Lower Caen

ND 01359 17834 A community excavation, planned to commemorate the bicentenary of the Kildonan Clearances, was undertaken 8–23 June 2013. The deserted township of Caen is located in the Strath of Kildonan, c2 miles upstream from the village of Helmsdale. Lower Caen is located on a river terrace above the Caen Burn. It is distinct from Upper Caen, which comprises a collection of ruins that extend N, up Caen Burn.

A programme of topographical mapping and geophysical survey was carried out. As the structural remains were clearly visible the fluxgate gradiometer survey aimed to identify the location of hearths, middens and other possible internal features.Unfortunately, the underlying granite masked not only more subtle responses but even that from a large corn-drying kiln. As a result it was only possible to identify features already visible, such as extant walls and ‘potato’ pits. Meanwhile, topographic information was shared with a team from the University of St Andrews ‘Virtual Worlds’ group in order to develop an immersive ‘Virtual Caen’ (now available at: http://openvirtualworlds.org/unity-static/web-Timespan/web-Timespan.html).

The first season of excavations at Lower Caen focused upon the final phase of occupation and subsequent abandonment of a single longhouse (Structure 1) and its associated outbuilding (Structure 2). Two L-shaped trenches were laid out to provide a half-section across both structures along the N–S (Trench B) and E–W (Trench A) axes.

The removal of turf exposed rubble spreads across Structure 1 and across the E half of Structure 2. At the W end of Structure 2, removal of the turf exposed a compressed earth surface. Removal of the rubble across Structure 1 (the longhouse) exposed a high quality cobbled surface in the S end of the longhouse (upslope and upwind). This surface was constructed of rounded river pebbles and had been finished with a concreted lime treatment. At the N end of Trench A (within Structure 1) the end of a sub-rectangular void was found in this flooring, and was interpreted as potentially representing the laying of the flooring around a large rectangular item of furniture – such as a box-bed or similar. We recovered a significant number of sherds of post-medieval glazed wares (18th- and early 19th-century in date, predominantly local wares but including some English wares) lying directly on top of this floor surface. This included one low profile bowl which had clearly shattered in situ with a complete half of the vessel recovered (the other half appears to lie under the baulk). In addition to the ceramics a considerable number of fragments of glass, iron objects (all badly corroded) and two fragments of leather were recovered.

Near the centre of Structure 1, at the S end of Trench B, a hearth was uncovered. Lying within a beaten earth floor, the hearth was made up of flat stones with a shallow depression at its centre. No accompanying features were identified, the central baulk ran immediately to the W and S of the hearth.

A single doorway was identified in the E-facing wall of the longhouse, with an accompanying internal and external area of rough paving. No threshold stone was present, although remains of a timber beam running across the entry at floor level were recovered from sockets on either side of the doorway, along with a substantial iron bolt and badly corroded iron object which may have been a door hinge. The stone-footing of an internal dividing wall was identified at the N of the internal paved area, along with a fragmented iron cauldron that had been deliberately incorporated into the paving during its construction.

The final northern third of Structure 1 represented the byre end of the longhouse, with an uneven and sloping surface of exposed natural. No drain was identified within this surface, although it is possible that it lay off-centre in the unexcavated W half of the structure.

In the space to the E of Structure 1, between the eastern flank of the longhouse and the gable end of two other (unexcavated) structures a metalled yard surface was exposed directly beneath the turf. Lying upon this, and butting against the E wall of Structure 1, we identified a stack of cut peat bricks.

The walls of Structure 1 consisted of a single skin of dry stone walling, using locally available and unfinished granite. The quality of construction was notably higher in the southern ‘domestic’ end of the building.

The excavations in Structure 2 revealed a very different structure. Here, removal of the rubble collapse spread revealed an uneven beaten earth floor across the interior of the structure. The construction of the walls varied significantly from E to W. The E gable end of the structure (separated from Structure 1 by a gap of c0.5m) was constructed of substantial dry stone walling, very similar in construction to that of Structure 1. However, the W wall was completely different in form, consisting of banked turf and rubble butting between two large granite boulders at the NW and SW corners of the structure. Meanwhile the N and S walls were thick double skinned turf and stone constructions.

Although removal of the collapse material produced finds including ceramic sherds, iron objects and sherds of glass, the density of finds from Structure 2 was far lower than in Structure 1. However, in the E end of the structure, the removal of the rubble collapse material revealed the presence of five fragments of riveted copper sheeting, almost certainly representing a fragmented copper pot-still, of the type used in 18th-century whisky distillation across the Highlands.

Archive: Timespan Museum (Helmsdale). Report: Highland HER and RCAHMS

Keir M Strickland, Orkney College, UHI, 2013

(Source: DES)