Excavations in the two Iron Age Hill-Forts on Castle Hill, Capel, near Tonbridge, 1965 and 1969-71

EXCAVATIONS IN THE TWO IRON AGE HILL-FORTS ON CASTLE HILL, CAPEL, NEAR TONBRIDGE, 1966 AND 1969-71 By J. H. MONEY, M.A., F.S.A. lN'.1.'ROl>UC'.1.'lON CASTLE Hn.L (N.G.R. TQ 608439), in the parish of Capel, lies 2 miles (3 ·2 km.) south-west of the centre of Tonbridge (Fig. 1). The O.S. 6-inch map of 1872 gives no indication that the earthworks on Castle Hill had by then been discovered. This map, however, refers to Castle Hill and Castle Hill Woods, which suggests a tradition that the place had onoe been fortified. On the top of the hill the map shows a cleared area of much the same size as it is today; this clearing was scheduled as grassland in 1849 when the estate was purchased by Somerhill. According to Winbolt the earthworks did not appear on the O.S. 6-inch map until 1912. It is not known whetl1er or not the fortifications survived in the area of grassland in or after 1849, but they had certainly been demolished before Winbolt's excavations of 1929,1 when the field was arable. It was no doubt the destruction of the defences in the arable field which misled Winbolt into assuming that there was only one fort, and this is what he showed on his plan, using field boundaries of no great age to link the visible remains of what were two separate fortifications. Winbolt found a few flint artefacts and some iron slag, which were deposited in the Tonbridge Public Library but have since been lost. In the east entrance (of what is now called Fort I, see Fig. 1) he recognized a roadway paved with ironstone no dules and what he calls 'sandstone slats', and revetting stones along its sides. In other respects his report is of little or no value. Credit for detecting the existence of two separate earthworks goes to the late Mr. E. Geary, field investigator in the Archreology Division of the Ordnance Survey,2 who in 1969 recorded (of what is now called Fort II): 'On the south-west of the spur is a kidney-shaped enclosure, with the characteristics of an Iron-Age contour fort, formed by a rampart and outer ditch. The north-east side is under plough but has slight indications of a causewayed entrance'. These details, including the ploughed-out north-east defences, were added to the O.S. maps. Geary also gives a description of the visible remains of Fort I, and 1 S. E. Winbolt, 'Castle Hill Ca.mp, Tonbridge', Arch. Cant., xli (1929), 193-5. 1 O.S. Record Card. 61 J. H. MONEY comments: 'The true nature of this earthwork is uncertain; it does cut off the promontory but its defensive value is negatived by the weak flanking slopes . . . It may be an unfinished work, possibly a later strengthening of the contour fort.' In 1964, I was asked by the Inspectorate of Ancient Monuments to keep an eye on construction work preparatory to the erection of the television mast which stands at TQ 607440 just outside the north-west corner of Fort I. The work took place in May and June 1965 (see pp. 64-5 below). I had already decided to explore selected parts of the earthworks, in the hope of solving some of the outstanding problems; the excavations which took place in 1969, 1970 and 1971 are described below.8 ENVIBONMENT, GEOLOGY A.ND DATE The two forts (Fig. 1) are situated on a spur of high ground running from north-east to south-west, around 400 ft. above sea level. Although the natural slopes are nowhere very steep, the site commands the surrounding area, and from it the inhabitants would have been able to exercise direct control over the ridge between Tonbridge (TQ 5845) and Pembury (TQ 6240)-a natural north-west to south-east route which the A21 follows today. They would also have been able to exercise indirect control over the crossing of the River Medway at Tonbridge, where the valley is narrowest and which has thus always been a vital point on an important north-south route across the Weald.4 A track (now partly demolished by clearance work beside the A21) descends from the entrance of Fort I in a northerly direction. Winbolt marks5 another track (of which traces still survive) running similarly south-east. If ancient, as seems possible, these tracks would have linked the fort with, or may indeed have been, the north-west to south-east ridgeway. Castle Hill is on an outcrop of Lower Tunbridge Wells Sand. Wherever we dug, the sandstone was encountered a short distance below the land surface. As well as providing a reasonably dry base for the occupiers, the sandstone also facilitated the digging of steep-sided ditches and provided good material for ramparts and revetments. A number of worked flints and waste material of mainly Mesolithic type came to light during the excavations (see Appendix: B): also three late Neolithic sherds (Appendix: A). There were no concentrations 3 Interim reports a,re in Arch. Oant., Jxxxiv (1969), 233-4; lxxxv (1970), 176-7; Jxxxvi (1971), 233--4; and lxxxvii (1972), 219. 'This route (Cr.oss-in-Hand-Mark Cross--Frant-Tunbridge WellsSoutbborough- Tonbrid􀂛e-Ightham) and its context are described in I. D. Marge.ry, Roman Ways in the Weald, 2nd Edition, London, 1949, 258, 259, 264 and 265. 1 Op. cit. in n.l, 194. 62 EXCAVATIONS AT CASTLE RILL, TONBRIDGE 0 \00 1/"l'"I" II I 200 300 400 500 I I I I I I I \000 --'r-,--,-1-,--'---'-----...--------....JJ FT. I M. 0 LONDON/\ vwijo 􀀇 Sc¥end 5 Fro. 4· .. . ·" 20 Post-hot« Sto.ko-holes Cvertica,I) Stake -hof&s Choriiontal> 30 FT. 10 M. 7,.-72 -- - --- fFortI. '""·" tedAreaofeaat n ™ 􀀵 '"" Detailed Plan ofe xcava E trance o (partially excavated) I I IJ 1-L ....L􀀄 L I p - - IA -􀀂 \•􀀄I ,s I I􀀂 I􀀅 1f .!. FIG. 5. FORT I E. ENTRANCE BURNT TIMBERS & TUMBUO REVETMENT IN NORTH OITCH•END Pianka Beam, Stake ?Strut 2 -11. 11. 14. 15, 17 1. 16 10 12 Mixed fragments 13 N J L 7r Above: Plan of burnt Timbers and tumbled Revetment in north Ditch-end of east Entrance of Fort I and seoondary Bank. (The partially excavated area waa not dug deep enough to reaoh timbers and stones.) Below: Section A-B through Ditch Silt, Timbers, Stones, and secondary Banlc. EXCAVATIONS AT CASTLE filLL, TONBRIDGE hold the posts in position. The post-hole on the north side was about 8 in. (0·20 m.) in diameter and flat at its base; that on the south side was 10 in. (0·25 m.) in diameter and ended in a blunted point. On the south side of the entrance, 3 ft. (0·91 m.) to the east of the latter, was a third major post-hole but there was no corresponding hole on the north side; this may also have been a gate-post (perhaps a later addition). These three major post-holes, together with the lesser holes set roughly in two lines across the road-way, possibly held the uprights of a bridge over the entrance. A few roughly shaped stones (five on the north and two on the south) appeared to be in situ and may represent the remains of a revetment. In the ends of both ramparts there was a number of stakehol es. The timbers they once held were probably designed to consolidate the rampart material and check erosion. On the forward slope of the rampart on both sides of the entrance were many more similar stake-holes. The ditch was about 24 ft. (7 ·31 m.) wide. On the south side it was excavated only to the extent of determining the rim of the ditch-end. On the north the greater part of the ditch-end was e:x;cavated, with remarkable results. The silt contained considerable numbers of revetting stones and burnt timbers (planks,- beams, stakes and a strut) which had tumbled from the inner rampart. Whether the destruction was deliberate or accidental is not known, but it must have been a deciding factor in causing the abandonment of the defences of Fort I and its eventual replacement by Fort II on another part of the hill. Mr. Christopher Giles excavated this part meticulously and made a careful study of the timbers, which are planned as a whole (regardless of their level in the silt) in Fig. 6; here there is also a section up the ditch-end which shows the actual position of four of the planks (11, 14 17 and one unnumbered) lying amongst the silt and tumbled revetting stones. There can be no doubt that the planks, which varied in thickness from 2 in. (0·05 m.) to only½ in. (0·013 m.) had been sawn. In detail the timbers (Fig. 5) were as follows: 1. Beam (oak) o. 5 in. (0·13 m.) in diameter, with a notch which possibly joined it to another timber. 2. Planlt o. l½ in. (0·04 m.) thick; only one surface was charred; traces of unburnt wood on the other side still survived. 3. Planlc o. 7 in. (0 · 18 m.) wide, twisted about its longitudinal axis. 4. Plank (oak) with notch (of. 1). 5. Plank o. 8 in. (0·20 m.) wide. 6. Plank c. 6 in. (0·15 m.) wide. 7. Plank c. 6 in. (0·16 m.) wide and l½ in. (0·038 m.)-2 in. (0·05 m.) thick, lying on edge. 8. Plank, of which only part projected into the trench. 71 J. R. MONEY 9. Plank c. 5 in. (0 · 13 m.) wide, 10. Plank (birch) c. 6 in. (0 · 13 m.) in diameter, with a sharpened point. 11. Plank (oak) c. 11 in. (0·28 m.) wide and½ in. (0·013 m.) thick. 12. Strut (ash) c. 2 in. (0·05 m.) wide and 1 in. (0 ·025 m.) thick. 13. Area of mixed fragments. 14. Plank (oak) at least 6 in. (0·15 m.) wide, of which only part projected into trench. 15. Plank (oak) c. 5 in. (0·13 m.) wide and possibly 2 in. (0·05 m.) thick. 16. Part of a beam (oak) standing nearly on end amongst revetting stones. 17. Plank c. 12 in. (0·30 m.) wide and ½ in. (0·013 m.) thick, lying almost on the natural and immediately under a stone slab. It will be observed that no. 17 and the slab overlying it are covered by a low bank, which must, therefore, be later than the collapse of the stone and timber revetment. It is likely that the material of this secondary bank was dug from the gully, which runs almost the whole length of the entrance on the north side and is matched by a similar gully on the south side. The date and purpose of these gullies are unknown. A possible explanation is that they belong to a peaceful phase after the collapse of the defences, when the top of the hill was given over to farming which required drainage. The gullies were peppered with small stake-holes which may represent fencing, designed to prevent livestock from straying into the ditches on either side of the causeway, or to control erosion, or both. The roadway, the surface of which had been worn away completely in the inner entrance, was well preserved between the ditch-ends; it was made of iron-stone nodules rammed into the old ]and surface. In and beside it was a group of medium and small post-holes which may have held the uprights of a removable or temporary timber barrier a.cross this pa.rt of the entrance. Here, below the level of the road and out into the old land surface, was the beginning of a small cutting, a.bout 3 in. (0·076 m.) deep and between 4 in. (O·IO m.) and 8 in. (0·20 m.) wide, which ran westward along the line of the road; if not a drainage gully, it may have been a 'marking-out' trench to guide the course of the entrance. A good part of the outer entrance had survived. It had stone revetments on either side (Plate III) and numerous post-holes and stake-holes which held timbers for supporting the revetment, checking erosion and, where there were a number of medium-sized holes in the roadway, possibly carrying a bridge between the rampart-ends. A group of large and medium post-holes 14 ft. (4·26 m.) to the west of 72 Rite K (Fort I). Section through rock-cut Ditch and outer Jtampart, looking North. (The arrow points at artificial cuts in tho rock; see p. 6.5.) East .Entrance of Fort I. looking South. Holes which held Gate-posts of Entrance. PLATE II East Entrance of Fort I, looking Sc-uth. General Vie\V of outer Entrance, sho\Ving revetted Rampart-ends. Gullies and Roadway. The vertical ranging rod stands in the southern gully; the hori?.ontal rod, along the middle of the roadway. 􀂎i!o D (Fort, 11 ), looking Nort-h-west. Section t,hrough JJ<>foncos, showing mRin Hampart nnd counterscarp Bank. Sito D (Fort II). Sido of rock-cut Ditch and main Rampart, looking South-cast. The arrow points at, the revetment stake-holes, somo of which are visiblo (se e pp. 73-4). PLATE IV Site E (Fort JI), looking ::S orth-east. Section through rock-cut Ditch in ploughed J,·ield. The light-colottred filling nt the top is demolished rampart material. EXCAVATIONS AT CASTLE HILL, TONBRIDGE the supposed bridge may be the remains of another removable or temporary barrier. On the crest of the outer rampart were post-holes, which in part probably represented a palisade (see also Fig. 3) and in part stakes to oheok erosion. Outside the rampart proper, but linked to it, were outer works flanking the roadway. The northern ex;tension was in the form of a platform, on which palisade post-holes were found over a length of 15 ft. (4·57 m.) where the top of the earthwork was e:x:amined. The cobbling of the roadway was patchy in the outer entrance and became generally more degraded between the rampart e:x:tensions. 􀃚,Once outside the earthworks the roadway meets the ridgeway whioh ,. passes below the defences (see p. 62 above). 1 , The site yielded ·eight pieces of pottery, including one rim-sherd 􀃛 ,(see Appendix: A). ·􀃜:Fort II ;":;. Fort II, which enclosed about 2·5 acres (l ·01 hectares), lies on the ·south-west part of the spur. Two-thirds of the earthworks still survive rt; the woodland; they consist basically of a single bank and ditch, , ex.oept on the north-west side where the slope of the hill-side is reasonably Wsteep and there is a low outer or counter-scarp bank. On the north-east , . side the rampart has been demolished but is still visible on the ground '.as a slight swelling running across the :field. There is no trace of an "entrance through any part of the surviving defences, and nothing convincing in the ploughed area shows up on any of the air photographs I have seen. On the analogy of Fort I, a likely place of the entrance is in the middle of the north-east defences (see Fig. 1). Sites D and F (Fig. 2; Plates IV and V) (1969) A trench 5 ft. (l ·52 m.) wide and 74 ft. (22 ·55 m.) long (Site D) was dug through the defences on the north-west side (see Plate IV), and behind it an area 10 ft. (3 ·04 m.) by 8 ft. (2 ·44 m.) (Site F) (Fig. 2). The inner rampart, which was composed of soil, broken rook and sand, all dug from the ditch (there was no tra.ce of quarrying behind the rampart), was 19 ft. (5 ·79 m.) wide at its base and had survived to a ma:x;imum height of 4 ft. (l ·22 m.). The butt-end of a palisade post-hole 9 in. (0 · 23 m.) across was found on the crest of the rampart. The forward face of the rampart was held up by a timber revetment, six post-holes of which were found, four at regular intervals of about 18 in. (0 ·45 m.) and two smaller holes close together besides one of the four. Some of the holes are visible in Plate V; one is shown in the section (Fig. 2). Their dimensions and characteristics are as follows: 73 J. H. MONEY Post-hole Depth Width (t 4 in. (0·10 m.) 2 in. (0 ·05 m.) f3 16 in. (0·40 m.) 4 in. (0·10 m.) y 21 in. (0·53 m.) 2 in. (0·05 m.) 8 7 in. (0·17 m.) 4 in. (0·10 m.) f: 11 in. (0·28 m.) 1 in. (0·025 m.) { 15 in. (0·38 m.) 1 in. (0·025 m.) f3 entered the ground obliquely so that the post emerging from it would have pointed forward down the hill at an angle of about 50° from the horizontal; it also pointed obliquely across the trench, i.e. almost due west. y also bent forward at an angle of about 50°, but did not lean laterally, e and { both tilted forward-e only forward like y, ( both forward and westward like /3. a. and 8 tilted slightly backwards. It is difficult to e:x:plain how the revetting posts were arranged without examining them over a greater width of rampart, but one possibility is that the post-holes into which the main uprights supporting the front of the bank were driven have disappeared entirely owing to erosion, and what is left is the remains of a rear line of posts (which would not originally have been visible) driven in at various angles in order to strengthen the rampart structure. The front of the rampart appears to have been supported also by some form of dry-stone walling (the layered rock which occurs in places in the sandstone would have been suitable for this). Several flat stones, of which the largest was 8 in. (0·20 m.) by 6 in. (0 · 15 m.), were found resting on the side of the ditch just below its lip . .A considerably larger group of stones, including some 12 in. (0 ·30 m.) across, also occurred in the silt, about half-way down the ditch. The old topsoil which had survived below and behind the rampart was a brown silt and rather loamy. Immediately behind where the topsoil is cut into the ditoh, there was a slight hump, which may be turf deposited either as a 'marking-out' bank or as· a footing for the rampart; no distinction, however, between deposited and undisturbed topsoil was noticed. There was a considerable amount of charcoal (oak, hawthorn, and lime) resting on and in the old land surface under and behind the back of the rampart. In one area, the surface was also reddened by burning, and there was a number of small burnt stones. It is possible that these were the remains of a hearth used by the builders, but equally they could have resulted from the burning of unwanted wood and undergrowth cut during clearing operations. The ditch, which was V-shaped and 20 ft. (6·09 m.) wide, was out out of the natural rook. Having been broken up, the excavated mat.erial was dumped on either side to form the main rampart and the counter- 74 EXCAVATIONS AT CASTLE HILL, TONBRIDGE scarp bank. Apart from the group of largish stones which had probably once formed part of the revetment, the silt and tumbled stone in the ditch contained nothing noteworthy; no occupation material except charcoal was found in the ditch. The counterscarp bank, which was 14 ft. (4·26 m.) wide, was made of soil, sand and a small amount of broken rock. It rested directly on the natural rock, with no topsoil intervening. It may be that the topsoil had been removed for rampart material before the counterscarp bank was built or had washed down the hill. A third possible explanation is that proposed by Dr. I. W. Cornwall at High Rocks, Site J, namely, 'that the buried soil was insinuated throughout the material of the counterscarp by worms operating under its shallow cover'. 8 Beyond the counterscarp was a thin yellow clayey layer of hill-wash stretching away do,vn the side of the hill. An area 10 ft. (3·04 m.) by 8 ft. (2·43 m.) was e:x:cavated inside the fort behind the rampart section, in the hope of finding occupation material. Apart from a few flint flakes, charcoal and pieces of burnt sandstone, no human remains were found. Site E (Fig. 6; Plate VI) (1969) This site was chosen because the O.S. 1/2500 map, following the late Mr. Geary's survey, marks an entrance here (cf. p. 61 above), but our work proved conclusively that none existed ( of. also pp. 78 below). There is no break in the rampart in Site E, and test holes at frequent intervals along 100 ft. (30·48 m.) of the ditch in Site E and south-east of it showed that there was no causeway and, therefore, no entrance in this sector. Dr. Anthony Harding made a skilful e:x:amination of the rampart and the area immediately behind it; the details are shown in Fig. 6. The plough soil, which covered the site, was 9 in. (0·23 m.) deep. Immediately below this, in the middle of the site, the base of the rampart was encountered. The rampart material consisted predominantly of orange loam with an admi:x:ture of yellow and light grey clay and small chips of sandstone. The fact that this tiny vestige of the rampart was undisturbed makes it clear that there had been no deepploughing of the field. In B4 a small pile of stones was found-possibly a 'marking-out' cairn, but more probably just part of the rampart material. The back of the rampart was delimited with some certainty, and in it was a number of small post-holes, some appro:idmately in a line, which may represent a fence that originally held up the material. Some traces occupation behind the rampart were noted. In 06 was a number of • Op. cit. inn. 6, 168. 75 B D E 6 . - _, ,.. •• - •• - - - • .• -- I (partielly •' ,f.: . 6 . . .. 5 4 3 : v ·;,-✓􀀉/ -/-✓- - - -· , • - · · · · · -􀀎- ·?0·' ?i- , (stokcholoY ch-1.-COII 0" okl // 􀀓 /( / // i land ,-u􀀘hoce 1//.1 /no recorded) : -􀀷·:. '!;􀀸'!.. - - 􀀹/_,c_/_/_/_'- - - 􀀺"'- ---- 11611.,,. •ft