1994:008 - SHANDRAGH, Knockans South, Rathlin Island, Antrim

The site lies in the south central area of the island with the southern cliffs of Rathlin Island located 300m to the south. Beyond lies the stretch of sea known as Rathlin Sound, with Ballycastle a little over six miles away. The site lies within a complex of largely derelict farm buildings set on a sheltered strip of level ground with a natural basalt rock face lying directly to the north.

Finds of porcellanite axes and flintwork from within the general location of the derelict farm had been reported to HMB Environment Service DoE (NI) and the Ulster Museum Belfast by the landowner on various occasions since 1992 resulting in several field visits to monitor disturbance, largely by field drains and cultivation of a small vegetable garden. During further drain laying in late 1993 substantial quantities of porcellanite, flint and decorated pottery sherds came to light near the landowner's cottage. The pottery in particular was recovered in large unweathered pieces and comprised Late Neolithic and Early Bronze Age wares. A field visit underlined the necessity for open plan excavation in advance of further planned development in the area.

A small-scale excavation funded by Environment Service DoE (NI) was undertaken by the Ulster Museum over an eleven week period from March 21 to June 10. The aim of the excavation was to recover a contextual sequence for the large volume of material collected by the landowner Mr Paddy Burns (who assisted in the excavation) and to record any subsoil features which may point to the nature and extent of the site.

An area approximately 45 sq m was excavated by hand (two cuttings) in an elongated area directly between the cottage (east) and a derelict farm building (west), being the area where the greatest concentration of finds occurred.

On excavation sod was found to directly overlie the 19th-century farmyard, consisting of a neatly prepared cobbled surface. A mixture of prehistoric and later finds were recovered from and within this surface including several polished porcellanite axes which had been re-used as cobble stones. Below the farmyard surface lay a thin layer of loose dark brown soil containing prehistoric and later finds (including possible souterrain ware sherds). Below this layer a substantial deposit of compact brown soil containing a vast collection of Late Neolithic pottery and lithics was uncovered. The deposit appeared almost midden-like in content and extent. At the southern extremity of the excavation a concentration of cremated human bone and sherds from a food vessel were uncovered, though no evidence for a cist or pit was found. Below the artefact rich 'midden' deposit lay a prehistoric cobbled surface through and below which a series of pits and a gulley had been cut. In total 10 pits, three small postholes and one curving gulley were uncovered. All the features are not contemporary, but they were nearly all found concentrated within a band through the middle of the upper cutting (Tr 1) lying on a relatively level ground surface north of the area containing the major volume of 'midden' material which lies within a hollow to the south (Tr 2). All the cut features contained finds of Late Neolithic pottery and lithic material as well as charcoal. Two pits in particular (112 and 118) were associated with finds of significance. Both pits contained a vitrified material resembling slag in composition and texture, though found to be very light in weight. Both pieces are presently being examined in Oxford and await full identification.

During excavation it was noted that on and below the prehistoric cobble surface areas of lithic artefact concentration occurred and this may reflect possible knapping areas, indeed while most features contained a combination of porcellanite and flint, several were exclusive to one or other category. The excavation was suspended before the cuttings could be 'bottomed-out', however it is apparent from the information already collected that there is substantial prehistoric activity within the area. The vast collection of artefacts recovered during excavation both from the midden deposit and other layers and features suggest a continued use of the site from Neolithic through Bronze Age periods and perhaps further.

The Neolithic pottery recovered prior to and during excavation displays a predominance of decorated forms including Goodlands bowls, Sandhills varieties and Carrowkeel Ware all in association. Plain Neolithic pottery also occurs. One aspect of the pottery noted prior to conservation was the presence of porcellanite used as a temper in several decorated Neolithic sherds; this factor appears unique at present to the material from this site. Early Bronze Age pottery occurs in the form of sherds from both bowl and vase food vessel forms. Sherds from coarse wares, possibly Bronze Age in date, are also present. The lithic assemblage is composed of a large number of artefact types as well as a large volume of flakes and cores of both flint and porcellanite; indeed the initial lithic frequency suggests concentrated industrial activity, especially in the production of porcellanite axes (large numbers of rough-outs) though other tools including scrapers and cores have been recognised. This predominance of porcellanite on the site is not surprising as the Brockley outcrop is located less than one mile to the east. Other lithic material recovered includes pitchstone from the Isle of Arran in the Firth of Clyde (confirmed by non-destructive XRF analysis), while numerous hammerstones and pounders have also been recovered. Other finds include many pieces of fine local sandstone and quartz flakes and pebbles. Some unburned bone survived and one broken piece is perhaps part of a perforated bone pin.

The results from the excavation at this stage are very promising: the site is clearly of an industrial/production nature and has evidence for contacts with Scotland. The resources are all apparently local and this includes the pottery with in some cases local stone used as temper. The excavation of subsoil features and the probable 'midden' on the site would suggest that the site was occupied on a permanent or semi-permanent basis, though no definitive structure can as yet be identified. Hopefully further excavation will corroborate this premise.

Botanic Gardens, Belfast, BT9 5AB