Dunion Hill

NT61NW 3.00 6255 1906

(NT 6255 1904) Fort (NR) (site of)

(Centred NT 6240 1903) Hut Circles (NR) (3 shown)

OS 25" map, (1969).

See also NT61NW 3.01 NT 6262 1915 Settlement: Unenclosed Platform

The structural remains on Dunion Hill comprise the fragmentary ramparts of a fort, perhaps of Dark Age date, seven hut circles, scattered over the W face of the hill, which are probably all later in date than the fort; and some later foundations and lengths of stone dyke. A large quarry has already bitten deeply into the outlying defences of the fort on the N, and is being steadily enlarged.

The main rampart of the fort, a massive drystone wall, appears to have enclosed the whole summit of the hill (1092 ft OD). This is an oblong plateau, of irregular shape, measuring 550 ft NE-SW by 180 ft transversely and bounded on the SE and SW by precipitous slopes, on the NE by a broad gully lying at right angles to the major axis of the hill, and on the NW by a rock ridge interrupted, a short distance NE of the centre, by a gap which affords the easiest approach to the summit area. No trace of the rampart survives on the NE or SE sides, and on the SW side all that is left is a slight bank bordering the edge of the precipice at the WSW corner; on the NW side, however, its course is still plainly shown by a stony bank ('A' on RCAHMS 1956, plan, fig.96), which runs along the crest of the ridge from the WSW corner as far as the gap just mentioned, and thence NE for a further 90 ft. The bank itself is of insignificant size at the present time, but some light is shed on the original character of the rampart by the fallen debris which has accumulated along the foot of the ridge on the NW. This debris, a thick stony scree, suggests not only that the wall was of massive construction but also that its core was solidly packed with loose stones, in place of the normal filling of earth and stones, since a large proportion of the stones forming the scree are unsuitable for use as facings. That the gap in the ridge was used as an entrance to the fort is made probable by the fact that the ends of the rampart on either side of it are not in mutual alignment.

From the point 50 yds SW of the gap where a kink occurs in the rampart, a branch rampart (B on RCAHMS 1956 fig.96) is thrown off south-eastwards at right angles to the ridge. This branch represented partly by a bank and partly by a mass of loose stones which chokes the gully on the SE side of the ridge, can only be traced for a distance of 40 ft, but it seems likely that it originally extended across the summit area, on the line of the later dyke, dividing the interior of the fort into two enclosures of nearly equal size. The kink in the NW rampart can only be explained on the assumption that this cross-wall was an original feature of the fort, and it is possible, therefore, that the southern enclosure, which possesses the stronger natural defences and which could only be entered by way of the northern enclosure, constituted the citadel, while the northern enclosure was simply an annexe.

The weakest sides of the fort are on the N and NW; additional ramparts have been erected on these sides. Two of these (C and D on RCAHMS 1956 fig.96) are drawn across a broad gully on the W, and the way in which they have been sited so as to incorporate rock outcrops in their line is a notable constructional feature. The lower of the two (D on RCAHMS 1956 fig.96) is represented by a line of debris similar in character to the debris of the main rampart, while rampart C is simply a low turf-covered bank with only a little loose stone on the flanks. Both ramparts are pierced by a hollow track which leads up the centre of the gully and is bordered at one point by four upright stones; the position of the gap in the upper rampart (C) suggests that the track and the entrances that serve it are secondary and perhaps coeval with the hut circles in the space between the two ramparts. A third outer rampart (E) can be traced westwards for 50 yds from the NE end of the summit plateau to the edge of the quarry. Like rampart C it is a low stony bank displaying none of the massive scree which characterises ramparts A and D; its SW end may have terminated against the same bastion of rock to which C and D are connected. Certain features of these defences, notably the absence of ditches, the solid stone cores of the walls, and the tendency to use natural rock outcrops as an integral part of the defensive system, are reminiscent of Dark Age rather than Iron Age practice.

Secondary occupation of the hill is attested by seven hut circles, four of which - one blocking the main entrance to the fort and three lying entirely outside the defences on the W - are patently later in date than the fort while the remaining three, situated between ramparts C and D, resemble the others and are presumably contemporary with them. All the huts have been stone- built, but are reduced to foundation-level. At the extreme NE end of the summit plateau there can be traced the outline of a small oblong building measuring 19 ft by 15 ft over walls about 3 1/2 ft thick built of rubble in mortar. OS 6" 1917 marks this structure as a Watch Tower, but it is much smaller than any tower in the ordinary meaning of the term. The foundations of a second oblong structure of similar character, measuring 22 ft by 14 3/4 ft over walls 2 ft thick, appear about the centre of the SE edge of the summit plateau. The age of these two buildings is unknown, but they may have been connected with a Border look-out post of the 15th or 16th C. That the hill was occupied at the time of Flodden is shown by the mention, in a report written by Lord Dacre in 1513, of a "place called the Dungyon".

A cottage, of which the foundations can also be seen, seems to be of no antiquarian interest, and the lengths of stone dyke, one of which seems to be intended to fence off the lip of a steep slope, may or may not be connected with this phase of the history of the hill-top.

RCAHMS 1956, visited 1933 and 1939

The ramparts and hut circles of this Iron Age fort were excavated. The relationship between walls and huts is interesting for more than one period of construction is proved, some of it after the ramparts had fallen into disrepair.

Ministry of Works 1961.

Excavation continued for a second season. Work concentrated on a sub-circular structure revealed hut foundations but no evidence of post-holes. Fragments of pottery suggested occupation in the later prehistoric period, though the fort is Dark Age.

Ministry of Public Buildings and Works 1962; E V W Proudfoot 1962

This site has been almost completely quarried away. Only three hut circles, of the seven shown on the RCAHM plan, fig.96 on the W, and fragments of walling round the quarry edge survive.

Surveyed at 1:2500.

Visited by OS(RD) 7 February 1967

Further quarrying has removed the last vestiges of the rampart and one of the hut circles.

Visited by OS(JP) 10 September 1976

The excavation at The Dunion was undertaken between 1984 and 1987 by the Central Excavation Unit of the Scottish Development Department (Historic Buildings and Monuments).

The Dunion, a hill formed by the eroded stump of a volcanic intrusion and now much reduced by quarrying, is situated c 3km SW of Jedburgh. The surviving remains on the hill, originally thought to be an unenclosed platform settlement, proved to be part of a fort (NT61NW 3.00) which had formerly been considered to occupy only the summit of the hill. The fort was tentatively identified as being of Dark-Age date. Later fieldwork, and the discoveries made by the CEU, have shown that the fort was much larger than originally thought, and, with a total enclosed area of between 6 and 6.5 hectares, was of a size comparable with the minor oppida of the Tweed Basin (see NT61NW 3.01). A composite plan of The Dunion, based on many sources, suggests that the site started as a small fort on the summit, which later expanded in stages. The crude outermost 'rampart' on the NE side, and two houses outside the fort to the SW, suggested that the defensive nature of the settlement was of less importance towards the end of the occupation of the hill.

J Rideout 1984; J Rideout 1987.