2007:857 - Prumplestown Lower, Kildare

This site was excavated as part of the N9/N10 Kilcullen to Waterford scheme, Phase 3: Kilcullen to Carlow. The topography of the site is highly significant; it extended over 350m from the southern bank of the River Lerr, a tributary of the River Barrow. The site incorporated the waterlogged river flood-plain and a sharp natural rise in the topography which led to a pronounced gravel ridge running parallel to the river. As part of the same road scheme a ring-ditch complex with associated cremation burials and structures (E2966; see No. 869 below) was identified across the river, 350m to the north-east. The two sites are closely associated and together provide an exceptional insight into the archaeological landscape focused on the River Lerr.
An area measuring 22,834m2 in total was stripped of overburden in two stages. A range of archaeological features dating from at least the Bronze Age, through the medieval to the post-medieval period were identified and excavated within this area. A large number of features that were investigated turned out to be related to the natural geology of the area. It should also be noted that the gravel ridge traversing the site had suffered extensive plough damage. The main archaeological features are outlined below.
Dry-land area
The first stage of the excavation focused on the dry-land element of the site. A prehistoric structure (as yet undated) was located on the southern side of the gravel ridge. It consisted of a segmented curvilinear gully (max. 0.35m wide by 0.18m deep), defining the northern, southern and eastern sides of the building, and a series of ten post-holes on the western side. The defined area measured 10.5m long (north–south) by 8.5m wide. A gap in the gully on the eastern side is likely to represent an entrance with further portions of gully extending away from the structure to form a ‘porch-like’ feature. A further 28 possible post-holes were identified in the interior of the structure as well as an inner section of gully on the northern side. Further post-excavation work is needed in order to comment on the structural arrangement of the interior. A subrectangular pit was identified in the interior, on the northern side. It measured 2.35m long (east–west) by 1.12m wide and between 0.42m and 0.92m deep. The lower fills contained a high proportion of rounded and sub-angular stones (mainly granite) measuring between 0.1m and 0.4m in maximum dimension. A fragmented granite saddle quern and a small quantity of animal bone were among these stones. No hearth was identified in the interior and no other artefacts were recovered in the vicinity of the structure. Four pits (two with evidence of burning in situ and charcoal-rich fills) were located less than 2m from the enclosing gully to the north-east and east and several post-holes were found close to the entrance feature on the eastern side.
A prehistoric pit-circle (as yet undated) was located c. 34m north-west of the structure. It consisted of a circular arrangement of ten large pits enclosing an area measuring 7m in diameter. The pits were all roughly circular in plan and measured between 1.75m and 2.5m in diameter. One of the pits on the eastern side was 0.27m deep, which was significantly less than the rest of the pits which were between 0.57m and 1.14m in depth. The average distance between the pits was 0.42m, except for the gap between the shallow pit and the pit to the south of it, which was 1.2m wide. This could be interpreted as an entrance on the eastern side. A small amount of prehistoric pottery was recovered from two of the pits, while a third contained a flint flake. A small amount of animal bone was recovered from a number of the pits including a complete antler. During excavation it was noted that there was no evidence that the pits functioned as post-pits for large timber posts or as sockets for a standing stone circle. The typology of the monument compares well with pit-circles excavated in Britain dating to the Middle Neolithic; however, further post-excavation work is necessary before a date and interpretation for this pit-circle can be put forward.
A segmented ring-ditch, which had been truncated by a modern field boundary, was identified 6m north of the pit-circle. It had two causewayed entrances: one on the west side measuring 2.8m wide and one on the east side measuring 2.6m wide. The ditch measured up to 1.5m wide by 0.65m deep and enclosed an area measuring 9m in diameter. Over 60 worked flints were recovered from the fills of the southern segment of the ditch while four flints and a small quantity of animal bone (including antler) were recovered from the northern segment. Two small post-holes were located inside the eastern causeway. Two extended supine human burials, orientated east–west, were located side by side in the southern half of the ring-ditch interior. These are likely to be related to two other east–west burials located 6m and 20m north-east of the ring-ditch. These skeletons are as yet undated and further post-excavation work is needed in order to comment on their relationship to the ring-ditch.
In the area between the ring-ditch and the pit-circle two cremation burials were excavated. Four very shallow features of similar dimensions to the cremations, which appeared to have been truncated by ploughing, were also identified in this area.
A network of gullies, measuring on average 0.5m wide and 0.3m deep, seemed to have been associated with the prehistoric activity on site. The southernmost part of this network consisted of two lengths of east–west-orientated gully which extended beyond the road-take on either side. Three lengths of gully formed a roughly parallel line 118.25m to the north; this also ran beyond the road-take on either side. These two sections were linked by two lengths of north-east/south-west-orientated gully, measuring c. 32m and c. 86m long. The lengths of gully were separated by causeways (between 1.5m and 4m wide) which appeared to have been entrances. It seems likely that these linear features would have served as some type of boundaries. While they would have been too small to have had a defensive function, they could have defined certain areas associated with the ritual and domestic activity on site.
The only prehistoric activity identified outside the limits of the gully network was a pit on the northern slope of the gravel ridge. The pit measured 1.09m long (east–west) by 0.8m wide and 0.16m deep. A cache of three high-quality stone artefacts (likely to be projectile heads) was recovered from the fill of this pit. Although yet to be analysed, two of these seemed to be made of chert and the other of a softer stone that had degraded significantly. These measured between 0.095m and 0.12m in length.
A curvilinear medieval feature was also located on the northern slope of the gravel ridge; it consisted of three overlapping ditch sections. On average the ditches were 1.5m wide and 0.65m deep. The first section of ditch started on the edge of the river flood-plain, on the eastern side of the road-take, and curved towards the south where it terminated; it measured c. 37.5m in length. The second section, which measured c. 118m in length, overlapped slightly with the first and curved around towards the south-west; it continued in this direction and terminated close to the summit of the gravel ridge. The third section overlapped slightly with the second and continued in a south-westerly direction beyond the road-take; 27m of this ditch was exposed. Several sherds of medieval pottery were recovered from the various ditch fills and a remarkably well-preserved ringed pin was recovered from the first section of ditch. Initial analysis suggests it is a plain, loop headed Dublin-type ringed pin dating to the 8th or 9th century. In the absence of artefactual evidence, no other features on the site can be dated to the medieval period in advance of further post-excavation work.
Wetland area
Phase 2 of the excavation focused on the southern flood-plain of the River Lerr. It must be noted that the River Lerr has been culverted for use as a mill-race but the flood-plain is still active. First-edition OS maps show the river to be slightly further north prior to culverting. The topsoil and upper levels of silt (0.6m in depth) were removed by mechanical excavator to reveal layers of waterlogged peat, silt and sand which were overlying coarse gravel natural. The maximum depth of these deposits was c. 0.8m. The site was divided into six cuttings for ease of access and to allow sections to be recorded. Several channels relating to old watercourses could be identified in the sections and a large ‘island’ composed of thin deposits of sand and peat was identified towards the centre of the site.
Cutting 1, the largest cutting, was located on the eastern side. Topsoil and upper layers of silt, sand and peat (totalling 0.5m in depth) were removed to expose an expansive, generally homogenous, layer of grey/black peat which had an average depth of 0.2m. This peat layer was overlying the natural gravel and petered out towards the south at the edge of the flood-plain and towards the east, where the topography also rose slightly. Archaeological material was identified within this peat in the form of a loosely arranged linear deposit of brushwood and worked stakes and pegs, which was parallel to the edge of the flood-plain (north-east/south-west). This was interpreted as a rough trackway which would have been laid down to enable movement through the flood-plain. Two similar, though less dense, deliberate arrangements of wood were identified between the main trackway and the river. Worked wood was also found randomly scattered on either side of these trackways. Few of the stakes were in an upright position and no deliberate pegging of brushwood could be identified. It is suggested that the constant water action associated with the flooding of the river would have slightly displaced the wood in the trackways resulting in the scattering of worked wood and the levelling of upright stakes/pegs should they have been present. In total 308 pieces of worked wood were recorded in Cutting 1. A finely made wooden spear and a perforated block of timber (yet to be analysed) were found in association with the main trackway in Cutting 1. A moderate quantity of animal bone was also recovered from in and around all three trackways, perhaps suggesting the movement of domestic animals across the flood-plain, though further post-excavation analysis of the bone is required in order to substantiate this.
In Cutting 2 a continuation of the minor trackways from Cutting 1 could be identified, although they petered out at the edge of the natural ‘island’ in the centre of the site. In total 107 worked pieces of wood were recorded in this area, including a possible oar, which is currently undergoing analysis. Some animal bone was also recovered in this area. A small part of a fulacht fiadh was identified in the south-west corner of Cutting 2 (see Cutting 6 below for details).
In Cutting 3 the stratigraphy of the peat and riverine was more complex than in the other cuttings. Following removal of topsoil and some upper levels of silt, a mound of heat-shattered stone and charcoal was identified in the north-west corner, overlying grey silt. It measured 9.25m long (north–south) by 6.6m wide and 0.11m deep. Although it appeared to be a fulacht fiadh mound, no associated trough was identified. A second area of fulacht fiadh activity was identified 15m to the south-east. Here a roughly circular trough measuring 1.05m in diameter by 0.42m deep was cut into peat and underlying silt. The base of the trough was lined with a lattice of trimmed branches (less than 0.02m in diameter) which were laid horizontally. A number of them had been driven into the sides of the trough. A ring of more substantial stakes (up to 0.1m in diameter) had been placed vertically around the circumference of the trough. These stakes had been driven through the base of the trough into the natural gravel. The trough fills consisted of peat and silt with a small amount of heat-shattered stone and charcoal. There was no associated mound other than a small amount of burnt material that extended beyond the edge of the trough.
Following excavation of the fulachta fiadh described above, further layers of silt were removed to reveal a part of a third fulacht fiadh in the south-east corner of the area (see Cutting 6 below for details). In the north-east side of Cutting 3 a very significant wooden haft for a late Bronze Age socketed axe (initial analysis) was recovered. Two stakes, which were not in situ, were recovered nearby but no other worked wood was identified in the cutting.
No archaeological material was identified in Cutting 4, which was adjacent to the edge of the flood-plain and contained only very shallow peat deposits.
In Cutting 5, which was also adjacent to the edge of the flood-plain, three spreads of charcoal were identified in the upper silt deposits at the wetland/dry-land interface. Two of these contained burnt bone, which is likely to be animal bone (analysis ongoing). A small concentration of worked wood (seven pieces) was recorded in the north-west corner of the area close to the fulacht fiadh in Cutting 6. A large number of tree roots were identified in the south-west side of the cutting, suggesting the area was wooded at some stage. There were no peat deposits on the east side of the cutting, where the main trackway from Cutting 1 was leading. This could mean the trackway was leading to/from that point or that any part of the trackway in that area has decayed in the absence of peat.
Cutting 6 was located at the junction of Cuttings 2, 3, 4 and 5 following removal of the baulks. Topsoil and upper silt layers were removed to reveal a small mound of charcoal and heat-shattered stone overlying peat. The mound measured 5m long (east–west) by 2m wide and a maximum of 0.2m deep. No definite trough was identified but some shallow depressions beneath the mound may have been associated with the water-heating process. Immediately south of the fulacht fiadh a concentration of worked wood (13 pieces) was recorded, and the area north of this was occupied by tree roots linked to those in Cutting 5. A large antler was found among these tree roots.
Analysis of the worked wood from the site is in its early stages but the majority of the stakes/pegs and trimmed branches seem to have been made with a metal axe. The Late Bronze Age axe haft indicated prehistoric woodworking on the site, while initial analysis of the perforated block from Cutting 1 suggests a medieval date.