-line approximating to that of the Romano-British period, though it must have been much the same before then - certainly from the Neolithic period onwards. During the Romano-British period - and since - there were undoubtedly alterations in the relative levels of land and sea, but the distribution of prehistoric and Romano-British sites, of which those shown on Fig. 3 form a small portion, and the absence of sites below the 'coast-line' confirm that its position is substantially correct. This line is unifonnly 12} ft. above O.D. It can be derived from the O.S. maps by observing the limits of the artificially drained marshes. This is well shown on the Agricultural Survey of the Kent County Planning Department. 1 Since George Dowker published his version in 18722 variations have appeared regularly in Arch. Cant. 1 Abercrombie, East Kent Regional Survey, 1925. 2Dowker, Arch. Cant., viii (1872), 14. 91 J. D. OGILVIE / LINE OF TRENCH ONE MILE Mill Lane ) ELMSTONE Longmete Road HOAOEN SANDW.CH ➔ OUALOCK STREAM Fig. I. Site of Trench (Based upon the Ordnance Survey Map with the sanction of the Controller of Her Majesty's Stationery Office. Crown Copyright reserved). 92 KENT/ THAMES Fig. 2. North- east Kent. Pre-Ro man Coast-line. J. D. OGILVIE :/·3 L2 LINE OF TRENCH, SHOWING 'LENGTHS' AND SITES @ 􀀼1\-iffLISHED PRE·t.lEDIEVAL Y SURFACE FUNT SITES ONE MILE Fi g. 3. Site of Trench in Relation to known archaeolo gical Sites (Based upon the Ordnance Survey Map with the sanction of the Controller of Her MaJesty's Stationery Office. Crown Copyright reserved). 94 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH ESTABLISHED PRE-MEDIEVAL SITES These are shown by the circled numbers on Fig. 3 and are listed below. Neolithic 1. The Wingham pit. Arch. Cant., lxxiv (1960), 58-72. 2. Large roughed-out flint axe-head; surface find at Briton Farm. Not yet published. Bronze Age 3. Stodmarsh Um. Arch. Cant., xliii (1931), 296-7. 4. Small hoard of bronzes at Hoaden. Not yet published. Pre-Belgic Iron Age 5. Goodnestone Hearth. Unpublished, but noted in Kentish Gazette, 24th February, 1961. 6. Enclosure and ring ditch. Rookery Farm. N.G.R. TR 253605. Colour slide in possession of Mr. J. Bradshaw, RAF A/P 5/7(70. Belgic 7. Wenderton Marsh. Pottery finds on Romano-Blitish site. Not yet published, but noted in Arch. Cant., lxxvi (1961), ltii. Romano-British 8. Roman Road. Canterbury-Thanet. 9. Preston Forstal pottery. Arch. Cant., xx (1893), 50, and local tradition. 10. Roman burials at Preston. Arch. Cant., xx (1893), 49. 11. Roman 'undertaker's kiln' at Preston. Arch. Cant., xii (1878), 47. 12. Wenderton Roman site. Arch. Cant., lxxvi (1961), liii. 13. Ickham Villa. Arch. Cant., xiv (1882), 139. 14. Wingham bath-house. Arch. Cant., xiv (1882), 134, and Arch. Cant., xv (1883), 351. 15. Twitham Farm samian bowl and amphora. Arch. Cant., xv (1883), 356. 16. Rubbish pit south of A257. Unpublished. 17. Second-century water-mill, Ickham. Excavated by Mr. J. Bradshaw, who has kindly let me see a pre-publication report by Mr. R. J. Spain. 95 J. D. OGILVIE Dark Age 18. Supperton Anglo-Saxon cemetery. Arch. Cant., xvii (1887), 6, and Arch. Cant., xxi (1895), Jvi. 19. Withersden Hall Anglo-Saxon cemetery. Arch. Cant., xv (1883), 356; Nenia Brit., pls. xi and xii; lnventorium Sepulchrale. 20. Wickhambreaux Jutish vessel. Arch. Cant., xxxix (1927), 36. Undated 21. Circular crop marks. Neavy Down. Univ. of Cambridge, Dept. of Aerial Photography, ADl 79. · DESCRIPTION OF TRENCH AND FINDS The land through which the trench passed is now basically chalk, except where this has been eroded and covered with clay, sand, gravel and brickearth. Stourmouth and Preston lie on a long low ridge, mainly of this brickearth, which juts out to the north into the marshes which now replace the Wantsum Channel. The southern end of this ridge terminates at Gobery Hill, where the chalk approaches the surface. South of Gobery Hill the chalk has been eroded to form the bed of the Durlock stream at Dambridge and the adjacent marsh. Beyond this the ground then rises gradually to Neavy Down, a prolongation of the chalk downs of Barham, themselves a spur of the North Downs. The sites and finds will be described from north to south as if the work had been carried out progressively in this direction. Although this was the contractor's general plan, variations were necessary depending on weather conditions, availability of materials, and technical factors. For convenience in recording finds, the trench was divided into lengths. These were primarily chosen for topographical reasons, but they appear also to coincide with the archaeological divisions. They are recorded on Fig. 3, LI to LI I. The description of the finds, the opinions thereon, and the corresponding illustrations are the work of Dr. T. C. Champion, except where otherwise stated. LENGTH I Stourmouth-Mill Lane, Preston (Site I) In this length the trench was dug to a depth of 12 ft. Brickearth, from 6 to 12 ft. thick, lay on gravel containing slightly rolled flints. The brickearth was uniform in texture and contained no flints. 96 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH INCHES Fig. 4. Neolithic Bowl from Site I (Scale: ¼). First examination showed nothing of archaeological interest. Careful re-examination, however, of the orchard of Oast Fann revealed a thin horizontal layer of pebble at a depth of 3 ft. extending for 30 yds. (Site 1). This was trowel-cleaned and at one point a band of darker soil, 3 ft. long, was found lying on the pebble layer. Examination of this band of soil, and of its extension into both sides of the trench, showed that the trench had cut eccentrically through a hearth, circular and of a shallow saucer shape, about 5 ft. in diameter and 9 in. deep. It was composed of fine black ash and charred wood. The soil of the hearth, both as recovered from the trench sides and from the spoil heap (where it lay practically undisturbed as deposited by the mechanical excavator), yielded worked flints and coarse pottery sherds. Site I. Dr. G. C. Dunning examined the finds and has already reported briefly on them. 3 From a rim of coarse dark grey ware with angular flint grits and some large particles he reconstructed a pot (Fig. 4) which he identified as a Neolithic carinated bowl (Form J). Dr. Isobel Smith reports on the finds as follows: Pottery There are some 60 sherds of plain Neolithic pottery of'Western' facies, all featureless wall sherds except four that join to form part of the rim and straight neck of a shouldered bowl (Fig. 4). As shown in the 3 Antiq. Journ., xlvi (1966), 13. 97 J. D. OGILVIE drawing, the rim has been rolled irregularly and at one end of the surviving part stands clear of the neck, but at the other is folded back against it; pressure along the inner surface has produced a marked facet. At least two ring-joints were visible in the section; a butt-joint at the shoulder and an overlapping one a short distance beneath the rim. The ware is grey-brown in colour and relatively hard. Neither surface has received more than a perfunctory wiping and both are rough and gritty with protruding particles of calcined flint up to 6 mm. in diameter. The straight neck of the existing sherds cannot readily be matched elsewhere, but in view of the generally crude finish of the bowl it may be suspected that this feature is accidental rather than intentional. The remaining sherds come from at least seven other vessels. These comprise two or three cups, a similar number of well-smoothed pots or bowls, and three or more pots with a rougher finish. Sherds from some of the latter are fairly thick (up to 13 mm.) and seem to come from rounded bases. Apart from a cup fragment that lacks any obvious filler, all contain particles of flint; carbonized organic material is also present in some sherds. Several pieces of rather soft dark clay contain reddened lumps of brickearth. Each piece has one fairly smooth surface, but none appears to have come from a pot and the substance is probably fire-hardened trampled mud in which are incorporated pellets of previously fired brickearth. Flint Industry The flints are all in fresh condition and no patina has developed since they were worked. Of the 44 pieces which retain patches of cortex, only one seems to have come directly from the chalk. The rest were evidently collected from gravel deposits, for seven have cortex that is somewhat abraded or discoloured, and 36, including both cores, have heavily rolled, black-stained natural surfaces, usually with an underlying ochreous band. This distinctive yellow streak is also present on a number of non-cortical flakes, but the axe fragment and a few pieces are made of flint from other sources. The 89 pieces (excluding chips and spalls) may be classified as follows: Leaf-shaped arrowhead (Fig. SA). Bifacial retouch, broken at tip, unusual subtriangular form. See also miscellaneous category below. Axe fragment (Fig. SB) from the cutting edge of a ground axe, the sides re-flaked to make a short pointed butt, presumably for rehafting. The flint is creamy grey in colour and one of the original round surfaces has a much higher lustre than the other. 98 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH A 􀀄􀀄􀀅 ;, :: - ,.:;.,i -􀀇-- 􀀈- .' ·--􀀄 B E ·--􀀆 G Fig. 5. Flints from Neolithic Hearth (Scales - A:½; B-H: ½). 99 J. D. OGILVIE Broken utilized flake (Fig. SC); this flake is made of a creamy grey flint like that of the axe fragment and may have come from the same tool. Five scrapers; in addition to the large, heavy specimen (Fig. 5D), there is a smaller one with an oblique working edge made on its bulbar end (Fig. SE); another three flakes have also some scraper retouch. Four serrated flakes. Fig. SF represents the largest of these; all bear traces of the narrow band of lustre that commonly appears along the working edge of such flakes and exhibit other signs of considerable wear. Forty-five utilized flakes; these are irregularly chipped along one or both edges (as in Fig. 5G); some also have narrow bands of lustre. In two instances the edge opposite the utilized one has been blunted by retouch, and in another four an edge is blunted, either intentionally or as a result of a particular kind of use, by the removal of fine, closely spaced squi!ls. Burin; a flake with one edge partially blunted has also at least one burin facet on its distal end. Miscellaneous retouched pieces; Fig. 5H illustrates a flake that has been struck from the edge of a bifacially retouched implement. At first sight, it appears to be an axe-sharpening flake and therefore out of place in the present context. On the other hand, it need not have come from a tranchet axe; a flake of this kind could be produced in the course of reworking a perfectly typical 'Western' Neolithic artifact - one of the relatively large bifacially retouched objects known as laurel leaves.4 A second (unfinished) leaf-shaped arrowhead may be represented by a small, thin flake with some invasive retouch on the bulbar surface. By-products; these comprise 24 apparently unused struck flakes, and two cores. One of the latter has been worked from three directions; the other, a small split nodule, has had a series of short parallel flakes removed round part of the circumference and may have been used as a core-scraper. There are some minor peculiarities in this assemblage - the atypical forms of the finished arrowhead and the oblique-ended scraper and the problematic origin of Fig. 5H, but at present there does not seem to be any reason for attributing particular significance to them, and the industry is otherwise a straightforward 'Western' Neolithic one of the kind that is well known from many sites in southern England. The affinity is confirmed by the presence of a relatively high proportion of narrow flakes similar to the one shown in Fig. SF. As will be seen from the table below, which gives the breadth-length5 ratios of the 50 4 PPS .• xxvi (1960), 223, 226. 5 Length was measured along the bulbar axis and breadth was taken as the widest part at right angles to this axis. 100 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH unbroken utilized, serrated and waste flakes, over a fifth are blade-like in shape, the breadth amounting to no more than two-fifths of the length, and in over a half the breadth is no more than three-fifths of the length. Ratio of breadth to length 1 :5-2:5 2:5-3:5 3 :5-4:5 4:5-5:5 Over 5 :5 Number 11 17 17 1 4 Per cent 22 34 34 2 8 The results may be compared with those obtained from 'Western' Neolithic assemblages in Devon, 6 Dorset,7 and Wiltshire. 8 LENGTH 2 Hardacre Farm (Sites 2-3) Immediately south of Mill Lane, in the orchard of Hardacre Farm, the gravel approached to within I ft. of the surface. Immediately below the topsoil were two small pits, each 2 ft. wide and 2 ft. deep (Sites 2 and 3). The Finds (described by T.C.C.) Site 2. Waste flint flakes and one scraper; 98 sherds of pottery, handmade and predominantly in coarse, flint-gritted fabrics. I. (Fig. 6) Hard fabric with abundant fine flint gritting, core orange to grey, exterior grey with black and orange patches; surface well smoothed on neck; profile incomplete, but one non-joining rimsherd shows plain rounded profile. Cf. Chinnor, Oxon.9 2. Hard fabric with much fine flint gritting; core and surface dark grey; interior and exterior well smoothed. 6S. H. M. Pollard, 'Neolithic and Dark Age Settlements on High Peak, Sidmouth, D􀂛von', Proc. Devon Arch. Exp/or. Soc., xxiii (1966), 35-9. 8 PPS., XXX (1964), 360. The results from Wi ndmill Hill (I. F. Smith, Windmi/1 Hill and Avebury: Excavations by Alexander Keeler, 1925-1939, (1965), Fig. 38) are comyarable if the serrated, utilized and waste flakes are taken together. Those from Knap Hil (Wilts. Arch. Mag., 60 (1965), 16) are not and the significance of this discrepancy is not as yet clear; it may, however, be 􀂜􀂝ted that the measurements include those of many flakes from upper layers in the 1\ches (ibid., 10) where Beaker sherds occurred. J, K. M. Richardson and A. Young, 'An Iron Age A site on the Chilterns'. Antiq. ourn., xxxi (1951), 132-48, e.g. Fig. 8, no. 52. 101 J. D. OGILVIE J I 5 ,,.. ..FF. , .. - 6 ' la􀀅1, . -.==-􀀄 --=--􀀉 =§:. -·--- -e. ----􀀇- ,..:,...-􀀄·􀀅- 􀀆-- · ·- 12 14 Fig. 6. Finds from Site 2 (nos. 1-6); Site 5 (nos. 7-11); Site 6 (nos. 12-14) (Scale:¼). 102 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH 3. Coarse fabric with much large flint gritting; core dark grey, surfaces dark brown to black; rim flattened. 4. Hard fabric with some fine flint gritting; core dark grey, exterior dark reddish grey to black. 5. Hard, with much fine flint grit; core and surfaces dark grey; exterior well smoothed. 6. Hard, with much fine and some large flint grit; core light buff, surfaces orange to brown. Site 3. One struck flint flake; 16 sherds of pottery, mostly of coarse, flint-gritted fabrics; one piece of tile and one fragment of glass. LENGTH 3 Shotfield Farm (Sites 4-13) The trench crossed the main arable field of Shotfield Farm. In this distance of 700 ft. eleven pits were revealed, varying from mere soil discolourations to the large well-defined pit (Site 6). In some cases excavations were made into the trench side to help to determine the nature of the pits. The whole section was surveyed and drawn (D.J.R.O.). Both of these tasks were rendered difficult by the flooding of the trench by heavy rain which caused collapse of the sides necessitating rapid re-filling of the trench by the contractors. The pit (Site 6) is shown in Fig. 7, derived from the survey of this section. The pit was of cylindrical shape with rounded base. Diameter was 4 ft. and depth 5½ ft. Six inches below the base of the pit, and separated from it by brickearth, was a layer of ash, 6 in. deep. This contained animal bone fragments, flints and a spindle whorl. The main part of the pit was filled with layers of ash, darkened soil, brickearth, sand and clay, and it contained many pottery fragments, broken flints and bone fragments. The upper layers contained pieces of metallic slag. The blue-green colouration of these suggested high copper content, but analysis by Mr. G. L. Bailey, of the British Non-ferrous Metals Association, showed it to be an iron slag - probably ferrous silicate with only traces of copper. The Finds (T.C.C.) Site 4. One struck flint flake. Site 5. One crackled flint, possibly a pot􀂞boiler and 29 sherds of coarse, flint-gritted pottery. 7. (Fig. 6) Much large flint grit; core dark grey; surfaces dark brown to black; rim flattened and neck smoothed. 103 0 .i:,.. SHOTFIELD FARM N PIT EAST WALL ONLY SUJ\VEYED OCT 1960 DJRO 6 PRESTON WATER MAIN TRENCH POTTERY 7 FEET Fig. 7. Pit (Site 6). s JOO 196.5 9 0 0 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH 8. Much fine and some large flint grit; core dark brown to grey; surfaces black, semicircular lip impressed in rim, with excess clay splayed inside and outside. 9. Much large flint grit; core and interior surface dark grey; exterior buff, orange and black. Angular shoulder decorated with rim of circular impressions; exterior surface below shoulder and above base marked with shallow scratches, predominantly vertical. Such scored decoration is most frequent in eastern England, 10 though probably of wider and longer occurrence than usually allowed. 10. Shoulder sherd of hard fabric with much flint grit; core brown; surfaces dark grey to brown. 11. Abundant large and some very large flint grit; core grey to orange; surfaces orange. Site 6. A few waste flint flakes, much slag, one lump of sandstone, bones of cow, pig and sheep, some calcined, one clay spindle-whorl, and 28 I sherds. 12. Some medium to large flint grit; core dark grey; surfaces light orange to brown. 13. Much medium and some large flint grit; core dark brown; surfaces light-brown with grey patches; exterior coated with light-brown slip now mostly abraded. Cf. Chinnor, Oxon., 11 and Risby, Suffolk. 12 14. Some medium to large flint grit; core and exterior dark grey; interior coated with light brown slip. 15. (Fig. 8) Much medium and large flint grit; core and surfaces dark grey; surfaces left rough; rim flattened. 16. Some small and medium flint grit; core dark grey; interior grey to black; exterior light brown and dark grey. 17. Much fine flint grit; core dark grey; surfaces light brown; rim folded over outwards and flattened. 18. Some fine flint grit; core black; interior black and smoothed, exterior light reddish brown, with traces of a bright cherry-red slip, possibly haematite, now mostly worn off; three wide flat grooves lightly impressed. 19 and 20. Some medium to large flint grit; core dark grey; surfaces grey to brown; exterior marked with vertical scratching (cf. 9). 21. Spindle-whorl of fine, brown fabric with some fine flint grit. Site 7. Some waste flint flakes, and ten sherds of coarse flint-gritted pottery. 10K. M. Kenyon, 'Excavations at Breedon-on-the-Hill, 1946', Trans. Leics. Arch. Sofr 􀂒6 (1950), 17-82, esp. 43-68. 12 Richardson and Young, op. cit., Fig. 8, no. 78. C. F. C. Hawkes, 'The Mamian Pottery and La Tene I brooch from Worth, Kent', AntiQ. Journ., XX (1940), 115-21, Fig. 14. 105 I J. D. OGILVIE 􀀓------􀀔---------·-· - - ---􀀅􀀆- ---- 16 /r,'\ 􀀌 21 I 􀀕 􀀍23 15 17 rw.6, •27 =- - 22 rr. Fig. 8. Finds from Site 6 (contd. nos. 15-21); Site 8 (nos. 22-28); Site 9 (nos. 29-30) (Scale:¼). 106 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH Site 8. Waste flint flakes, bones of cow and 98 sherds. 22. Soft fabric with much medium and large flint grit; core black; surfaces dark brown to black; exterior crudely smoothed, leaving vertical corrugations. 23. Hard fabric with much medium and some large flint grit; core brown; surfaces light reddish brown; decorated with a row of circular impressions. 24. Much fine and some medium flint grit; core and interior black; exterior and base dark reddish brown. 25. Much fine flint grit; core dark grey; interior black with traces of smoothing; exterior light brown; rim flattened and folded down outside. 26. Very coarse fabric with some large flint grit; core and surfaces reddish brown; rim flattened. 27. Hard fine fabric with some fine flint grit; core light grey to buff; surfaces buff. 28. Much medium and large grit; core and surfaces dark brown; rim flattened. Site 9. Waste flint flakes, slag, bones of cow and sheep, and 128 sherds. 29. Some fine flint grit; core grey; interior black and burnished; exterior now abraded but probably originally similar to interior; rim flattened inwards. 30. Some medium and large flint grit; core dark brown to dark grey; interior black· exterior dark brown with traces of horizontal ' smoothing; rim flattened and expanded, with finger-tip impressions. Site JO. Burnt flint lumps, possibly pot-boilers, bones of cow, and 61 sherds. 31. (Fig. 9) Some fine flint grit; core grey; surfaces dark grey to black; rim turned out and flattened. 32. Much fine flint grit; core dark grey; interior grey; exterior reddish brown. 33. Much fine and some medium flint grit; core and surfaces dark grey; rim slightly flattened. 34. Much medium and large, very angular flint grit; core and surfaces orange-brown. 35. Some fine flint grit; core grey; surfaces dark grey to black; surfaces smoothed; rim flattened. Site 11. Ten sherds of coarse flint-gritted pottery. Site 12. Waste flint flakes, slag, bones of cattle and sheep, and twelve sherds of coarse flint-gritted pottery; also one sherd with dark-brown glaze, post-medieval in date. 107 􀀂· ., ' J. D. OGILVIE 37 \) 39 40 ( , I 41 F \, > I \u l \ } 1-----1 _c \ .46 45 \,cc,ococccccc I 7: .a: ....... -􀀏- 􀀐/') 􀀌 .•. • , • • • • 49 Fig. 9. Finds from Site 10 (nos. 31-35); Site 13 (no. 36); Site 15 (nos. 37-38); Site 16 (nos. 39-40); Site 17 (nos. 41-49) (Scale: ¼). 108 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH Site 13. Unidentifiable bone fragments, slag and six sherds of flintgritted pottery. 36. Much small and some medium flint grit; core dark grey and dark brown. Sites 2-13 are evidence for an Iron Age occupation site or sites, though the circumstances of discovery make it impossible to say much about the nature of such occupation. The pottery is potentially a useful criterion of the site's chronology, but little Iron Age material has been published from east Kent, although extensive collections survive from sites such as Worth 13 and Dumpton Gap, 14 which may be broadly contemporary. In the absence of well-dated local groups, it is not of any great value to cite parallels from further afield, except in the case of some of the more distinctive and datable forms, such as nos. 1 and 13. These belong to a series of pottery forms found in southern and eastern England which imitate the styles of early La Tene in Europe, and can therefore be dated to the fifth to third centuries B.C. If all the groups from Sites 2-13 are indeed roughly contemporary, they present a good picture of local Iron Age pottery at that period. Fabrics are predominantly flint-gritted, features such as flattened rims are common, and firing conditions regularly produced black or browncoloured surfaces, with only rare examples of lighter tones. Forms are varied, but include large jars with rounded shoulders (9, 12, 16, 25), wide-mouthed jars with barrel-shaped profiles (3, 15, 28, 30, 3 I), jars with everted rims (27, 29), small angular bowls (10), and wide dishes with lips (8), as well as the forms (l and I 3) derived from early La l'ene types. Decoration is rare but fingertip impressions occur occasionally on the body (9, 23) or the rim (30); vertical scoring is also found (9, 19, 20). It is also clear that bases with low pedestals or foot-rings are a frequent element (I, 5, 14, 24). Shotfie/d Farm (Site 23) A few months after the completion of the excavation, the tenant of Shotfield Farm, Mr. C. Thornby, produced a large collection of pottery, which he had recovered while digging deeply to make a potato clamp at a point a few feet from the line of the trench where it left his farm. N.G.R. TR 25156055. Site 23. A large, but very mixed, collection of pottery, including, (a) some hand-made sherds with some or much large flint grits; some sherds are from very large jars (see also nos. 50, 51); (b) wheel-turned :34W. G. Klein, 'Roman Temple at Worth, Kent', Amiq. Journ., viii (1928), 76-86. H. Hurd, 'On a Late Celtic Village near Dumpton Gap, Broadstairs', Archaeologla, 61 (1909), 427-38. 109 I J. D. OGILVIE -:::.:... · .. 50 /111• 􀀂􀀃---.; --..... -::::::C::::. 51 I\ 57 59 \, Fig. 10. Finds from Site 23 (nos. 50-57); Site 25 (nos. 58-60); Site 30 (nos. 61-64); Site 36 (no. 65); Sites 38 and 39 (nos. 66-67) (Scale: ¼). 110 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH sherds, some in fabrics typical of the late Iron Age or early RomanoBritish period, some in more sandy Romano-British fabrics; (c) fragment of South Gaulish samian; (d) mortarium in cream fabric with quartzite grits; Oxfordshire, mid-third to fourth century; (e) cordoned, carinated bowl in sandy pinkish cream fabric with red colour-coat and rosette stamps; Oxfordshire, Fulford's Portchester Type 34.2; such Oxfordshire stamped wares are not found at Porthester before 330/340.15 50. (Fig. 10) Hand-made, some medium to large grits; core reddish brown; surfaces reddish brown to black. Cf. Minnis Bay, Kent. 16 51. Some medium and large flint grit; core and surface light brown; exterior very crudely smeared. 52. Sparse fine flint grit; core and surfaces dark-grey to black; chevron pattern lightly burnished on shoulder. 53. Slightly sandy light brown fabric; surfaces reddish brown to black; some grog temper. 54. Hard sandy, dark brown fabric. 55. Lid in dark grey, slightly micaceous fabric. 56. Fine grey fabric with black surfaces; not BB2, but possibly an imitation. 57. Soft, grey, sandy fabric. This assemblage is derived from at least three chronologically distinct phases of occupation. First, there are the hand-made wares with flint gritting (nos. 50, 51); these may belong to the very beginning of the Iron Age pottery tradition, possibly to the period of the eighth to fifth centuries B.C., though the bowl (no. 50) is such a simple and obvious type that it may have had a much longer life. Second come the South Gaulish samian and the wheel-turned vessels such as no. 52 in forms and fabrics typical of the final Iron Age and early RomanoBritish periods. Finally, a fourth-century date must be assigned to the Oxfordshire products. It is clearly a mixed collection, apparently from distinct phases rather than continuous occupation, but the reasons for such an assemblage cannot be suggested without further investigation. LENGTH 4 Longmete Farm (Sites 15-21) Immediately following the pits of the Shotfield length, a series of partly confluent ill-defined V-shaped trenches was crossed. It was not possible 15M. G. Fulford, 'The Pottery', in B. W. Cunliffe, Exca1•ations at Portchester Castle. I (1􀁻15), 316, and Fig. 174. B. J:. H. Worsfold, 'A Report on the Late Bronze Age Site excavated at Minnis Bay, irchington, Kent, 1938-40', PPS., ix (1943), 28-47, Fig. 6, no. 3. 111 J. D. OGILVIE to discern any distinctive pattern of the trenches. The filling of the trenches yielded numerous sherds. This length of trench was just west of the enclosure and ring ditch crop-marking shown on the map, Fig. 3,(6). The trench would not have cut through this presumably Iron Age site. The Finds (T.C.C.) Site 15. One struck flake, one oyster shell, and 14 sherds, including one South Gaulish Form 27 and two flagon sherds. 37. (Fig. 9) Cream fabric; grey colour-coat. 38. Soft, dark grey, sandy fabric, slightly vesicular. Site 16. 47 sherds and one sheep bone. Sherds not illustrated include two mortarium fragments in a dark-red sandy fabric with white slip. 39. Hard, slightly sandy grey fabric; exterior red brown. 40. Soft orange fabric. Site 17. 223 sherds and two fragments of millstone-grit rotary quern. Sherds not illustrated include, (a) stamp of PERPETVS of Rheinzabern, Hadrianic-Antonine; 17 (b) two other sherds of samian, one originally decorated but now too abraded for identification; (c) seven sherds of flagon; (d) mortarium fragments in a red sandy fabric with white slip (cf. Site 16); (e) mortarium fragment, in an orange fabric with multicoloured quartzite grits, from Oxfordshire, mid-third to fourth century A.D.18 41-46. Sandy fabric ranging in colour from dark-orange to brown. 47. Hard sandy bright-orange fabric with grey core. 48. Black-burnished fabric BB2, visually comparable to those produced in north Kent. 19 49. Mortarium in sandy cream fabric with quartzite grits; an 0'5-fordshire product, mid-third to fourth century. 20 Site 18. Bones of cattle and sheep. Site 19. Seven Romano-British sherds. Site 20. Seven Romano-British sherds and one iron nail. Site 21. Eight Romano-British sherds. 17F. Oswald, Index of Potters' Stamps on Terra Sigillata (1931), 238-9. 18C. J. Young, 'The Pottery Industry of the Oxford Region', in A. P. Detsicas (Ed)., Cu1r9rell/ Research in Romano-British Coarse Potlery (1973), 105-15. D. F. Williams, 'The Romano-British Black-Burnished Industry; an Essay on Characterisation by Heavy Mineral Analysis', in D. P. S. Peacock (Ed.), Pottery and Early Commerce: Characterisation and Trade in Roman and Later Ceramics (1977), 16230-220. Young, op. cit., 109. 112 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH Sites 15-21 represent a Romano-British occupation site of unknown nature. The chronological evidence extends from the first to the fourth century, though occupation may not have been continuous. The earliest sherd is the South Gaulish Form 27 (Site 15), which may date from the middle of the first century A.O. The stamp of PERPETVS (Site 17) can be assigned to the middle decades of the second century while the black-burnished pottery industry of Kent (no. 48) does not seem to have survived long into the third century. 21 Later occupation is indicated by the Oxfordshire mortaria which post-date the mid-third century. LENGTH 5 Longmete Road- Carpenters Farm (Sites 24-31) The trench here ran along the top of the flat ridge, mostly through orchards, to Four Turnings. The soil consisted of brickearth of depth varying from 1 to 10 ft. lying on sand and gravel. There were many small poorly-defined pits and soil discolourations. These contained flints, some showing signs of working, and fragments of soft gritted pottery, many of which were too soft for preservation. In the neighbourhood of one pit (Site 34) a saddle quern rubbingstone (Fig. 12) was dug up. It was submitted to Mr. S. E. Ellis, of the Petrological Department of the British Museum, Department of Mineralogy, who reported that 'it was a sarsen - a fine-grained siliceous sandstone - the presence of occasional glauconite grains and phosphatic fragments indicates that it cannot be one of the Wealden sandstones. It most closely resembles sarsens derived from the Woolwich Beds and may have been taken from these or from a river gravel, or found as a surface stone. It was probably obtained north of the Downs, in the Thames basin and not in the Weald'. The Finds {T.C.C.) Site 24. One flint core; some struck flakes; and five sherds of coarse flint-gritted ware, probably Iron Age. Site 25. Struck flint flakes, cattle bones, and 25 Iron Age sherds. 58. {Fig. 10) Much small to medium angular flint grit; core and surfaces dark red-brown. Cf. Site 13, no. 36. 59. Much small flint grit; core dark brown; surfaces black; rim rolled out and slightly flattened. Cf. Site 2, no. 3; Site 8, no. 28; Site I 0, no. 31. lfWilliams, op. cit., 213. ll3 J. D. OGILVIE 60. Much small flint grit; core and surfaces dark red-brown; vesicular exterior crudely smoothed. Site 26. Three struck flint flakes. sue 27. One struck flint flake. Site 28. No finds. Site 29. Struck flint flakes and one flint-gritted sherd, probably Iron Age. Site 30. Struck flint flakes and one core; and 32 handmade flint-gritted sherds. 61. Much small flint grit; core grey; surfaces grey to black; exterior smoothed. Cf. Site 5, no. 7. 62. Some medium flint grit; core and surface black. 63. Some medium and occasionally large flint grits; core and surfaces dark grey to black; rim flattened. Cf. Site 8, no. 26. 64. Some fine flint grit; core and surface dark grey; exterior lightly burnished and decorated with small diagonal incisions. Such decoration is rare in England, but is found especially in Kent, as at Dump ton Gap. 2 2 Other forms of surface treatment are also found, including rustication, combing, and even applied flint grains, but whatever style of decoration is used it is typically limited to the area below the shoulder, leaving the shoulder and neck plain, as well as a strip around the base. A similar range of decorative motifs is also found on the other side of the Channel, especially in Belgium2 3 and Southern Holland24 in the period contemporary with early La Tene. Site 31. Struck flint flakes and two coarse flint-gritted sherds; probably Iron Age. Sites 24-31 represent an extensive area of Iron Age activity, apparently contemporary with that of Sites 2-13. Although the quantity of pottery found is not large there are good similarities between the types found at the two sites, noted above. Furthermore, the incised decoration of no. 64 can be dated by its continental parallels to the same period, approximately fifth to third centuries B.C. One interesting feature of these sites is the occurrence of cores and struck flakes of flint. Unfortunately, the number of utilized flakes or other finished tools is very small, and the true nature of the industry 22 Hurd, op. cit., Fig. 3. 23 M. B. Marien, La Periode de La Tene en Belgique; le Groupe de la Haine (1961), 167-71. 24 G. J. Verwers, Das Kamps Veld in Haps in Neolithikum, Bronzezeit und Eisenzeit (1972), 137-40 and Abb. 62-75. 114 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH cannot be assessed; it is, of course, possible that nothing was being produced other than very crude blades. In the absence of more definite evidence, it cannot be certain that the flint industry is not of a much earlier phase, and the possibility of a crude flint industry in the Iron Age must still be allowed. In that case, other features such as Sites 32 and 33, though lacking diagnostically Iron Age pottery, may also belong to the same occupation. LENGTH 6 Four Turnings - Rushams (Sites 32-39) The ground here is a continuation of the ridge of the last length until it suddenly drops away into the water-logged hollow of brickearth at Little Rushams. The trench was dug through sand with a surface covering of black pebbles. Site 38, and an extension dug sideways from it for 3 ft., Site 39, were exactly at the point where the ground fell away to the hollow. The Finds (T.C.C.) Site 32. Struck flint flakes. Site 33. Struck flint flakes and one blade. Site 34. Struck flint flakes and four sherds, probably Romano-Brit1sh. Site 35. Struck flint flakes. Site 36. Nine coarse flint-gritted sherds, probably all Iron Age. 65. Much medium flint grit; core grey; interior red-brown; exterior black; internal bevel on rim and finger-tip impressions on outside. Site 37. Three Romano-British sherds and one piece of brick. Sites 38 and 39. Struck flint flakes and one core; one fragment of tile; cattle bones and I 06 sherds. 66. Hard sandy fabric; slightly micaceous grey core; dark-brown to black surfaces. Cf. Borden. 2 5 67. Some small flint grit; core and surfaces black; exterior treated with vertical combing. 68. (Fig. 11) Some small flint grit; core brown; surfaces black. 69. Soft grey fabric; surfaces burnished. 70. Hard sparse fine flint grit; core and surfaces black. 71. Hard; some fine flint grit; core and surfaces black. 25F. H. Worsfold, 'An Early Iron Age Site at Borden', Arch. Cant., lxi (1948), 148-55, Fig. 3, no. 10. 115 J. D. OGILVIE , ? 69 68 , 􀀁 70 / 􀀂 71 J (,, \ a ) 73 74 Fig. I I. Finds from Sites 38 and 39 (co111d. nos. 68-73); Site 60 (no. 74) (Scale:¼). 72. Some fine grog temper; core and surfaces brown. 73. Sandy, grey; surfaces brownish purple. This group (Sites 38, 39) belongs to the very end of the Iron Age or possibly even to the post-conquest period. No. 73 is important, as a local copy of a Gallo-Belgic plate; it must therefore date to later than 15/10 B.C. The rest of the assemblage is dominated by large jar forms, which can be paralleled at a variety of sites such as Faversham. 26 It is possible that this phase, characterized by Gallo-Belgic pottery and its imitations and local wares, mainly large jars, some with constricted, cordoned necks (as no. 68), can be divided from an earlier phase belonging to the second half of the first century B.C., in which GalloBelgic pottery is, of course, absent, but the local wares are also different, with fewer narrow-necked jars and a wider variety of small 26B. J. Philp, Excavations at Faversham, 1965 (1968), Figs. 23-5. 116 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH 0 2 3 INCHES Fig. 12. Saddle Quern Rubbing-stone (Site 34). 117 J. D. OGILVIE jar and bowl forms; this earlier assemblage might be seen at Snargate,27 Teston, 28 and Rose Lane, Canterbury.29 LENGTH 7 Little Rushams (Sites 40-42) This is the water-logged hollow which has since been drained and planted with fruit trees. It was not possible to identify any clear soil discolourations and finds were few. The Finds Site 40. Struck flakes; two coarse sherds possibly of Iron Age date; medieval and post-medieval glass and pottery, including stoneware. Site 4 I. Five coarse flint-gritted sherds, Iron Age. Site 42. Struck flakes, one core, one coarse flint-gritted Iron Age sherd. LENGTH 8 A257 (Canterbury-Sandwich) Road (Sites 43-44) The trench was here driven under the road through sand, for a distance of 30 ft. Close observation was therefore difficult, but medieval pottery was obtained. There was no evidence of road metalling below the modern layers. Dr. G. C. Dunning reported on the pottery and Dr. D. F. Williams on its mineralogy. One pot (Fig. 13, A) came from underneath the road, and a second pot (Fig. 13, B) from some vague pits a few feet south of the road. Report by Dr. G. C. Dunning, F.S.A. Since both pots appear to be about the same date in the latetwelfth or early-thirteenth centuries, they are evidence that a medieval settlement of some extent, but undefined in character, existed in this vicinity on Gobery Hill. A. Cooking pot of grey sandy fabric, with dark-grey surface red-toned on base. On the body the surface is uneven and much worked over by the fingers, and above the base it is knife• 27D. B. Kelly, 'Archaeological Notes from Maidstone Museum: Snargate', Arch. Cant .. lxxxiii (1968), 265. 28M. A. Ocock, 'Late Belgic Pottery from Court Lodge Farm, Teston', Arch. Cant., lxxxix (1974), 206--7. 29S. S. Frere, 'Canterbury Excavations, Summer 1946; the Rose Lane Sites', Arch. Cant., lxviii (1954), 101-43, Figs. 3, 6. 118 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH ------􀀌- --- I -- / / I :.=::=====·:::.:..-_ -------- r---- I\ -----------,---------------, ,, I ( I I I I / \ I I I Fig. I 3. A. Cooking•!)Ot (Site 43) (Scale: ¼). B. Pitcher (Site 44) (Scale: ¼). trimmed. Rim rounded and beaded on outside. The neck passes in a smooth curve into the rounded bulge of the body, and the basal angle is well defined. B. Sherd from upper part of body of a pitcher with tubular spout. Thick, grey sandy fabric, with red inner surfaces and black core. Light-grey surface, knife-trimmed in horizontal planes. Tubular spout attached above the bulge, and knife-trimmed as on body; inner margin of hole also enlarged by trimming. The cooking-pot belongs to a large class well represented in east Kent, notably at Rochester, 3° Canterbury, 31 and Dover Castle, 32 both in pit-groups and from stratified deposits. This 30 Arch. Cant., Ixxxvii (1972), 142, Figs. 15 and 17. 31 Arclr. Cant., lxi (1947), 38, Fig. 17, 10. 32 J.B.A.A .. 3 ser., xxx (1967), 110, Figs. 10, 11-16. 119 J. D. OGILVIE simple form of rim is typically a twelfth-century type rather than later, but probably continuing into the early-thirteenth century. The regularity of the rim and the smooth profile of the body point to the latter part of this period. The pitcher with tubular spout is late-Saxon in origin. 33 This form is found occasionally in east Kent, again at Rochester, 34 Canterbury, 35 and Dover Castle. 36 At these sites it is assigned to the twelfth and thirteenth centuries. Elsewhere, the latest example is from a deposit dated 1294 at 9ungay Castle. 37 Thus, the date brackets for both pots from Gobery Hill are rather wide, though there is some indication that they are later rather than earlier in the range. On these grounds, a central date about the turn of the twelfth and thirteenth centuries is assigned to them. Sherds of both pots were submitted to Dr. D. P. S. Peacock, Department of Archaeology, University of Southampton, who kindly arranged for Dr. D. F. Williams to carry out the petrological analysis. Report by Dr. D. F. Williams. Both samples are in a hard, very sandy darkish-grey fabric. Thin-sectioning shows numerous well-sorted sub-angular quartz grains, average size 0·40 to 0·80 mm., and a small amount of flint and mica. A heavy mineral separation was also carried out on both sherds, but the resulting assemblages yielded too few grains to give a reliable reading. However, it was noted that zircon, kyanite and tourmaline seem to be well represented. The two sherds appear to be of a similar fabric, though this is not immediately comparable to any known fabric type. Moreover, the raw materials are too commonplace to indicate a likely area of origin. Kyanite is found principally in post-Triassic sedimentary rocks, but this would include the majority of the eastern half of the country up to southern Yorkshire. All that can be said therefore is that the mineralogy is entirely in keeping with a Kentish source, though an origin further afield cannot be ruled out. 33E.g. Med. Arch., xvi (1972), 128, Fig. 37, d. 34 Arch. Cant., lxxxvii (1972), 149, Figs. 15, 25 and 27. 35 Arch. Cam., !ix (1946), 76, Fig. 7, 2. 36 J.B.A.A., 3 ser., xxx (1967), 112, Fig. 10. 37 Proc. Suffolk Inst. of Arch., xxii (1936), 338, Fig. 13. 120 STOURMOUTH ADISHAM TRENCH SECTION 25 24 l.ONGWETl ROAD ROMAN 15-21 23 􀄔Rt..Y IRON AG! 4-13 3 2 Pl:tESTON MILL U.Nt l'IEOUTHJC WET IRl(:K UATH 9RK:k EARTH I I I I I S.t.NO 4"'0 GRAV(L AT 5 F£( T &RC.lt tAA.TH GAAV(t.. 4 T 6 F£ET FEET FROM SOUTH 16 500 F'OUA. TUAN!f4GS 36 35 60 0 120 ,o 60 1'20 00 EARLY IRON AGE 34 &RIO: EARTH 33 BLACK PE8Gl£ AT 4 FEET SANO II SOO 12 000 WINGHAM RIVl!:A. 49 p 7000SILT 32 12 500 7 500 NII.AVV ROAD) CHALK 2 00 2 500 31 13000 48 8000 DOWNS DAMBRIDGE 'INLET., < 10 u. 0 3: 􀄕 10 1- ... 20 . . . . . . . . • CHALK • • • 65 00 66 00 ovu 6 7 00 6800 6900 70 00 7100 18 SOO 30 BRICK EART"f AOL.LEO 1'1.tNTS AT .t ,e:e:T ..... <:,OBERY HILL EA􀀊LY IRON AGt 14000 47 46 45 10 500 20 29 14 SOO ! LITTLt': fl!USHAMS ROAD A 257 Mf.01AEVA\. 44 43 42 41 2 0 SOO 28 27 26 5.UAVACE P!&!Li!S ON CLAY WET SANOl' CI..AY 8 SOO 9 SOO 10000 10 SOO !RONlE AG! IRON AGE. 60 S9 56-S8 S2-55 51 BRICK EARTH CHALK AT 8 F't£T 4000 SECTION P-Q ENLARGED 0Alr.18RIOG!. $..TA.liAM • • • CHAU( • • 7200 7> 74 00 WINGHAM AJll(R 75 00 SA.IQC. [AR.TH TO 10 f'UT ◄ 500 5 000 76 00 7700 78 00 79 Fig. 14. Longitudinal Section of Trench. N STOURUOUTH 21 16 81!LGIC 39 38 37 " 1•0•0• l 50 ,ooo THE STOURMOUTH􀀜ADISHAM WATER-MAIN TRENCH LENGTH 9 Gobery Hill. A257 - Wingham River Marsh (Sites 45-48) The trench passed over Gobery Hill through sandy brickearth, with a thick covering of Woolwich Bed pebbles. Near the top of the hill (Sites 45-48), N.G.R. TR 249578, there was a wide scatter of worked flints, cores, flakes and hammer pebbles, with numerous small illdefined pits. This was obviously a flint working site and the scatter of worked flints is exactly paralleled by three similar sites on the extensions of this ridge at Stone Down, Wenderton Manor and Wenderton Hoath, marked Y on the map (Fig. 3). It is hoped that a report on these will be published later. Site 48 was an isolated find in the trench where it entered the marsh. The Finds (T.C.C.) Site 45. 96 coarse flint-gritted sherds of Iron Age date. Site 46. Struck flakes: five coarse flint-gritted Iron Age sherds. Site 47. Struck flakes; 21 coarse, flint-gritted Iron Age sherds. Sites 45-47 represent another area of Iron Age activity, probably occupation. The sherds are all small and no forms can be reconstructed, but the fabrics would indicate a date in the earlier part of the Iron Age. Site 48. One handle fragment of medieval date. LENGTH 10 Durlock Stream at Dambridge The water􀂗main was here carried overground to cross the valley, for a distance of about 300 yds. and was supported on small piers. The contractors drilled the marsh vertically at regular intervals to ascertain at what depths firm soil lay. 􀂘e contractors generously supplied a copy of their longitudinal section of the trench. Fig. 14 is derived from this and the Dambndge Inlet enlargement shows the nature of the surface geology there. This accords with the supposition that this was probably an extension of the Wenderton tidal inlet. 121 J. D. OGILVIE LENGTH 11 Wingham - Neavy Down (Sites 49-60) This length consisted ofbrickearth overlying chalk. The chalk, at first, near the river, was at a depth of over 8 ft., but it became shallower as the downs were climbed, until it was immediately below the ploughed topsoil. There were numerous black W oolwich Bed pebbles on the surface. At' site 60, N.G.R. TR 244564, on the 75 ft. contour, V-shaped discolourations appeared in both sides of the trench. These were repeated at a distance of 100 ft. In these, and in an adjacent small pit, only small flint chips were found, but in the spoil piled up alongside by the mechanical excavator the workmen found a few pieces of pottery. Search through the spoil-heap produced further sherds. Together they were sufficient to reconstruct the beaker described. It can therefore be deduced that the trench cut through the ploughed-out remains of a barrow. The circular crop-markings revealed by Dr. St. Joseph's aerial photograph (Fig. 3, (21)) are at the top of the chalk ridge, less than half a mile away. The Finds (T.C.C.) Site 49. Struck flakes. Site 50. Pig bones. Site 51. Struck flakes; one piece of tile; two sherds, one certainly Romano-British, one possibly Iron Age. Site 52. Struck flakes; cattle teeth; and ten coarse sherds, probably Iron Age. Site 53. Ten coarse sherds, Iron Age. Site 54. Struck flakes; six coarse sherds, probably Iron Age. Sites 52-54 are yet another concentration of Iron Age activity, but are also very difficult to date. The fabrics would again argue a date in the first half of the Iron Age rather than later. Site 55. One struck flake; one post-medieval sherd. Site 56. Struck flakes; two coarse, flint-gritted sherds, probably Iron Age; one fragment of glass. Site 57. No finds. Site 58. One sherd, possibly Iron Age. 122 THE STOURMOUTH-ADISHAM WATER-MAIN TRENCH Site 59. One struck flake. Site 60. One struck flake; two sherds of coarse, black, flint-gritted pottery, probably Neolithic; and beaker sherds including no. 74 below, and two other sherds in a different fabric from another beaker. 74. (Fig. 11) Beaker in reddish fabric with much small to medium grit; surface decorated with zones of overlapping horizontal rows of comb-impressed dots, with about nine such rows between each plain horizontal band. This beaker has been published by Clarke,38 but with an inaccurate and over-schematized drawing, which omits the undecorated horizontal rows. With its rather globular body and short everted rim it falls into his East Anglian group; in the terminology of Lanting and van der Waals, 39 it would be assigned to Step 3 of the East AnglianKentish group, and on their tentative chronology be dated to c. 1900-1800 B.C. CONCLUSIONS (J.D.O.) Apart from the discovery of three new sites - Neolithic, Bronze Age, and medieval - the finds show that the low-contoured ridge from Stourmouth through Preston to Gobery Hill was occupied throughout its length in Iron Age times. Dr. Champion points out that there must have been at least three separate occupation sites on this line. The area round the modem Preston village must have been an important centre, continuously inhabited for a considerable period, as shown by: the Shotfield Farm Iron Age site; to the east the probable Iron Age site shown by aerial photography on Rookery Farm; between these the area of Romano-British occupation; and to the west the gravel pit area, where Dowker made his finds of Romano-British material, and from where came the quantities of samian ware described by the local inhabitants. This populous centre must have significance when considering the layout of Rutupiae and the Roman routes from Richborough to Canterbury. The original objective in observing the trench was to look for traces of the Roman road. Although the trench was examined throughout its length, except for about 200 ft., no trace of this road was found. The Ordnance Survey maps mark the A257 as a 'Roman Road'. The absence of deep road-metalling, and the presence of medieval pottery below the road show that this is unlikely. 38D. L. Clarke, Beaker Pottery of Great Britain and Ireland (1970), Corpus no. 409 and Fig. 389. 391, M. Lanting and J. D. van der Waals, 'British Beakers as seen from the Continent', Helinium, 12 (1972), 20-46. 123 J. D. OGILVIE ACKNOWLEDGEMENTS I am grateful to the landowners and tenants who allowed access to their land, and who have allowed all of the finds to be given to the museum, including particl1larly, the Kent County Council, the Viscount Hawarden, Messrs. Chandler and Dunn, C. Shepherd, C. Thomby, J. Marchant, C. Hulme and P. Holden; to the contractors, Messrs. Whittaker Ellis, their agent, Mr. L. Fisher and their staff; to the Thanet Water Board, their engineer and manager, Mr. J. L. Unitt, and their local engineer, Mr. Hoile, for their co-operation on the site, their advance notices of their programme of work and for their gifts of plans and records of borings; and to local archaeological helpers, including Mr. and Mrs. R. Carr, Mr. D. G. Downes, Mr. A. Southam, and my wife and family. I particularly thank David Ogilvie for his work on the site; for his surveys and drawings; and for his help with this account, especially in the preparation of the maps and the drawing of Fig. 12. I am deeply indebted to the late Major F. W. Tomlinson, F.S.A., to whom I reported the first finds, and who, realizing their importance brought to the site Dr. G. C. Dunning, who then introduced me to the chain of experts, starting with Miss Jean Cook, F.S.A., then at the Royal Museum, Canterbury, and ending with Dr. T. C. Champion, without whose help this report would probably never have achieved completion. The finds and records are now deposited at the Royal Museum, Canterbury. 124