Monument record 454 - DUN EISTEAN, PORT OF NESS, LEWIS

NB56NW 1 535 650.

Dun Eistein is a flat-topped island, about 75 yards by 50 yards, separated from the mainland by a ravine 30-40ft broad and of similar depth. Along the edge of the ravine are the remains of a strong wall, through which, it is said, there were squints or loop-holes.
Towards the NE corner of the island are the ruins of a dun, sometimes called 'Tigh nan Arm' (the House of Arms), built of drystone masonry and reduced to a height of 4 1/2ft: externally it is a rectangle, 23ft x 18ft, but the central area is oval, 6 1/2ft x 4 1/4ft. There is no trace of a doorway and the entrance was probably on the first floor.
There are remains of houses on the island said to have been built by the Morrisons, about 1597, and on the S side is a ledge, 'Palla na Biorlinn', where they used to haul up their boat.
The 'tigh' appears to have been an incipient peel, but is probably 12th century.
F W L Thomas 1890.

Dun Eistean: in 1914 the wall along the ravine was 6ft wide and 1 1/2ft high. Behind it were the foundations of a group of small huts impinging on one another and an oval enclosure, 29ft x 13ft internally, containing a hut circle, 6 1/2ft in diameter, and three other small oval compartments.
The 'tigh' was described as 10ft high, lying ESE and 115 WNW. 35 yards south of it there was an artificial pond, banked on the E and excavated on the W, still showing moisture.
RCAHMS 1928.

Dun Eistean appears to be a settlement, probably medieval, comprising a complex of small rectangular stone built huts with rounded corners, now heavily turfed, and defended by the remains of a stone-built wall c. 1.8m wide and c. 0.5m high, which runs along the top of the ravine. The pond, now dry, has been created by damming the E end of a small hollow. The oval enclosure described by the Commission is actually a rectangular hut with rounded ends, with settings of stones within, suggesting oval compartments. This is one of apparently three main huts. There is nothing to suggest monastic associations. ???
The 'tigh' is an oval, turf-covered stony mound c.13.5m by c.10.5m, and 2.2m maximum height. Many of the stones are small rectangular blocks, suitable for building, but there is no trace of built walling. An excavation in the centre, probably for material to build the modern marker cairn on the west side of the mound, has revealed a content of earth and stones only. Impossible to classify.
The name 'Tigh nan Arm' is not known locally.
Surveyed at 1/2500.
Visited by OS (R L) 16 June 1969.

Eroding along the eastern side. Burgess & Church 1997

'The monument...comprises of a fortified settlement, probably occupied during the Iron Age and later periods. The area to be scheduled is irregular measuring a maximum of 140m E-W by 90m N-S, to include the entire island.'
HS Scheduling Document 19 March 1992.

Archaeological and topographical survey, including description of the archaeological structures on the island, undertaken by Chris Barrowman and a team from GUARD (Glasgow University) at the end of May 2000, sponsored by the Clan Morrison Society. The island was accessed at low tide and the team camped on the site for the duration of the survey. Eight groups of structures were recorded, Structures A to H, and a possible boat landing place/noost 'Palla na Biorlinn'. The immediate environs on the mainland opposite the island were surveyed in April 2000 and four features, I to L were identified, amongst an extensive system of well-preserved lazy beds.

Structure A, the best preserved of the ruins, consists of two rectangular elements linked by a common W wall. The primary element, A1, is a dry-stone construction rougly rectangular in plan, 4 x 2.5 m, with walls approx 1 m wide and 0.25 m high. A possible entrance is located in the middle of the E wall. A2 lies to the S and is less well-poreserved. It appears to be a completely separate building from A1, and the N wall of A2. The walls stand 1 m wide and 0.25 m high. A stretch of wall extending on the same axis as the W wall of A, extends about 15 m to the N before turning E and petering out. This wall appears to be part also of the perimter wall of the site (structure H) and stands 1 m wide and 0.25 m high. This appears to be a post-medieval dwelling house.

Structure B occupies the extreme SW of the island hard on the edge of the cliff at the narrowest point across the ravine. It is an irregular conglomeration of six dry-stone cells, which share a common S wall (this may be also an extension of Structure H). The largest cell, B1, measures 5 x 3 m and its interior is almost circular. It is set apart to the W of a tight cluster of five roughly oval cells, which measure from between 2-5 m in internal diameter. The walls are approximately 0.5 m wide and 0.5 m high. Bi is more substantial, walls are 1 m wide and 0.5 m high with a possible entrance to the NW. This may have served as a 'gatehouse' to the island.

Structure C is a substantial sub-rectangular building lying parallel to Structure A and is set back from the S cliff edge. Overall it measures 11 x 6 m and its relatively well-preserved walls stand up to 1.5 m in places. It has three or four internal divisions, but this may be tumble. Neither of the two small cells to the N (C1 and C2) have obviousl entrances and only measure c 2 m x 2 m and c 1 m x 2 m respectively. C3 is situated right in the middle of the building and is similarly small and may be an artefact of collapsing walls. C4 is larger, 3 x 2.5 m, and more regular. It appears to have an entrance from the E and a linking passage leading towards the three small compartments. The curving end walls and slightly bow-shaped side walls are suggestive of the Norse architectural tradition, but the degree of collapse makes it difficult to give a more accurate impression other than it looks like a dwelling house. The degree of preservation suggests it was occupied later than B or D, perhaps at the same time as A.

Structure D is the largest and most complicated on the island. It is similar to B, being composed of small cells and set against a wall running along the cliff edge. The form of the cells is difficult to discern due to the clumpy grass and tumble. However it is possible to identify at least 15 individual cells. The W side of the complex is better preserved. Overall the structure measures approx 30 x 12 m and the walls generally stand to between 0.25 and 2.5 m internally. There is no obvious entrance and Structure H forms its S wall. On the W there is a circular cell, D1, which is measures x 2.5 x 2.5 m and does not share any common walls with the other cells. It has no obvious entrance and structure H forms its S wall. Cells 2 and 3 to the W of the structure are rectilinear in plan and open ended to the W. They share a central wall. N of these lies a larger cell, D4, 3 x 3 m, which appears to have an additional element, 2 x 1 m, adjoining it to the NE. Cells D5 - 8 run round the N edge of the structure and have internal diameters of between 2 x 2 m and 4 x 3 m. D14 is of a similar size and is situated in the middle towards the E. A long rectilinear, 6 x 2 m, cell is situated on the S side of the structure (D9) running adjacent to structure H. This is the best preserved cell and suggests that the other cells may originally have taken this form. D10 - 13 are situated in the middle of the structure and have rectilinear elements. The larges of thesem D12, measures 5 x 3 m. At the E end of the complex is a tiny, open-ended cell, D15, measuring 2.2 by 1.4 m, which is formed by a stretch of wall jutting out from the main structure and the perimeter wall. This complex is presumably the main residential component of the site, but it is particularly difficult to interpret due to the degree of tumble and vegetation.

Structure E is an artificial pond dug into the turf and contained by a low earthen bank in the centre of the island. At present the pond is circular and measures 7 m diameter and approx 0.3 m deep. Judging from the bank and the extent of marshy ground around it, the original size may have been c 7 x 20 m. A channel or runnel cuts through the E edge, presumably to allow higher levels of water to drain away. Overspill has formed irregular terraces and a marshier area below the pond to the E. It was presumable built to provide fresh water for the occupants of the island.

Structure F is two slight hollows which have been quarried into the foot of the slope occupied by the dun (Structure G). These measure 4 x 2 m and 3.5 x 2 m respectively and lie adjacent to one another. An upright slab of rock is situated to the back (N) of the E hollow (F2). Directly to the NE of these hollows is a natural gully Palla na Biorlinn (Ledge of the galley) and was used to haul boats up onto the island. It is possible that F1 and F2 were used as boat noosts.

Structure G is an irreglar mass of masonry situated on the highest part of the island. It dominates not only the stac itself but can also be seen for considerable distances along the north Lewis coast. The structure stands on a natural terrace which runs around the southern side of the dun and extends approximately 3 m beyond the edge of the rubble spread. On the N side the structure stands on the edge of a steep slope which drops onto the cliff edge. It is now very ruinous and the substantial amount of tumble, expecially at the E end, obscures all structural detail. Nevertheless, it seems likely that this represents the robbed base of a dry-stone fortification. From the spread of the rubble it appears that the extent of the structure was about 8 x 15 m. A smaller walled structure G2 has been built on top of the rubble. It's walls stand approximately 1 m high and are about 1 m wide, defining a space c 2 m x 2 m. The original form of this structure has been modified through the construction of the modern cairn, G3. Capt Thomas reports in 1890 that the interior was a small oval construction without a door. Allowing for the disturbance by the construction of G3, this is what survives today. It is not clear whether the building described in the 19th c as the Tigh nam Arm, 'House of Arms', referred to the whole mound or the secondary building, G2. Even if roofed, G2 seems most likely to have been an outbuilding, presumably it could have served as some sort of armoury.

Structure H is a low wall which follows the cliff edge on the landward, E and W sides of the island. Generally its construction is obscured by the grass. It appears to be of coursed rubble construction, but no sign of mortar could be seen. The wall stands 0.7 m high and 1.5 m wide and is not accompanied by large amounts of tumble.

Structure I consists of fragments of two concentric banks with a slight internal ditch on the cliff edge opposite the island. It is heavily eroded, 1 m wide by 0.2 m deep, and defines an area with an external diamter of 12 m. This may be natural feature or sheep tracks.

Structure J is a low mound, 0.9 m high, occupying a slight enimence and is well-defined on the S side, opposite the island. The summit is defined by a slight bank with a diameter of c 6 m and a max height of 1.5 m. This sits on a platform with a diameter of about 16 m. The lazy beds to the NW of the feature lap over the edge of the mound, indicating that the feature predates the cultivation. It could be a natural mound.

Structure K is a low mound 5 x 3 m by 0.5 m high at the head of a long cultivation ridge. It may be due to stone clearance.

Structure L is similar, measuring 4 x 4 m by 0.5 m high.
Barrowman and Driscoll, 2000

Geophysical (resistivity and gradiometer) survey across the whole island and four small trial excavations were undertaken on Dun Eistean in May and June 2001. A number of interesting anomalies were revealed by the geophysics, especially around the main structure G, which as been described as a dun. These anomalies suggest that t rock-cut foundation trench and a robbed-out wall are present running around the southern edges of the structure. Further anomalies were recorded around the rectangular building, which lies at the western side of the island (Structure A). It is possible that this represents a large area of burning, suggesting a working or industrial area. These findings increase the potential of the island's archaeology, demonstrating that there are buried features relating to known upstanding remains.

The subsequent trial excavations were situated on Structure G (Trench 1), over the geophysical anomaly on the plateau to the west of Structure G (Trench 2), over cell D12 and the perimeter wall (Structure H), and over cell A1 of Structure A (Trench 4).

In Trench 1 (2 x 8 m) a well-preserved clay-bonded wall of a medieval rectangular building, possibly a keep or tower, was revealed beneath the rubble of what had previously been interpreted as a dun. This wall stands to a height of approx 1.5 m. A stone roof tile was also recovered. It is still not known whether an earlier structure is present underneath this building, although the results from the geophysical survey have led to the speculation that a robbed out wall is present running around the circumference of the existing structure. An alternative suggestion has been made that this possible rock cut trench may be the result of the quarrying of stone to build the structure itself. Eight courses were excavated of the wall across a 1 m face. Although the courses are uneven, they are consistent, and built with large stones, 0.15 - 0.5 m long by 0.1 to 0.3 m high, with smaller stones filling the irregular gaps. The external measurements of the building are now clearer- externally approx 7 x 4.5 m, with the long axis aligned E - W. The internal measurements are approx 2 x 1.5 m, although not all of the internal faces are located. The height of the building is difficult to guage. Judging by the fallen masonry, and given that a large proportion of this will have been lost over the cliff on the north side, it is possible to speculate that the building was at least twice the height of its present state, ie. Approx 4 m tall. An occupational layer ws recovered outside the builiding containing six sherds of handmade undiagnostic pottery. Frequent flecks of charcoal also show that burning occurred in the area. Three stages of collapse sealed this layer. The internal deposits suggest a more recent history of use and reuse. The lack of an inner face to the wall, the presence of a cut and a variety of fills point to this hypothesis. The interior of the building is an obvious choice for any reuse as a temporary shelter and the stone from the inner face has also been used for the construction of the modern cairn.

Trench 2 (3 x 5 m) revealed no archaeological remains and the geophysical anomaly was revealed to be a natural, geological fault.

Trench 3 (4 x 2m) confirmed the relationship between one cell within the large group of cells, Structure D, and the perimeter wall, Structure H, and also the nature and depth of the archaeology. The perimter wall is constructed mainly of turf and had a stone facing. This would have been incorporated into the wall at the same time as the turf was laid, and the full height of this structure may have been at least twice its current height. The cells of Structure D were built up against this wall. The earliest phase recorded suggests that turf was dug out from the area of the cell, and either used for the construction of the main wall, or the side walls of the cell. These were also built of turf, with no stone. The side walls abutted the external wall. It is possible that some form of facing to the external wall had been constructed, possibly of wood. After removal of the turf, and the constrction of the walls, the area may have been levelled to create a suitable floor. Pottery was scattered across the full extent of the area, an occupation deposit is recorded, but no hearth. The abandonment of the cell is evidenced by the slumping of the external wall over the floor deposits.

Trench 4 (4 x 2 m) was limited by time, so only main structural components were revealed. The external wall of the building was built of stone, with a possible core of turf. A small stone partition divided the main room of the building. The majority of the pottery came from a deposit overlying the original ground surface outside the building. This appears to be a midden deposit. Occasional charcoal flecks from the internal floor deposits suggest that there was a hearth inside the building, although it was not located.

C Barrowman 2002 (SWE41041)
See also C Barrowman 2001, 99 (SWE 41185)

…described the rescue excavation of two main areas at Dun Eistean, Near the But of Lewis, which took place in June 2002. The excavations were undertaken to fulfill plannning and Scheduled Monument Consent conditions set prior to the construction of a steel footbridge across to the island. The two trenches were located on either side of the chasm which divides the island from the mainland of Lewis.

The island trench produced many artefacts, as well as previously unrecorded structural deposits. This raises questions about the interpretation of the results of the geophysical survey conducted in 2001, and suggests that there may be more structures that cannot be identified as topographical features or geophysical signals.

Over 300 sherds of local hand-made pottery were recovered fromr the island trench (1). Many of these were from the slumped layers of the external wall which runs around the landward side of the island. On excavation this was found to consist of an external and internal stone facing, with a turf and earth core. The footings rested on carefully prepared foundations: the exterior face rested upon a quarried terrace of bedrock; the interior face rested on small stones and gravel set into the subsoil. The nature of these deposits suggested a rebuilding of the wall at some stage in the past.

A second wall was also found in this trench, toward the island's interior. This again consisted for half its length fo stone facings and a turf core, and ran across the trench from east to west. It also had a gravel foundation. A break in the wall measuring half a metre wide coincided with a post hole, in which there remained waterlogged deposits of wood as well as some pottery. It seems tha the wall had a timber post set into it. Further pottery sherds and lithic impliments were recovered from within and adjacent to this structure.

An occupation floor was excavated between the two walls, and again thi held many pottery sherds, pieces of wood, corroded iron and lithics, along with several unworked beach pebbles. The most significant find was a lead pistol or gun shor, measuring 9-10mm in diameter, with a sprue running around the circumference. This had not been fired, and may be contemporary with the structures, although it was recovered from the topsoil

Trench 2 on the mainland was not as productive, although some pottery, a lithic artefact and two possible features were recorded.

The area stripped for the new access road to the bridge was also examined, but it was difficult to detect any archaeology as the area had been stripped with a toothed bucket. No features were revealed, although two prehistoric stone tools were recovered.

A possible defensive wall on the mainland was also examined. This ran between two inlets and would have defended the peninsula adjacent to the Dun. The opportunity to examine this feature arose because the access road passed through a denuded part of it. It was composed of agricultural soils typical of lazybeds in the immediate vicinity. No ditches were evident, although the secontion of the feature examined was of poor quality.

C. Barrowman 2002 b (SWE41056)
C. Barrowman 2002 c (SWE41091)

Excavations in 2005 revealed that structure A consisted of two late- or post-mediaeval domestic buildings. It also clarified that structure G, the tower, is constructed on a natural rock platform, and located the internal face of the tower wall, giving a wall thickness of c. 2m. Both domestic buildings and tower appear not to have been in use before the 16th century.
Barrowman 2006

Thomas, F. W. L., 1890, Archaeol Scot (Bibliographic reference). SWE7756.

Thomas, F. W. L., 1890, Archaeol Scot, 365-9 (Bibliographic reference). SWE7756.

RCAHMS, 1928, The RCAHMCS 9th Report & Inventory: Outer Hebrides, Skye and the Small Isles, 7, No. 15 (Bibliographic reference). SWE5587.

Burgess, Christopher & Church, Michael, 1997, Coastal Erosion Assessment, Lewis. A Report for Historic Scotland (Bibliographic reference). SWE41000.

Barrowman, C and Driscoll, S, 2000, Dun Eistean, Lewis: Archaeological and topographical survey - description of the archaeological structures (Bibliographic reference). SWE41040.

Council for Scottish Archaeology, 2001, Discovery and Escavation in Scotland (Bibliographic reference). SWE41185.

Council for Scottish Archaeology, 2002, Discovery and Excavation in Scotland (Bibliographic reference). SWE41091.

Chris Barrowman, 2002, Dun Eistean Excavations 2002: Bridge Footings (Unpublished document). SWE41056.

Chris Barrowman, 2002, Dun Eistean: Geophysical survey and trial excavation (Bibliographic reference). SWE41041.

Rachel Barrowman, 2006, Dun Eistean Archaeology Project Excavations 2006: Data Structure Report (Unpublished document). SWE41159.

Rachel Barrowman, 2006, Dun Eistein Archaeology Project Excavations 2005; Data Structure Report (Unpublished document). SWE41093.