Small SwandroOrkney logo rgb 2nd-ver-5-8-17penny-logo

The Excavation

Summary of work so far.


A number of set upright stones just visible among the pebbles on the beach in 2010 indicated the presence of orthostatic settings. Subsequent excavation indicated archaeological survival on the beach below the erosion face that forms the boundary between land and high water. The presence of these deposits and their subsequent investigation has completely changed our understanding of this enigmatic mound. Initial clearance of the overlying beach material revealed the remains of an Iron Age structure. This was confirmed by an AMS radiocarbon date of 25BC-AD130 at 95% confidence for carbonized barley from a midden, which sealed flagging in one of the compartments. Work in 2012 enabled the nature of the erosion to be more fully understood indicating significant archaeological survival and potential. The sea had created terraces or steps within the archaeological mound, with each of these eroded scars being covered by re-deposited beach material.


In 2012 on the north western side of the cleared archaeological surface the remains of a substantial outer wall forming the arc of a large circular building seemed to form the continuation of a crescent shaped ridge at the top of the mound and it was thought at first to be the outer wall of a large roundhouse of broch proportions.  However the presence of a series of stepped concentric outer wall-faces containing a rubble core suggests that the mound represents a Neolithic chambered cairn.


Work in 2013 also concentrated on the continuation of the site south east of the mound. Investigation in 2014 has demonstrated a Pictish phase, indicated by cellular structures built within the infilled remains of more substantial Iron Age structures which themselves show there is a continuation of the site on the foreshore and under the boulder beach. The truncated remains of the Norse Hall clearly overlie this Pictish/Late Iron Age settlement.


Excavation at Swandro in 2014 also clearly indicated that the top of the mound forming the Neolithic Chambered Cairn had been partially robbed of stone in the Iron Age and infilled with Late Iron Age (Pictish) midden.


On the seaward area of the beach under the boulders the truncated building (Structure 1) was further investigated. Midden was found to continue to seaward but is clearly being affected by tidal action; deposits of midden located by coring in the intertidal zone in 2011 have now disappeared.


Work on the beach in 2014 concentrated on the excavation of the later Iron Age (Pictish) elements of the site. Excavation revealed a complexity of structural development with building forms found to be nested in earlier, larger structures.  The sea had partially destroyed both sets of buildings. The truncations were cleaned as sections, sampled and recorded.  The part-excavation of one of these later truncated buildings (Structure 2) in 2014 saw the sampling of floor surfaces down to the primary flag floor. The excavation and sampling of the infill of a third building form revealed the presence of slag and crucible material suggesting copper alloy working. A broken flagstone within a floor surface of one structure proved to be a capstone to a well. The well was accessed by steps and corbelled on three sides, with clay bonding present in the lower part; it is still filled by a freshwater spring.

The Knowe of Swandro looking north west down Eynhallow Sound.

The Knowe of Swandro looking north west down Eynhallow Sound. The top of the chambered tomb mound is visible on the right, with the person  standing on top of it.

Aerial photo showing the Iron Age spring fed corbelled well to the left, with further associated structures at the Knowe of Swandro, Rousay

Aerial photo showing the Iron Age spring fed corbelled well to the left, with further associated structures

The excavated remains of the Iron Age settlement and Neolithic Chambered Cairn at Swandro in 2015. The photograph was taken from a drone showing the Iron Age settlement and the chambered cairn to the right. Note the erosion of the seaward side of the cairn.

The surviving structural remains of the Iron Age settlement and on the right the casement walling of the chambered cairn

The surviving structural remains of the Iron Age settlement and on the right the casement walling of the chambered cairn after the beach overburden had been removed in 2015.

The Knowe of Swandro is in Rousay, Orkney (HY 3753 2966 ) and consists of a mound with obvious stone inclusions, which is situated immediately behind a boulder beach on the Bay of Swandro. On its eastern flank is the Norse settlement site known as Westness, excavated by the Norwegian archaeologist Sigrid Kaland in the 1970’s.  Described by RCAHMS in 1946 as ‘the much disturbed remains of a stony mound’, this knowe has generally been considered to be the remains of an Iron Age broch.  At the top of the mound a crescent-shaped wall or ridge faces towards the sea, which looked like the disturbed remains of a curving wall, surrounding an area which had large tumbled stones visible in the grass. Ordnance Survey records suggested it had been investigated at some point in the past but there is no published record.  The mound may have been disturbed during Radford’s investigation of the nearby Westness Norse houses in the 1950’s or 60’s.


As part of the Gateway to the Atlantic Project a number of coastal erosion sites were selected for investigation on the Island of Rousay. Due to the vulnerability of the remains at Swandro, work has concentrated on the investigation of this site. This research builds on the site and landscape studies undertaken at Tofts Ness, Sanday, Orkney and  Old Scatness, Jarlshof, and the Viking Unst project, all in Shetland by project directors Steve Dockrill & Julie Bond.

Excavation in 2015 continued to define the partially eroded structures on the beach and the excavation of the Pictish building (within the extension started in 2014). The passage to the chambered cairn was identified and the upper fill contained evidence of Viking Age activity with the finding of a coin of EANRED (King of Northumbria in the first half of the ninth century AD). Gareth Williams (in discussion of this find and within a Northern Isles context) was happy to see this as being a Viking Age deposition. The upper part of the beach containing the eroded buildings was recorded with both photogrammetry and 3D laser scanning.


The storm beach area containing the chambered cairn, which had been recorded in 2012, was re-examined in 2015 to identify and record the effect and damage caused by the sea since the site had been carefully covered and re-packed with beach cobbles. The tomb revetment demonstrated that heightened wave activity over the intervening winters had shifted deposits and revealed one of the lower revetment stones before clearance. Once uncovered, this area was cleaned and recorded using digital imaging and 3 dimensional scanning. It is worth noting that the cairn has suffered greatly from the effects of erosion in the intervening years, with much of the lower (seaward) circuit of the outer casement wall and the packing contained by it having been removed by the sea. Several of the large blocks from this lower revetment have been torn out and have completely disappeared. The water level at high tide regularly comes to this outer part of the tomb. The large stones that remain were angular when recorded in 2012 and now show significant smoothing by the action of the sea and movement of smaller beach material.

Observations on the Erosion


The terracing of the structures and associated deposits that survive on the beach is due to the unique combination of factors affecting the erosion and as such presents a rare opportunity to examine and sample such an extensive site and understand the erosion processes acting upon it.


The erosion is rapid; midden deposits found during an examination of the beach at low tide in 2011 have been completely destroyed by the sea. The sea is actively destroying the eastern part of the site under the earthwork remains of the Norse houses; stonework still survives but most of the sediments and midden deposits have been washed away and the front stones of the remaining features show battering and wear. The main part of the site survives in the shadow of the eroded remains of the chambered cairn and it is noticeable that midden deposits survive in this area. This earlier monument appears to be suffering badly from the sea on the beach itself. This is evidenced clearly by the Chambered Cairn when the results of the 2012 season are compared with those from the 2015 season. The assessment of the erosion and effective monitoring demonstrates the importance of developing recording methodology. The re-scanning of the monument using a 3D laser scanner before re-packing, has provided an important record enabling the monument to be visualised from different perspectives and dated to a point in time. The outer packing of the cairn’s casement walls was found to have suffered significant erosion since it was first recorded in 2012. A section in a trench 1.0m wide through the casement retaining wall and packing in order to investigate the structure of the cairn demonstrated that the packing had been badly eroded or replaced by washed in beach material.


On the beach to the east of the eroding Chambered Cairn, there are walls and features which appear to pre-date the Iron Age sequence and possibly represent an entrance passage to the tomb. These too are under threat; much of the lower terrace material has been destroyed and the remaining deposits are threatened. These middens are of vital importance as they contain an irreplaceable cultural and economic biography of the site


2016 excavation


Evaluative excavation continued upon the eroding beach at Swandro, Rousay, Orkney for a 4 week season in July 2016. The archaeology is suffering from erosion from the sea which has cut into a settlement mound (Iron Age to Norse) developed upon the eastern flank of a Neolithic chambered cairn. Previous seasons have established the presence of a sequence of structures which has been exposed by the sea cutting into the archaeology forming a series of terraces. The erosion has provided an opportunity to examine and sample this archaeological sequence. Unfortunately the deposits surviving at the lowest terrace have suffered from extensive erosion with much of the midden material having been washed away and the larger structural  stones having been smoothed by the movement of water and beach material.


The Chambered Cairn

The outer casement wall of the Neolithic chambered cairn is butted by a single faced alignment of stones, suggestive of a retaining wall. This was first observed in the 2012 season and again in 2015, when a much greater degree of erosion was noted. In 2016 this feature was investigated further in order to determine whether an old ground surface or underlying archaeology had survived the effects of the erosion by the sea. A number of large water worn boulders were found to be re-deposited, implying the movement of large 'storm thrown' rocks. These sealed a dark yellow sand and a black compacted sand containing decayed seaweed with no surviving evidence of any anthropogenic deposits or an in situ old ground surface. Despite this sequence of re-deposited material in the scoured area seawards of the cairn wall, the remains of a reddish ashy midden deposit was found to be sealed by the secondary retaining wall.


The entrance passage leading into the chambered cairn was also defined in 2015, upon the uppermost (landward) erosion terrace. The passage walls are single faced and the upper infill formed by a layer of small angular stone (shillet) containing copper alloy fragments, large fish and mammal bone appeared to reflect late activity. This was confirmed by the finding in post excavation of a coin of EANRED, King of Northumbria 810-840 AD, together with the near complete skeleton of a cat. This disturbance and infilling might represent Viking period activity. Work in 2016 continued to define the top of the passage and to assess the nature of this later activity. The further excavation of the passage revealed more faunal remains including of several sheep displaying metal butchery marks. This deposit sealed large angular rubble which appears to be the infill of the passage.


Structure 1: A truncated Mid Iron Age Roundhouse

This roundhouse is represented by just one segment of its circumferential cells, the interior and southern portion having been lost to the sea. The circumference of the building is formed by orthostats; the floor of the northern radial cell was formed by a single flag, which had been made to fit the cell. Several notches had been cut into the flag, which appear to be post settings. It seems likely that this would have supported a mezzanine level around the circumstance of the structure. The presence of such mezzanine structures is paralleled by Middle Iron Age remains at Old Scatness, Shetland.


Structure 2: A Late Iron Age Roundhouse

Definition of the upper eroded terrace of the beach identified the remains of what appeared to be half of a cell-like circular structure. In 2016, investigation indicated that an orthostatic divide with flagging either side which had been identified in 2015, were found to be clearly later elements forming a modification to the building. These were removed to reveal the original form of the building, the curved line of orthostats together with a door sill (threshold stone) indicating a western entrance. The seaward section demonstrated a greater degree of erosion and did not survive. A floor level was formed by large flags with evidence of orthostatic radial divisions. Entering this structure via the threshold stone, one of these radial orthostats blocks any turn to the left (north/landward side), confining movement to the centre or to the right. This phenomenon of barring entry to the left has also been observed within other Iron Age roundhouse structures in the Northern Isles at Old Scatness (Shetland). The flag floor and the hearth are clearly part of a sequence of floors representing several modifications to the building as the remains of part of a rectangular stone tank could be clearly identified under the flags.


Structure 3

Structure 3 is a cellular structure with structural features suggesting a Pictish date, mostly still sealed by the northern landward section. The southern wall was identified and excavated first and revealed evidence of an intramural cupboard, adjacent to a complete in situ cupboard. The contexts in the lower sequence under the rubble infill were found to contain some evidence of metal working with finds of slag, small crucibles and mould fragments together with evidence of fragments of copper alloy. A series of steps were found to lead from the north (landward) section into the building, in a curved passageway whose stones demonstrated wear consistent with rubbing caused by the passage of the past occupants. The presence of a threshold stone in the narrow passage, together with a bolt hole clearly indicates that there would have been a physical door dividing the passage and the central area of the structure.







The key research questions for the project are:


1.What is the extent of the Iron Age settlement and how does this change over time? The understanding of the Iron Age settlement sequence in cultural and economic terms by the excavation and sampling of these truncated archaeological surfaces will provide a current and informed understanding for people living on Rousay in the Iron Age and how this changes over time.


2.What is the stratigraphic association with the Norse settlement and how does this inform on the question of the Pictish/Viking cultural interface? The taking of existing estates by Scandinavian settlers is still a contentious issue in terms of its nature and date. Only with more detailed excavation will it be possible to gain an insight into this important transition on what increasingly seems to be a vital site for this transition period.


3.What is the potential of the Chambered Cairn in providing new data to complement the burial monuments excavated previously in Orkney? The site has the potential to establish the relationship of this monument form to the later Iron Age settlement, a phenomenon observed at a number of sites in Orkney, as well as providing an unique opportunity to investigate the construction of the mound due to the erosion.


4.The investigation of this eroding site takes place within a research framework, which also demonstrates the relevance of the disappearing record. The long settlement history or “biography” revealed by the erosion enables the study of human behaviour in this particular place through major changes in culture, climate and environment.


This would facilitate the following:


1. To provide an understanding of the erosion processes and the archaeological survival and to develop recording methods so as to inform future management of such sites.


2. A definition of the extent and nature of the archaeological survival (structural elements and sampling the in situ midden deposits) on the eroding beach with the sampling strategy informing on the economic and environmental exploitation and change, the site’s chronological development and its cultural biography within the archaeological record of Rousay and Orkney as a whole.

The  concentric walls of the chambered tomb at the Knowe of Swandro with the encroaching Atlantic Ocean

We have more information on the excavation including downloads of some of our preliminary reports. You can also read more on our dig diary pages


For more general Orkney background information please visit our page on 'Orkney's Archaeology'

Iron Age buildings in the foreground during the 2016 dig at Swandro