2006:1758 - Tulsk, Roscommon

Excavation at Tulsk took place over a ten-week period for its third of four proposed seasons. The season was organised along the same lines as in 2004–5 (Excavations 2004, No. 1484; Excavations 2005, No. 1352) as a field school for students from around the world. The year also saw the development of collaboration with the University of Minnesota, enlisting the specialist assistance of Dr John Soderberg, bones specialist. The acquisition of a laser scanner at the end of 2005 presented the opportunity for the first systematic use of this device on an archaeological excavation in Ireland.
The excavation at Tulsk continued to focus on establishing the primary sequences of development on the mound. The excavation cutting was extended to the north-east to expose the fuller corner of the stone tower, and to the east to expose a wider area of the building that was constructed against the tower. A significant body of additional data was acquired and the main components are summarised here.
Pre-ringfort
The recovery of a leaf-shaped flint arrowhead and a barbed-and-tanged arrowhead from levels on the east side of the mound attest to a prehistoric horizon, and support the discovery of a saddle quern in 2004 from the cutting opened beside the well, off the north-west slope of the platform. Geophysical survey by Dr Paul Gibson and Dot George, NUI Maynooth, in the course of the 2006 season, identified that the eastern half of the platform area was formerly cut away to a depth of c. 3m to accommodate the footprint for the later stone tower. The prehistoric flints came from redeposited soil that is most logically associated with this work and consequently it is entirely likely that the natural hillock which forms the core to the mound retains, or retained, a prehistoric horizon.
Ringfort
No additional insight was achieved into the main ringfort phase to the site in 2006. Excavation has confirmed an outer fosse, or ditch, to the east of the stone tower, but it remains to be seen whether this ditch is a primary cut or was recut subsequently to accommodate works in the later medieval period.
Stone tower
Excavation outside the tower focused on the north-east corner. It exposed more fully the rounded corner, which is constructed above a right-angled corner plinth. The base of the external wall was not achieved due to a confined excavation area, but it was traced 1.5m below the batter, at which point a narrow outer plinth was highlighted. Taking the distance between the lowest exposed levels and the internal levels surviving within the building, it is apparent that the foundations of the tower extend c. 3m below the top of the present-day mound. This observation is confirmed by geophysical survey, which indicated that the footprint for the tower occupies an area that was actively cut away from the mound. The excavation of the mound was no doubt undertaken to accommodate the tower.
Internally, the tower consists of a main chamber and a narrow north chamber. The southern extent of the tower lies outside the excavated area. The main chamber is further divided into two halves, the northern one of which is flagged and retains three localised ash spreads or hearts. The southern half of the main chamber is cobbled. It aligns with a break in the west wall of the tower, as well as with the c. 15m-long by 9m-wide stone building that was added outside the tower to the east. It is unclear what this alignment means but a working hypothesis is that it represents an entrance feature. The main chamber in the tower was filled with collapsed stonework, which might represent a fallen vault. The north chamber is narrow and, while it would logically represent a stairwell, the floor surface is still too confused for this supposition to be clarified.
The eastern external wall of the tower retains a garderobe chute. The chute falls vertically to a culverted base, which is then channelled further east to empty into the external fosse. The culvert was destroyed when the tower was collapsed, but some of the arch stones survive, and the channel is filled with loose rubble. Excavation did not extend to the channel in 2006 but will do so in 2007.
Eastern building
Some time after the tower was built, a second mortared stone structure was erected. This building lies east of and against the stone tower and measures c. 15m east–west by 9m. Where the boundary wall has been exposed on its northern face, the external façade is well finished and was designed to be seen. The internal façade is constructional and served to accommodate a series of introduced layers of redeposited boulder clay. No definite features were identified in these layers, which can be traced some 0.5m above the surviving height of the wall. The infill layers are themselves truncated by a later field boundary ditch and bank and possible trackway, which is 19th-century in date.
It is still not clear what purpose this eastern building served. It crosses the line of the ditch that surrounds the platform, and the building was built after the tower was erected. The alignment mirrors the internal division within the tower’s main chamber and it is possible that the two are interlinked. However, whether the eastern building is an entrance feature that provided access to the tower or whether it is an external hall structure cannot be defined at this stage. Given that so small an area has been exposed, this is hardly surprising, and it begs the question of whether more extensive excavation is warranted.
Both the eastern building and the tower were collapsed at the same time, and this event is recorded in the sections.
A series of small finds from 2005 in the upper levels of the ditch might indicate a 13th-century date, which presupposes that the tower and the eastern building post-date this period. However, as excavation was focused on removing the post-fosse overburden, and only looked at the ditch layers very superficially, it would be unwise to read too much into the small finds assemblage at this point.
The 16th century
If the excavations in 2005 revealed a surprising level of detail from the mid-1500s and later, the work in 2006 has only made this horizon more detailed. The small rectangular building that protruded above the mound on the northern perimeter prior to excavation was more fully exposed in 2006. It is reasonably well built and extends more than 1m in depth. The southern long wall is, however, only a skin-deep revetment and cannot have been load-bearing; it is suggested that this building served as a cellar feature. It is built against a series of dark glacial till deposits that were pit fill from the 16th century. Two horse skeletons were exposed in the pit fill. A series of silver coins found at the base of the stone building are from the late 1500s as well, which suggests that there was no real gap in time between the filling of the pit and the construction of the building. The building suffered burning, as indicated by a significant charcoal deposit that retains a mixture of wood types. It is thought to represent a former roof and samples have been taken for dating purposes.
Outside the stone building and to its east, excavation revealed a broadly curving low mud bank that ran directly above the internal walls of the early stone tower. It is suggested that the mud bank is the foundation to a breastwork defence that would have supported a latticed timber frame, to provide added defence to north-east corner of the stone building.
The range of equipment and features associated with the 16th-century levels supports the contention made in 2005 that this phase represents a reoccupation of the mound at the time when Sir Richard Bingham, the Queen’s Governor in Connaught, garrisoned the settlement. The stone tower was collapsed by this stage and certain stones have been identified in reused 16th-century contexts. The range of coins recovered continues into the 1640s and suggests that the mound was still in use during the 1641 rebellion.
Looking forward
The excavation project on Tulsk mound was carried out to identify the lordly residence of the O’Conor lords during the ‘invisible’ 13th century, where it has been difficult to see clearly settlements and structures in the non-Anglo-Norman landscape. The results to date suggest that there is real merit in the choice of excavation site. There are already indications of the elusive century. There is, however, a significant overburden in terms of a later medieval tower and associated building, and indeed of the subsequent 16th century. The sum total of these features presents the site as a microcosm to the larger period under review.
The range of material recovered and the scale of the features exposed present legitimate grounds for considering this to be the site of Tulsk Castle, which is recorded in the annals as being built in 1406. If this is the case, then it raises questions concerning the traditional location of the castle as being the stony mound behind Kelly’s pub. It is entirely possible that Tulsk had more than one encastellated structure. It is equally possible that the mound in Kelly’s pub belongs to the Elizabethan period.