2008:1048 - Tulsk, Roscommon

The extent to which ringforts can retain evidence for occupation in the later medieval period is a question that the Discovery Programme’s Medieval Rural Settlement Project has been pursuing actively since 2004, when the programme commenced excavation on Tulsk Fort, a raised ringfort located at the centre of Tulsk village (Excavations 2004, No. 1484; Excavations 2005, No. 1352; Excavations 2006, No. 1758; Excavations 2007, No. 1520). The ringfort lies 60m east of a stone-filled earthen mound in Castleland townland that is considered to be the site of Tulsk Castle, and is c. 100m north of a Dominican priory, founded in 1448. The ringfort retains a dominant position in the settlement’s wider topography and suggests that this was a location of primary importance throughout the medieval period.
The excavations have now revealed a sequence of five main horizons of activity, two of which appear to be early medieval in date when the site served as a ringfort, and two of which are later medieval and early modern/Elizabethan in date respectively. The fifth and lowest horizon predates the construction of the ringfort and belongs to the prehistoric period. The earliest medieval stratum seems to date to the 10th–11th centuries, based on small finds including lignite bracelet fragments, bone pins, copper-alloy and iron ringed pins, glass beads and a gaming piece. The assemblage suggests the presence of a reasonably affluent family and is typical of many ringfort excavations. The base of a corn-drying kiln and a series of other burning events and associated ash spreads over the occupation level produced the bulk of the small finds. The bottom of the ringfort bank and a deep V-shaped fosse associated with it has also been identified. The bank is constructed from boulder clay and is supported by a drystone revetment on its interior. The western part of the site has revealed a sequence of clay deposits introduced to raise the interior surface, and to transform the site into a platform. A depth of c. 1.2m of these clays survives, but no occupation levels have been associated with the raising event as the uppermost levels were truncated by later activity.
A series of trenches were cut through the introduced clays to accommodate massive rubble foundations which extend in places to more than 2.5m in depth. The foundations supported a masonry tower and this sequence of events represents the first clear appearance of the later medieval horizon. The tower is aligned north-north-west/south-south-east on the eastern side of the site. It retains battered external walls with a rounded north-east corner, the only corner exposed by excavation. At this stage, the tower measured 20m long by 10m wide externally. A garderobe set into the wall fed directly into the external fosse. The interior space appears to be divided into a main chamber and a narrow north chamber. There is no obvious entrance to the tower and it remains possible that this lies in the unexcavated southern half. In creating the tower, the builders modified other parts of the ringfort. They recut the external ditch to make a broader but shallower fosse. They also cut a ditch internally to separate the tower from the wider interior of the site, thereby creating a bawn area from the larger ringfort space. In time the tower collapsed, or was collapsed, and a period of rebuilding occurred that included a 15m long by 9m wide extension to the tower on its eastern side. The extension effectively transformed the site by filling in the fosse at this location and adding what was probably a fortified house or hall to that side of the tower.
All of these events occurred before the mid-1500s, it seems, because at that stage the ruined encastellated ringfort enjoyed one last horizon of activity when it was reoccupied to serve as part of garrisoning works associated with the Queen’s governor to Connaught, Sir Richard Bingham, who visited Tulsk in the 1590s. The discovery of building elements and a stone-lined cellar, along with small finds dominated by musket and pistol shot and silver coin, indicate that Bingham’s works extended to what would have been a very suitable platform from which to protect his northern flank.
The excavations at Tulsk Fort roundly inform issues surrounding continuity of ringforts beyond the first millennium. The results have not yet revealed continuous settlement from the early medieval period into the later medieval period because each new significant building horizon effectively truncated the upper levels of preceding deposits with the resultant loss of contextual information. There are some indications for continuity from the small finds recovered but these matters need to be teased out in greater detail. The complexity of the site calls into question the association of the historically referenced Tulsk Castle with the mound in the adjacent Castleland townland. That site has yet to be investigated but it is relatively slight in comparison. It is perhaps worth suggesting that the ringfort may have served as the castle site. The foundation of a castle in Tulsk in 1406 must be associated in some way with the emergence of the O’Conor Roe as a separate and distinct line of the O’Conors, following a schism in 1385 which resulted in their line and that of the O’Conor Don, whose primary base lay in Ballintober, to the west. Tulsk became a principal residence of the O’Conor Roe and it is therefore quite appropriate to consider the ringfort in this context. A final season of excavation is planned for 2009, to bottom-out the medieval horizons.