South Uist, Bornish, Dun Vulan
Broch (Iron Age), Midden (Period Unassigned)
- Council Western Isles
- Parish South Uist
- Former Region Western Isles Islands Area
- Former District Western Isles
- Former County Inverness-shire
Dun Vulan, 150-50 bc Iron Age broch (or galleried dun) which recent archaeological excavation has revealed to have been a structure of some significance. Its position on the southern edge of the promontory (originally on a freshwater loch) is now protected by a newly constructed sea wall. The broch appears to have stood to some 10m in height, with internal chambers and galleries linked by stone stairways. Later Iron Age re-use raised the entrance about 2 m above the original. Just to the south lies evidence of two paved outbuildings of 2nd-4th century ad - the earliest known rectangular buildings in the Western Isles. After 400 ad a Pictish house was built within the broch walls. Nearby, an extensive 2nd or 3rd century midden has yielded significant finds.
Taken from "Western Seaboard: An Illustrated Architectural Guide", by Mary Miers, 2008. Published by the Rutland Press http://www.rias.org.uk
NF72NW 1 71407 29815.
(NF 7140 2980) Dun Vulan (NR)
OS 6"map, Inverness-shire, 2nd ed., (1904)
Dun Vulan: a dun, about 56ft diameter externally, NW - SE. The outer face of its wall is broken down in several places, and only the foundation course is traceable on the eastern arc. The inner portion of the wall remains about 5ft high, and a possible gallery is suggested in the NE. A later internal structure (18ft x 14ft 8 ins, an irregular oval in plan) is built towards the western side.
RCAHMS 1928; R W Feachem 1963.
Dun Vulan, a galleried dun, is as described above.
Resurveyed at 1/2500.
Visited by OS (WDJ) 7 May 1965.
Excavations at Dun Vulan revealed a sub-triangular stone tower with a diameter up to 19.5m, a surviving height of 4.5m, and a wall 4m thick. There is a wall chamber with a door to the interior, and a stairway to the first floor gallery. The wall chamber and gallery were filled with deep layers of soil containing only Iron Age pottery. An excavation 50m by 8.5m along the beach revealed deposits up to 4m deep. At sea level these are banded midden and sand layers. Above is a curved revetment wall which encloses a very large midden, created from house floor and hearth sweepings. Pottery, carbonised barley seeds, and a few shells occur but no bone or antler tools (found everywhere else on site). S of the midden wall sherds and bones are larger, dominant seeds are Chenopodium and Polygonum, and there are coprolites. There were also wooden artefacts (pointed tools and joinery) preserved in wet deposits. Structures were subsequently built on top of the revetment wall and E of it, including a paved area 3m wide and, last of all, a house of which the floor layer, hearth and wall partially survive.
The bones indicate exploitation of sheep, cattle, pig, deer, seal, whales, fish and birds, but not crab or lobster. Much smaller quantities of winkles and limpets were collected than at the nearby Kildonan wheelhouse.
Sponsors: HS, University of Sheffield.
M Parker Pearson 1992.
In 1993 work continued on the deposits under the modern beach outside Dun Vulan, noted as a broch by Pearson. A trench 60m by 8m was opened by machine. The large midden adjacent to the broch was excavated to its base. It provides a 2m-deep sequence of refuse deposits, probably deriving from inside the broch, ending in the 1st century AD and probably beginning soon after the broch's construction.
In front (E) of the broch is a large platform c20m by 30m in size and delineated by a low wall on its S and E sides. On the S side of this platform were constructed two rectangular stone buildings, the earliest around the 1st century AD and the second in the 2nd/3rd centuries AD. These were succeeded by a circular dwelling with hearth but were not themselves habitations. Each rectangular structure had a foundation trench (or 'drain') running under its threshold. The trench under the later rectangular building contained half a human mandible. Overall, the deposits outside the broch provide a stratified sequence of long-term occupation spanning at least 500 years.
A pilot field survey in the 7 square kilometres around the broch revealed the presences of thirty-two archaeological sites, most of them hitherto unknown. Those of Iron Age date, contemporary with occupation at the broch, include large settlement mounds which are spaced approximately 1km apart.
Sponsor: Historic Scotland
M Parker Pearson 1993
Nf 7140 2980 Excavations on the Iron Age broch and its accompanying settlement have been carried out since 1991 (Parker Pearson & Sharples 1995). The main purpose of excavation in June 1995 was to clear rubble from the inside of the broch and from its N side in order to enhance the monument's presentation for improved public interpretation. This was carried out in tandem with the first stage of the construction of a concrete sea defence. These aspects were funded primarily by grants from LEADER and from the TSB Foundation. Further work, funded by the University of Sheffield, was carried out on the deep deposits underneath the beach. These largely predated the construction of the broch and associated structures and some deposits were rich in organic remains, notably insects and wood. One of these deposits was full of animal bone and Late Bronze Age pottery, indicative of occupation on this site prior to the broch. Additionally, the sediments on the S side sloped against the angle of the beach and seem to include lake silts, suggesting that the broch was built within a freshwater or brackish loch, which was later encroached upon by the sea.
Within the sub-circular interior of the broch, an area of approximately 11m diameter, a sequence of later rebuildings was identified. The earliest of these was the construction of a cellular roundhouse with two doorways and an external room. This was modified by the construction of a straight wall within the roundhouse. Finally, a small fisherman's hut was constructed in the 1940s, covered by an upturned German lifeboat as a roof. The entrance to the broch was cleared down to the top of the roof lintels. Outside the broch, on its NE side, a rectangular building, probably medieval in date, had been constructed against the broch wall. One entrance led into the building from its S side next to the broch entrance. Its N wall was revetted into fallen masonry outside the broch's N side. The wall was retained but the fallen masonry was removed to a depth below the present roadway to provide a view of the broch's impressive walling.
M Parker Pearson, N Shaples and H Smith 1995.
NF 7140 2980, NF 7151 2983 Excavations on the Iron Age broch and its accompanying settlement have been carried out since 1991 (see unpublished reports, Dept Archaeology, Sheffield University). In June 1996 we continued conservation work around and inside the broch. No archaeological layers were removed but rubble was cleared from the S side of the interior to reveal the top of a semi-circular chamber which forms one of three rooms of a later building within the broch. This later building is not dated but its layout is very similar to the Pictish period cellular roundhouse inside the Loch na Berie broch on Lewis.
Outside the broch on its NE side, a building, probably medieval in date, had been constructed against the broch wall. This building was partially excavated to its uppermost floor layer in 1995 and we had hoped to locate its S wall. Unfortunately this was not found.
Sponsor: Sheffield University.
M Parker Pearson, N Sharples and H Smith 1996
NF 7140 2980 A short decorated and nail-headed hipped pin probably of 7th to 8th century AD date. Recovered from high in eroding midden deposits immediately E of the new sea wall built to protect Dun Vulan complex Atlantic roundhouse. Deposited with the National Museum of Scotland (NMS) pending formal allocation (Arch.DB.2000/106).
S Gilmour 2000
Change Of Classification (12 August 2010)
The site is a Complex Atlantic Roundhouse in the modern terminology and is now considered to meet the criteria to be classified as a Broch rather than a Dun. When initially examined by the Ordnance Survey it had not been excavated and the fact that surveyors were looking at the 1st floor of the site rather than the ground floor meant that the details of layout and organisation of the site were not appreciated at that time.
Information from Dr S M D Gilmour, 12 August 2010.