1997:002 - BALLYGORTGARVE, Antrim

This site was the proposed location for a 60-house development. It lay in the northern part of Ballygortgarve townland, about half a mile north-north-west of the centre of Crumlin village, Co. Antrim. The above grid reference refers to the only site within the development recorded in the SMR prior to this excavation-the site of a historically attested holy well and an enclosure detected by geophysical survey. The immediate area of the development was a splendid location for settlement at all periods, with good drainage, fine soils, and a salmon-, eel- and trout-bearing river close by.

A geophysical survey had been carried out in 1995 at the north-east end of the field. The main site discovered during this process was an L-shaped section of ditch, presumably representing a right angle in a rectangular enclosure, in the northern corner of the field. This work was prompted by early map references to a holy well, 'Mulleague well', in this area. It has been suggested that this is linguistically cognate with the 'Church of Miloc' placed in this general area by many documents dating from the 12th to the 17th century. The rectangular enclosure, the well and the historical reference suggested the possibility that a medieval or even earlier ecclesiastical site lay in the northern corner of the field in which the development was to take place.

The site was initially subjected to a brief three-day field-walk (6/10/97 to 8/10/97). This confirmed the existence of a large ditch in the area indicated by the geophysical survey, as well as locating two new sites and an impressive collection of topsoil finds-predominantly either worked stone of the Neolithic/Early Bronze Age period or pottery dating from the last two centuries. It was decided to undertake a supervised soil-strip using a machine with a smooth-bladed bucket. This revealed 119 'features', generally standing out as dark areas against the orange-brown subsoil. The only other large site revealed was a ring-ditch, approximately 16m in diameter, with a circle of pits lying to the south-west and another dense cluster of pits to its south-east. The vast majority of features in the rest of the field were individual, isolated pits averaging c. 0.4m in diameter. The excavation commenced on 17/11/97 and ended on 27/2/98.

The enclosure
The L-shaped stretch of ditch in the northern corner of the field showed up as a dark, clayish, charcoal-flecked upper fill against the orange sand/gravel subsoil and closely matched the form indicated on the geophysical plot. Two 1.5m-wide sections were put though this ditch. Both revealed a moderately steep-sided ditch cut into the subsoil and measuring 2.8–3.9m wide at the top. The bottom third dropped very steeply and abruptly down to a flat bottom, approximately 0.9m wide. The ditch was probably originally around 1.8m deep. No trace of a bank survived on the surface.

A small amount of charcoal was recovered from the lowest layers of the sections-enough for an accelerator radiocarbon date. Only the upper layer produced any finds: two unimpressive flint flakes. However, this layer lay above a turf-line that had developed when the ditch was two-thirds full. The finds are probably in a derived context, possibly the result of ploughing.

The ditch stratigraphy was suggestive of fairly rapid silting, followed in places by bank slump from the internal side, only stabilising (there were traces of a turf-line) when the ditch was more than half-filled and rendered ineffective. The apparent rapid refilling need not have taken more than a year or so. The lack of any trace of revetment stones in the backfill or a slot for a timber revetment inside the ditch suggests that turf must have been used. Only a very small part of the interior of the enclosure was exposed by the stripping of the development area (perhaps only 10m by 10m), and within this small area three very truncated shallow pits were found. Unfortunately these contained no finds or carbon.

The geophysical survey and the soil-stripping had to be limited to the area of the development. For this reason, all we can say for certain is that this was a large ditch with a near-right angle. It would seem likely that it represents part of a rectangular/square enclosure. Geographical considerations suggest the possibility that the ravine that this backed onto may have formed one side of the defence and that the enclosure may have been only three-sided. The same considerations suggest that it cannot have enclosed an area much larger than 30m on each side. It also seems very likely that the site of a well, mentioned above, lay within or immediately outside this enclosure.

It is possible that this was an ecclesiastical site of the medieval or Early Christian period. As mentioned above, the nearby well was called 'Mulleague Well' on OS map editions 1–4. Reeves identified this with the 'Church of Miloc' mentioned in the 1306 Taxation. He also showed that this church is mentioned in many documents of the 12th–17th century (Reeves 1847, 3–4). It is obvious from the order of its appearance in various itineraries that the historic ecclesiastical site referred to was located in the Crumlin area.

The rectangular shape of the postulated enclosure would strengthen the argument as vallums of this form are known from the Early Christian period. Although the circular plan is more common, rectangular perimeter ditches are occasionally associated with Early Christian ecclesiastical sites–Iona and Clonmacnoise are large examples.

The lack of debris in the ditch and its apparent rapid silting could perhaps be a result of the church being a chapel rather than a monastery. Most references do not indicate a monastic site. The Ulster Visitation book refers to the ruined 'church of Meleeke' in 1622. All the references to an actual functioning church date from between 1180 and 1306, a period of only 126 years. Perhaps its usefulness as a chapel was undermined by the big drop in population known to have affected much of Europe in the mid-14th century.

If the ditch represents the historic church site's vallum, then the site must have been, in effect, undefended for the bulk of its life. Alternatively the real site of the church may lie elsewhere in the townland.

Another strong possibility is that this could be a medieval moated site. The form (flat-based ditch, internal bank), dimensions and setting of the site (riverside, low-lying, good land, beside natural springs and in an area of tentative second-generation Anglo-Norman settlement) exactly match the definition of this type of site (Barry 1987, 93). There are no certain examples of this site type in Ulster, but there is a small cluster of possible examples in the south Antrim area.

The ring-barrow, pit circle and cist complex
Soil-stripping in the north-east part of the field revealed an area rich in interesting features. The most significant group was a cluster comprising a ring-ditch with a central cist burial (not revealed by initial stripping), a small ring of pits and a patternless cluster of substantial pits.

The truncated ring-ditch measured c. 16m in internal diameter, the ditch only surviving at subsoil level. The surviving depth of the cut varied from 0.1m to 0.3m. Its width was between 0.4m and 1.1m. On plan, it was penannular with the open end (1.6m wide) facing approximately north. The profile varied from a steep 'V' to a flat 'U' (a gentle 'V' was typical). The unmonitored soil-strip had truncated the west side more than the east. Five uninformative small pits lay within the ring-ditch, and a small, very truncated cremation deposit lay just outside to the east. The only artefacts recovered from the ring-ditch fill were a struck flint, a piece of unworked obsidian and a piece of unworked porcellanite. Only one small area of ring-ditch fill produced any charcoal.

The main feature inside the ring-ditch was a centrally placed cist burial. This was set in a bowl-profiled, flat-bottomed pit, about 0.8m in diameter and 0.75m deep, cut into the subsoil. It is likely that the top 0.3m was truncated by soil-stripping and earlier ploughing. This hole had been floored by six flat slabs approximately 0.2m thick. It seems likely that the next stage was the deposition of the collared urn and the vase food vessel (both in inverted positions) directly on the floor slabs. The collared urn contained a large amount of cremated human bone, while the vase food vessel was empty. The only grave-good found was a plano-convex/slug knife, found on the floor paving beside the urn. The bones are currently being analysed by Ms Eileen Murphy. The urn and food vessel are currently undergoing conservation by Mr Malcolm Fry (EHS).

The cist was ingeniously constructed in drystone with a strong suggestion of corbelling. However, we have lost the upper 0.3m through ploughing and soil-stripping. It is likely that the top was covered by a central slab. At some stage a lot of topsoil had percolated into the cist, occupying most of the voids.

The pit circle
About 1.5m south-west of the ring-ditch lay a cluster of eight pits, forming a circle approximately 2.4m by 2m in internal diameter. The individual pits were around 0.44–0.74m in diameter and 0.07–0.2m deep. They were all subsoil features, the cuts being bowl-profiled. The fills were all homogeneous and were generally dark-mid-brown silty loams containing charcoal. The finds included seventeen flint flakes or blades, three flint end scrapers and some burnt bone. No pottery was recovered. The finds indicate a Neolithic or Early Bronze Age date for the circle. Enough carbon was retrieved for dating.

The apparent homogeneous and relatively shallow nature of the pits may be illusory, as the upper layers (probably the greater part) of the pits have been removed by truncation. The very small diameter of the circle precludes these pits representing a structure. Their proximity to the ring-barrow immediately suggested a ritual interpretation. However, this is far smaller than any of the other known Late Neolithic/Early Bronze Age pit or post circles and can hardly have performed the same function. On the other hand, they could represent the surviving scatter of post-holes/pits around a totally truncated and removed surface hearth.

The pit cluster
A dense, patternless scatter of eight pits was situated about 4m south-east of the ring-ditch. These varied from 0.5m to 0.8m in diameter and from 0.15m to 0.4m in depth. The profiles were generally bowl-shaped with shallow to medium sides. The overhangs of natural, giving the bowl-like profiles, were probably the result of the pressure of the Hymac during soil-stripping. These pits contained dozens of sherds of Coarse Ware, several flint flakes and blades (including some burnt), and a broken and burnt porcellanite axe. There were also traces of burnt bone. This mixing of Neolithic/Early Bronze Age-type lithic material along with Coarse Ware normally associated with the Mid-Late Bronze Age is interesting.

References
Barry, T.B. 1987 The archaeology of medieval Ireland. Methuen, London.
Reeves, W.B. 1847 The ecclesiastical antiquities of Down, Connor and Dromore. Dublin.

Editor's note: previously labelled as "Ballygarve: 1"

31 Madison Avenue, Cavehill Road, Belfast