More information : (ST 112031) Hembury Fort (NR) (1)
The fort on a narrow spur of the Blackdown Hills was excavated by Miss Dorothy Liddell in 1930-35, when it produced evidence of occupation at three periods.
Neolithic: the ditches of a causewayed camp, eight in number and up to 7 feet deep, had been dug in an arc (A on illustration) across the southern half of the ridge and a similar line of ditches seems to have been cut below the northeastern lip of the ridge. Here, however, only one ditch was found under the east entrance to the later Iron Age fort but it is believed that other ditches underlie the ramparts. A recent re-assessment by I.F. Smith (3) suggests that these ditches represent adaptation to a promontory position of the usual type of causewayed camp, in which Wilson (4) sees the ditches as defining one side only; natural features supplying the others. The southern enclosure was backed by a rampart and had a timber-framed entrance and substantial hut (B) inside it. At the point of the spur (D) intensive occupation had taken place, with evidence of trade in Cornish greenstone axes, jet and steatite beads besides an extensive flint industry.
Iron Age B : Initially the 9-acre Iron Age fort, was given defences consisting of two palisades set in stone-packed trenches revetted behind with scraped soil. Later the two great ditches were dug and the three ramparts raised all round except to the north-east and east sides where the outer rampart was never completed. Inturned entrances approached by hollow-ways occupied the northeast and west sides, probably bridged by gate-towers above double gates. No hut circles or evidence for settled occupation were found, but pottery was of southwestern Iron Age B.
Iron Age C: Shortly before the Roman occupation, judging by a Belgic bead-rim bowl of AD 45 found in the primary silt of the southern of two transverse embanked ditches, the Iron Age B fort was divided into two enclosures. The southern of these, probably a stock enclosure, was through the former west gate and an obliquely set gateway in the south transverse bank and ditch. No huts were found for this period but Roman pottery and a coin of Claudius indicate occupation to AD 70. A coin of Nero (5) was found in a wall at Hembury in 1963.
Traces of a hollow-way from the east gate of the fort can be followed (ST 11350321 to 11390313). An alleged hut circle with an entrance through its bank (at ST11280313) yielded an arrowhead, 2 scrapers and 32 other flints, 17 sling stones and an Iron Age potsherd. Another alleged hut circle, a bare area in young bracken (at ST 11280311) was barren except for 3 slingstones and flint flakes. A circular platform about 80 feet diameter (at ST 11240298), cut into the hill on its west side but embanked to a height of 3ft. on its east and having a slight ditch, yielded only one sling stone. A semi-circular bank 25 feet wide, enclosing the south point of the fort (at ST 11240291), sealed an Iron Age coloured glass bead on the original turf line. Apart from the last enclosure Miss Liddell attributed the other features to scythe stone quarrying. (2-5)
Surveyed at 1:2500. Ridged and pitted ground on the steep hillside in Hill Plantation immediately below the east gate of the fort (at ST 11360321) is probably due to quarrying for scythe stones. (6)
A research project in Hembury was commenced in 1980. The results of the 1980 excavations may be summarised as follows:
1. the inner cross-bank was sectioned and found to be modern. It is presumed that the outer cross-bank is also of modern date. These earthworks may be connected with the fair which was held on the hilltop in the eighteenth century.
2. a small area inside the eastern iron Age rampart produced the remains of a well built rectangular structure some 4m wide. Its constructional details was stongly reminiscent of Roman military work and pottery associated with its occupaton is provisionally dated to AD 50-70. The Roman building was laid out over the site of an Iron Age round house some 7.0m in diameter. This is the first certain Iron Age structure to be recorded at Hembury.
3. Neolithic features were recorded in the area immediately behind the eastern Iron Age rampart. Two shallow pits containing pottery and flint implements were found, but there were no traces of structures.
It is anticipated that work will continue in 1981. (7)
Hembury hillfort is an elongated pear-shaped enclosure of 3 ha., defended by close set double ramparts ith ditches. The outer ditch is accompanied by a counterscarp bank on the north and west, and is obscured on the east. The interior is divided near the middle by two relatively slight crossbanks. Entrances are on the east and west. The hill was originally occupied by a Ne. causewayed camp.
The initial IA defence was a single rampart (following the line of the present inner bank), built up from an internal quarry ditch and turf faced, with an outer facing formed by a palisade of close-set posts. At the entrances a second palisade seems to have had looped outworks for additional protecton, but details are unclear and this work may have been unfinished. In the second and main phase both ditches were dug, the outer bank and counterscarp bank constructed and the inner bank enlarged; all were probably of glacis type. The outermost on the north was unfinished. Gateways were reconstructed as timber revetted passages, probably bridged, between short inturns. Both phases were associated with IA 'B' pottery.
The two cross banks were built later and Belgic or IA 'C' pottery has been found in association. The banks run parallel, 13.0m. apart, with ditches on the south. The northern one is longer though neither is strong and no more than 10.0m wide. At the southern bank postholes between the overlapping ends of the ditch presumably held a gate. These works are not the conversion of the fort to a 'citadel' but yet seem rather substantial for cattle control. They may imply social division. It is noticeable that the inner rampart of the fort is stronger to the north of the crossbanks than it is to the south. (8)
Hembury Fort (c) occupies the narrow southern end of a spur where precipitous slopes present a natural defence. The interior is level and at 267m. O.D. The fort is generally as described by authority 8, though some points can be qualified as a result of the present re-survey.
The Inner Rampart : On the north the inner face is up to 2.5m. high, and is fairly strong along the west side of the fort but on the east it is exceedingly weak, mostly less than 1.0m high. A true glacis type bank occurs for only 120m. along the southern part of the east side; elsewhere on the outer face there is a berm, 2.0m to 5.0m wide, interrupted occasionally by natural slip. The berm appears to be the undisturbed natural hillslope and not a terrace that has resulted from scarping.
The Second Rampart : Of glacis type, with a wide ditch around much of the circuit and a fragmentary counterscarp bank, much rebuilt as a parish boundary bank. Externally, along the east side and part of the west side the natural hillslope appears to have been steepened for a distance of about 10m., possibly by depositing upcast from the ditch.
The Third Rampart. An additional third rampart extends for 80m, covering part of the vulnerable north end. It is of considerable strength but unfinished; the intention to extend the work across the spur is indicated by a further piece of embryonic bank at the northeast corner.
The West Entrance : A hollow way set obliquely through the ramparts, approached by a terrace way which may be the original route, re-cut in modern times. A parallel terrace in pasture field to the west is another possibility though this is now blocked at the northern end.
The Northeast Entrance : Broad, low, inturned banks are the primary feature. The approach through the ramparts has been much disturbed by excavation, slip, and modern pathways. There is no obvious route to the ramparts; hollow ways and terraces coincide with various pits and quarry diggings.
Internal Crossbanks : The northern crossbank, 1.0m high, has a distinct step in its southern face with a berm 1.0m wide. No accompanying ditch was seen although one was noted in 1953 and depicted on the OS 25" 1957 and on the plan by authority 5. The southern crossbank between 0.5 and 1.0m high, is staggered, with an outer ditch 0.4m. deep and an almost central entrance gap.
Field System : The possibility of a field system within the fort is suggested by a few prominent lynchets, but bracken growth at the time of investigation inhibited proper assessment.
Holloway : Outside the hillfort, and obviously of later date, a well defined pack-horse route extends from ST 11190339 at the top of the spur to ST 11140323 near the bottom. It is mostly between 1.0m. and 2.5m. deep, encroached upon at one point by an old sand pit.
Hembury Fort, surveyed at 1/2500 for MSD and at 1/1250 (annotated copy).
No profiles of the earthworks were possible because of bracken up to 2.0m. high. (9)
Re-examination of the results of Liddell's excavations of1930-5 showed the high quality of the early Roman pottery. Secondly, at least one of the fragmentary structures recorded by Liddell and interpreted as a rectangular Iron Age hut, appeared to be of unmistakably Roman Military construction.
Excavations of Prof. M. Todd between 1980-2 within the NE part of the hillfort revealed large timber structures of early Roman date and clearly military construction. One of these, a courtyard building associated with considerable quantities of iron slag was probably a fabrica. The occupation of these buildings appears to have been brief as relatively few finds were recovered. An occupation of c 50 to 60 A.D. is suggested. There is clear evidence of demolition at the end of the occupation. The buildings were carefully dismantled, the uprights being usually sawn off at ground-level.
Miss Liddell's plans of the western Iron Age gate shows the latest phase as a series of square post-holes each being some 18cms square. These form a simple entrance gate closely paralleled by the Roman gate inserted into the NW angle of the Iron Age ramparts of Hod Hill. It is suggested that the rebuilding of this gate at Hembury early in the Roman period might imply that much or probably all of the northern end of the hillfort was adopted for use by the Roman army.
(Excavation summaries in Britannia 12, 13 and 14 erroneously refer to Hemburg hillfort (SS 41 NW 1), in Buckland Brewer parish (SS 427178).) (10-13)
Three radiocarbon dates have been pbtained from the Neolithic enclosure:
BM-138. 3300 +- 150 bc. Charcoal from middle silting of ditch.
BM-136. 3240 +- 150 bc. Charcoal from the occupation area.
BM-130. 3150 +- 150 bc. Charcoal from the ditch bottom. (14)
A rapid examination of air photography (15a) shows the earthworks of this site, including the 1983-1984 excavations. (15)
The site was included in RCHME's national project to record Industry and Enclosure in the Neolithic (16). Since all the upstanding earthworks are thought to be of Iron Age or Post Medieval date, no detailed fieldwork was undertaken, although a brief site visit was carried out. (16)
Further references. (17-28)
The Roman occupation at Hembury began circa 55 CE and ended before 70 CE. (29) |