Cashel in Rinnaraw

In lowlying coastal pasture, encompassing a small bedrock platform located within a few hundred meters of the SW coastline of Sheep Haven Bay. This enclosure is not indicated on the 1838 OS 6-inch map, but is shown as a roughly circular enclosure, marked as ‘Cashel’, on the 1905 edition. It was recorded as a cashel in the Archaeological Survey of Donegal (Lacy et al. 1983, 151, no. 904) where it is described as follows: ‘Internal diam. c. 20m-25m. A 7m arc of massive stones on edge to the NW is all that survives of the 'Cashel' marked on the 2nd and 3rd editions of the OS 6-inch maps. On the NW interior is the remains of an oval (6m NW-SE and 5m NE-SW) grass-covered stone foundation. The wall is up to 2m wide and 0.5m high. The site is located on a subcircular, flat-topped, grassed-over rock platform surrounded by pasture and close to the coast.’ The site was excavated by Professor Tom Fanning, NUIG over five seasons (1987-90; 1992) as part of a programme of training and research (https://excavations.ie: Fanning 1988; 1989; 1990; 1992). Following the untimely death of the excavator, the post excavation analysis was completed under the direction of Michelle Comber (Comber et al. 2006, 67-124). The enclosure encompasses the level top of a natural rock platform (diam. c. 25m), with a steep fall of ground at W–NW, and a more gradual slope at E. The excavation uncovered low remnants of the original enclosing wall which had been built around the perimeter of the rock platform; it consisted of large, set boulders, but in parts also utilised bedrock outcropping. The entrance was in the western arc of the wall, with stone paving lead from it towards a subrectangular house (DG016-016001-) of drystone construction, notable for its rounded external corners, which occupied much of the space in the W half of the enclosure. Two shell middens (DG016-016002-; DG016-016003-) had accumulated immediately adjacent to the house. The E-SE half of the enclosure is thought to have served as an open haggard or yard area, with paving slabs laid in parts to level the surface. Two small shell deposits were found overlying bedrock in this area. There was clear evidence that iron working was undertaken in the SE quadrant. Here, a shallow pit, interpreted as a ‘fire-pit’, was cut into the underlying bedrock. The rock base of the pit was fire-reddened, and was overlain by a layer of almost pure charcoal. A spread of charcoal-flecked soil and friable stones surrounded the pit, and an area of stone paving and possible wall foundations –perhaps part of a lean-to structure- were close by. Finds from this area included considerable quantities of iron slag, iron ore and fragments of furnace bottoms, suggesting that both smithing and smelting took place. The finds recovered during the excavation throw light on the activities and diet of the occupants of the enclosure. A small assemblage of 33 identifiable animal bones, which included bones of the main domestic species, cattle, sheep, pig and horse, as a well as fish bones from ling and cod, a single bone from a young grey seal, and bird bones probably of domestic fowl. The shells retrieved from the middens were predominately from limpets and periwinkles. The processing of cereals is indicated by the finding of several fragments of quern stones, including rotary querns, two fragments of possible trough querns, and a possible saddle quern. Approximately 100 pottery sherds were recovered. All were identified as Man-made underground passage ware, a type of pottery strongly associated with NE Ulster, comprising handmade pots of relatively course texture, burnished black on exterior, with simple, slightly everted rims. The finds assemblage included an array of simple tools: flint artefacts, whetstones, fragments of tanged iron knives, and perforated stones, which may have served as loom, net or line weights. Also recovered were fragments of sepentinite rings or bracelets, stone discs and undiagnostic fragments of worked bone. Several radiocarbon dates were returned for the site, and ‘offer a slightly confused chronology’ (Ibid. 105). A date of late 8th- to early 9th-century was produced by a sample from a layer underlying the wall of the house, and the house walls appear to have been built later in the 9th-century. However, shells from one of the middens produced dates that suggest a phase of occupation in the 11th century or later. Comber (2006, 107) concludes that ‘the main phase of activity at Rinnaraw dates to the ninth century, with the possibility of reuse.' She notes that the house shows interesting similarities in form to Norse houses excavated on Orkney Island, off the N coast of Scotland, and adds that the ninth century saw Scandinavian activity in this part of Donegal, evidenced by documentary references and by various Viking artefacts recovered in this region. The fragments of possible trough querns also hint at a possible link with Scotland. This was a small, self-sufficient settlement, with evidence of on-site iron-working. The form and size of the enclosure and house, and the nature of the finds suggest that ‘the occupants of the house were of relatively modest status, dependent on agriculture and marine resources.’ (Comber et al. 2006, 109). Compiler: Jane O'Shaughnessy Date of revised upload: 12 December 2019