St Kilda, Hirta, Ruaival

NF09NE 15 0984 9852

At Ruaival an excavation trench was established over two of the apparently circular/sub-circular stone scatters, situated just to the W of the most northerly of the four extent drystone enclosures (Enclosure 1). These were noted and recorded, and soil samples taken in 1992 and seen to be associated with a system of earlier, turf-covered banks and terrraces that clearly underlie the surviving drystone enclosures.

Excavation suggests that the circular/sub-circular sites were of coarse rubble construction, much of material of which had collapsed into and downhill of the structures. After removal of the collapse, both structures were seen to contain remnants of walling, possibly only an inner face, delimiting a sub-circular central 'activity' area. In the most northerly of the two structures, the outlines of what appear to be traces of an 'entrance passage' were beginning to emerge at the close of the excavation. No material that would provide chronological or cultural indicators were recovered, but some fragments of what appear to be rudimentary bowls or mortars were found.

A Morrison 1992b; A Morrison, P Johnson and A Pollard 1993

At Ruaival (centered on NF 098 985), excavation of two circular structures was completed this summer. One of the biggest problems was the attempt to distinguish the structures from the scree, since most of the main stone material for the structures was probably collected from the vicinity. It was eventually possible to see traces of facing stones and arrangements that suggest that the structures were built and are not, for example, simply the remains of quarry pits. Apart from the remains of a few broken mortar-like stone vessels in the structural material there was nothing that could give definite evidence as to function or dating.

This season also saw the first examination of the earthen platforms at Ruaival. These lie within and under the more recent drystone enclosures and show similarities to the terraces on An Lag. Unlike An Lag, Ruvial is freely drained. Because of this, little trace of stratigraphy (eg tip lines) could be found in the Ruaival platforms. No traces of agricultural, horticultural or domestic activity were discovered and no finds of any chronological significance. Despite the missing evidence for, eg, pottery or tip lines, as found on An Lag, the results of the Ruaival investigations suggest that here, too, a series of artificial earthen platforms has been constructed for growing purposes.

During this season's field work, a built structure in the scree, higher up on the slopes of Mullach Sgar to the N of Ruaival, had been noted. This structure is not visible from lower down the slope and is not easy to see even from a few feet away. The remains stand to a height of well over 1m; it is drystone built, with traces of corbelled construction and there is a small kidney-shaped chamber at one end, plus a small niche built into one wall and a drain-like feature running through the other (see Fig 38, p106). There is what appears to be a blind passage leading out of the chamber for a distance of about 2m and there is no apparent entrance to the structure. One large roofing lintel survives in situ across the passage, with what might be the broken remains of another close to it. The method of construction is interesting. It appears that a level area had been quarried back into the slope of the scree had been replaced around the structure to support but also possibly to conceal it. As a result of this work further examination of the general area was undertaken and it became apparent that there are many examples of walls, enclosures, structures and other features.

Sponsor: National Trust for Scotland.

A Morrison, P Johnson and J Huntly 1995.

NF 098 985 Work at Ruaival consisted of a final sampling of the terrace-like structures in Enclosure 1, mentioned in earlier reports, in an attempt to identify possible variations in form and possible function. It became obvious that there was a much greater quantity of stone involved in the construction of the edges of some of those terraces located closer to the cliff edge than had been discovered elsewhere. Preliminary interpretation suggested that the stones appear to have served as a possible supporting wall or 'revetment', with such supporting edges being stronger and thicker where the ground beyond slopes away. The terraces have undoubted similarities to those on An Lag and, despite the scarcity of evidence, an agricultural or perhaps horticultural function still seems the most likely interpretation of these structures.

Last year's work on the scree (Morrison, Johnson and Huntly 1995) extended into a much wider area. Different forms were identified, including circular foundations, some with traces of corbelling still surviving,

elongated forms such as the structure planned last year, and small circular enclosures without visible openings. As noted last year, it appears that a level area had been created for some of the structures.

Various environmental samples were taken from locations in Gleann Mor, in preparation for survey and fieldwork in that area.

Sponsor: Historic Scotland, National Trust for Scotland, Universities of Glasgow and Durham.

A Morrison and P G Johnson 1996

NF 098 985 (centre) Several further scree structures at the base of Mullach Sgar (Morrison and Johnson 1996) were investigated. Excavation showed the constructional sequence: spaces were hollowed out within the scree; a horizontal platform was made; then the walls of the structure were built and the whole structure was concealed in the scree.

Structure SS1 was cigar-shaped, with evidence of two possible entrances, both blocked. An area of burnt material is not thought to represent an in situ hearth. Very fragmentary sherds of thin-walled coarse pottery were recovered, but no floor or activity surface was identified. The destruction levels of SS1 contained a single coarse stone implement.

The corbelled dome of structure SS4 protruded slightly above ground level. This was a sub-circular domed structure, probably a classic 'beehive', with a readily identifiable entrance surmounted by a substantial lintel and a small drainage aperture adjacent to, but at a lower level than the entrance. A deliberately laid floor of flattish dolerite blocks was found, above a layer of dressed stone fragments bonded in clay. Four coarse stone implements were discovered in the foundation layers.

Finds from structure SS11 suggest a more recent date, perhaps post-medieval. Structure SS14 survived as two sections of curvilinear corbelled walling at the back of a degraded platform.

Sponsors: Historic Scotland, National trust for Scotland, Weston Trust, Universities of Glasgow and Durham.

P G Johnson and J Huntley 1997

NF 098 985 (centre) The 1998 season of excavation focused upon a series of terraces located on the lower slopes of Mullach Sgar, near Clash na Bearnach or 'the Chimney' on the W side of Village Bay. A major aim of the four week project was to evaluate the character of a drystone ?horned enclosure? which had been identified in previous surveys. It was thought that this structure might bear some similarity in form and function to those located in Gleann Mor to the N of the Village Bay area. It is known from contemporary accounts and photographs that these features were used as lambing pens and shieling sites into the 19th century. It is, however, possible that the surviving features were built on top of much earlier structures which may date back as far as the prehistoric period. A series of trenches was centred on the enclosure and its environs. A long trench cut across the upper terrace, upon which the enclosure is situated, revealed a deposit of scree buried beneath humic topsoil. This buried scree contrasts with the large areas of exposed mobile scree which rises up from the rear of the terrace. The scree terrace had been modified, prior to the formation of the topsoil, with a bank created across the western edge of the upper terrace, thus accentuating its separation from the lower terrace. The denuded remains of a drystone wall could also be seen running across the crest of the bank.

Trenches inside and immediately outside the enclosure revealed some degree of structural complexity and temporal depth. A possible wall footing was identified within rubble deposits inside the enclosure and a number of coarse, undiagnostic pot sherds were recovered from a possible occupation surface beneath the rubble (it is hoped to obtain a TL date from these sherds). Outside the enclosure a substantial curving wall was located, which appeared to pre-date the construction of the previously described terrace bank. Both of these features, the wall inside and outside the enclosure, do not appear to relate to the upstanding structure and are more likely to relate to the earlier use of the terrace.

Geophysical survey was carried out across the larger, lower terrace, in an effort to locate remains which may relate to the lost site of St Columba's Chapel, which had previously been thought to exist within the village head dyke. Trial trenching of anomalies resulted in the identification of two shallow, rectilinear stone-lined cuts. Although no human remains were found they may relate to graves, and as such point to the ecclesiastical use of the terrace. The excavation of a corbelled, cell-like structure at the rear of the lower terrace was also begun, but time did not permit anything more than clearance of rubble in an attempt to define the structure.

Work also continued on the scree structures located on the higher slopes behind the terraces (Johnson and Huntley 1997). This involved the backfilling of structures previously investigated and the completion of the excavation of Scree Structure 4. Work outside this circular, corbelled structure provided an insight into its construction. A platform had first been created on the scree by removing stone, the structure was then built on this platform. In order to stabilise the structure, scree material was then mounded up against the outer walls, a technique which also served to obscure the location of the cell. It is possible that at least some scree structures may also relate to the ecclesiastical use of the area, bearing some similarity to Irish monk's cells.

Work on the palaeoenvironmental side of the project also continued in 1998, with soil samples taken from profiles across the terraces.

Sponsors: National Trust for Scotland, Universities of Glasgow and Durham.

T Pollard and P Johnson 1998

NF 098 985 (centre) Excavation continued on a series of terraces located on the lower slopes of Mullach Sgar, near Clash na Bearnach or 'the Chimney' on the W side of Village Bay (Pollard and Johnson 1998). Trenches first examined in 1998 were reopened and extended during this second season of work. This exercise confirmed that the horned enclosure did overlie a series of earlier structures, which included a midden deposit located between two earlier walls. A thick layer of heather charcoal underlay the midden material and has been sampled for radiocarbon dating. Although the full character of these earlier structures has yet to be clarified there can be little doubt that the site has been a focus for human activity over a considerable period of time.

It now seems clear that the horned structure is much simpler in form than those in Gleann Mor, lacking the various cells and chambers, although the discovery of an iron shepherd's crook does confirm its use as a sheep pen.

Further examination of the lower terrace and features previously suggested to be graves established that these were in fact natural features related to the exposure and weathering of bedrock. Although the location of St Columba's Church on the site now seems less likely, work will continue on these complex features in 2000. (GUARD 702).

Sponsors: National Trust for Scotland, GUARD, Peter Stormonth Darling Charitable Trust.

T Pollard and B Will 1999

NF 098 985 (centre) Excavation continued in summer 2000 on a series of terraces located on the lower slopes of Mullach Sgar, near Clash na Bearnach or 'the Chimney' on the W side of Village Bay, Hirta. Trenches first examined in 1998 and 1999 (DES 1998, 99-100; DES 1999, 90; NMRS NF09NE 15) were reopened and extended during this third season of work. This exercise confirmed that the horned enclosure does overlay a series of earlier structures and that these extend beyond the limits of the horned structure.

Excavation within the horned structure recovered prehistoric pottery and stone tools. Excavation within Structure 3, to the E, revealed the remains of a corbelled stone-built structure that had been built into the hillside and survived to a height of 1.5m. Excavation within the structure uncovered a large amount of collapsed building stone and midden material containing a large amount of prehistoric pottery and stone tools - probably ard tips.

The pottery is believed to date to the middle Iron Age. This is the first time that pottery and stone tools have been recovered from secure archaeological contexts from St Kilda. (GUARD 855).

Sponsor: National Trust for Scotland.

J Duncan and R Will 2001

NF 098 985 (centre) Excavation continued on a series of terraces located on the lower slopes of Mullach Sgar, near Clash na Bearnach or 'the Chimney' on the W side of Village Bay. Trenches examined previously were re-opened and extended during this fourth season of work (DES 2001, 101). Further excavation within a semi-subterranean structure (3) uncovered more collapsed building stone and midden material containing prehistoric pottery and stone tools, in addition to an internal wall divide and partial blocking of the entrance. Excavations within the area between the horned structure and structure 3 revealed the presence of a linear wall which may relate to other structural elements uncovered in 1999. A trench was also excavated across the stone bank which surrounds the terrace on the E side. (GUARD 1182)

Sponsors: NTS, Scottish Heritage USA

J S Duncan and B Will 2002

NF 098 985 (centre) This year saw the conclusion of five years of excavation on the site (see DES 2002, 120), carried out in June and July 2003 by NTS work party volunteers under professional supervision. Excavation focused on the area outside the entrance to a corbelled semi-subterranean structure (3) uncovered in previous seasons, the occupation of which has been radiocarbon-dated to the last two centuries BC and the first two centuries AD. Its entrance-way proved to be stepped and revetted with stone.

Excavation immediately W of this structure further investigated a midden, radiocarbon-dated to the 10th to 13th centuries AD. The midden was found to fill a small, bean-shaped structure (4), which had been constructed using an earlier wall of possible prehistoric date. The area between these two structures contained prehistoric walling and occupation deposits that produced prehistoric pottery and stone tools. The wall of the adjacent horned enclosure was found to overlie the Norse/medieval midden, dating its construction to after the 13th century.

Report to be lodged with Western Isles SMR and the NMRS.

Sponsor: NTS.

L Innes, O Lelong, I Shearer and R Will 2003