RCAHMS Archaeological Survey of Swona

Date August 2014

Event ID 1000859

Category Project

Type Project

Permalink http://canmore.org.uk/event/1000859

2014 RCAHMS Special Survey

Introduction

Swona is a small uninhabited island in the Pentland Firth, part of the parish of South Ronaldsay, Orkney. It measures 2.2km from NE to SW by about 640m at the widest point and it comprises a higher central section dominated by two hills (Warbister and Keefa), separated by a small loch, and two low-lying headlands, one known as The Tarf , which lies at the SW end, and the other North Head. The island narrows before the headlands, in the north to only 140m, and much of the area of ground beyond each headland is dominated by pools of standing water.

There is early map evidence for settlement on Swona. Blaeu’s Atlas (1662-5, vol. 6) includes a map which depicts a churches or chapels and a single settlement (NLS: EMW.X.017). In contrast, the South Isles of Orkney, surveyed by Murdoch Mackenzie and published in 1750 (NLS: EMS.X.005), is probably the most useful of all the pre-OS mapping. It depicts two settlements on the W coast, most likely the township ND38SE 15 and the farmstead ND35SE 17. Mapping in an archaeological sense effectively began in the late 19th century with the depiction of the remains of St Peter’s Chapel on the 1st edition of the Ordnance Survey 25-inch map (Orkney 1881, Sheet CXXVI.1). This was followed by a cursory mention in the Royal Commission’s Inventory (1946, v2, 296). The Ordnance Survey’s Archaeology Division undertook a more through revision of archaeological records in 1973 and again in 1981, the later trip including chambered cairn expert Audrey Henshall and Orkney archaeologist Raymond Lamb. A brief survey of Swona was undertaken by G Geddes and A McCaig of RCAHMS on 21-22 August 2014, in conjunction with Dr K Strickland of Orkney College and Caroline Levey, Photographer. Known and new sites were recorded with differential GPS, notes and photography. Site descriptions were edited by JR Sherriff, RCAHMS.

Prehistory

The most obvious evidence of prehistoric activity on the island is the chambered cairn at the SW end. Although the S edge of the monument has been lost to an eroding cliff-edge and the surviving part has largely been robbed of its cairn material, enough survives of the chamber to identify it as an Orkney-Cromarty type tomb. A number of other cairns and mounds on both the E and W coasts of the island have at various times been put forward as either prehistoric burials or prehistoric settlements sites (i.e. individual houses). In most cases, however, the size, character and location of the features strongly suggests that they are probably burials of Bronze Age date (e.g. Canmore 9422, 9499). The exception is a group near Tray Geo which could be of relatively much more recent date (75268). One site has been positively identified as that of sailors’ graves (75267), but there are numerous records of shipwrecking, and it seems likely that other, similar, grave sites exist. Whether or not place names such as the ‘Crew of the Slade’ or the ‘Crew of the Quoy Dyke’, which appear on OS maps in the 20th century, are evidence for this is difficult to say.

The most likely candidate for a prehistoric settlement is the large mound that underlies the modern farmsteads and the medieval chapel at The Haven (9426). While the local report of the discovery of ‘Pictish Houses’ must be treated with caution, the mound does appear to be partially artificial in character.

Medieval

The site of St Peter’s Chapel (possibly founded in the 14th century) survives as a rectangular mound of stones and a fragment of enclosure (9421). While no grave markers have been identified, two possible fonts were located in other parts of the island. A possible second chapel, Twinly Kirk on the Tarf, is more equivocal in its character (9454). The suggestion that it is a chapel appears to have been based largely on its E-W orientation, but the presence of relatively high quality of masonry and a linear feature that isolates the surrounding headland may support this interpretation.

Post-medieval to modern

The last members of the Rosie family left the island in 1974, and the visible remains of human settlement are dominated by groups of roofed and roofless buildings, generally constructed after 1800. These represent the houses and associated farmsteads and kiln-barns of the farmers and fisherman who lived on the island through the 19th and 20th centuries, numbering seven households in 1881 (at Ayre, Middle Ayre, North House, and Upper and Lower Biggings). An earlier phase of settlement is represented by the buildings that were roofless and/or ruinous by the middle of the 19th century, particularly those that lie along the shoreline at The Redd (182454) and those that lie in the enclosure immediately W of the remains of St Peter’s Chapel. Other, perhaps slightly earlier, buildings survive as footings near the area known as The Rental. A rotary quern built into an enclosure near the chapel site also attests to earlier settlement (349761).

The island has been managed for both stock and arable, with the focus of the latter occurring in the northern half, nearest the settlement. Most of the cultivated ground has been heavily improved, drained and cultivated by tractor (two are still on the island) and plough, but there are at least three large areas of rig-and-furrow cultivation remains, one of which survives as an isolated patch farther south. A substantial earthen bank runs from NW to SE across Warbister Hill, presumably originally running to Dyke End, while another, perhaps earlier, extends NE to SW along the foot of Warbister Hill. These major dykes are supplemented by at least three others of different dates that delimit The Tarf, and two that delimit North Head, the latest iteration being a stone dyke of later 19th century date. These principal linear features are supplemented by smaller sections of dyke which prevent animals reaching the shore along the precipitous SE coastline (e.g. ND 3857 8378).

While the areas S of the main dykes have clearly been used for rough grazing, there is some evidence of areas in which the turf has been stripped for use as fuel, particularly on the SW slopes of Keefa Hill, where small mounds that may be peat-stack stands survive, as do ephemeral edges of turf-cutting. The islanders’ fishing activities are attested by references to a 19th century pier on the W side of the island and the surviving piers on the E side, built with the support of the Congested Districts Board in the early 20th century (NRS: AF42/3267). A large capstan was presumably part of the same project, as may have been an unusual boathouse, cut into the cliff. A boat, apparently the Hood which was built on the island, lies by the capstan. Elsewhere, numerous skeos (drystone drying sheds), can be found by the beaches at the N and S of the island, and there is little doubt these were used for the drying of fish (e.g. 9423, 9454). The skeos are found in close proximity to numerous enclosures, though this is probably just a coincidence resulting from the freely available building stone on the beach. The enclosures which are found in groups on both headlands and as isolated examples along the W coast, seem to be for sheep a well as cattle. At the Tarf, the largest pond has been dammed in two places, perhaps to ensure a ready supply of water for the animals.

The island is pock-marked by a large number of stone quarries, presumably exploiting the sandstone for buildings on the island and possible also for export (e.g. 345428, 347495). There are long open quarries at North Head, and at Warbister Hill, as well as a whole series of smaller pits on the summit ridge of Keefa Hill. Associated structures such as quarryman’s huts survive, as do a variety of cairns and walls built out of the quarried material. At Warbister Hill, the large mound previously identified as a possible settlement appears to overlie rig and furrow and is therefore probably a spoil heap (75265).

The Tarf is topped by a modern solar-powered navigational light built in 1983 (74481), but a circular concrete foundation on lower ground may represent an earlier navigational marker, possibly that built by Charles Stevenson in 1906 (349747). At North Head, a collection of features relate to navigation and quarrying (347162).

Information from HES Survey and Recording (GFG), August 2014.