1986:72 - DUNMISK FORT, Dunmisk, Tyrone

Following the discovery of a large and intensively used Early Christian burial ground, a probable chapel/shrine, and extensive evidence of glass-and metal-working during the 1984-5 excavations at Dunmisk Fort, the owners, McGarrity Bros., agreed to defer their proposed gravel quarrying in order to allow HMBB to continue investigating the site during 1986.

Excavations, commencing in June and continuing for 8 weeks, were concentrated in the N half of the site, although a limited area was also excavated in the SW quadrant.

The entire monument occupies an area of about 3,500 sq. metres, and of this about 860 sq. metres have now been fully excavated.

The main results of the 1986 excavations may be summarised as follows:

Areas A and D
Further extensions of the industrial (glass/metal- working) area previously found in trenches 28, 36 and 37. Several small hearths were found associated with many fragments of mould, crucible, slag, glass-rods and beadwasters. Several graves were also found cut through these industrial deposits, at least one of which was flexed.

Area B
This trench was sterile apart from a number of shallow and irregularly spaced pits.

Area C
The whole of this area was littered with small stakeholes, which seem to be the remains of a number of ephemeral structures. It is possible that these were the internal structures of a large house of which no other traces were found, apart from a substantial hearth setting and several stone-packed postholes. Alternatively, these stakehole alignments may be no more than numerous rebuildings of screens and shelters associated with the industrial activities on the site. Apart from a little slag and one glass-rod virtually no artefacts or domestic refuse were found in this area.

Area E
This area was examined to try and identify features found in the S corner of trench 2. A combination of a rabbit warren and rather deeper cultivation than usual effectively destroyed most of this area. Several small, slag- and charcoal-filled pits were revealed.

Area F
This trench was opened to try and establish the chronological relationship of the two burial patterns which seemed to merge in this area. The burials found in the central part of the site were well spaced and regularly laid out, while those in the E of the site consisted of dense chains dug head-to-toe. The two patterns did indeed merge in area F, but in such profusion that it was difficult to establish a precise stratigraphic sequence. However, it seems probable that the well-spaced central pattern was the later, in this area at least.

Area G
Apart from a number of post-medieval pits the only feature found in this area was the tumbled remains of a section of walling. This appeared to have been set in a slight trench cut along the brow of the hill. A fragment of a probably Early Christian quern was found amongst this tumbled stone, but no other clear dating evidence.

There is no reason to alter the general conclusions reached at the end of the 1985 excavations: 'It seems likely that the fort was originally a prehistoric earthwork which was subsequently modified and utilised by a small monastic community and became a focal burial ground for the area'.

The distribution of finds and features over this extensively excavated site does allow some additional comment. It would seem that the site was divided into several quite specific functional areas. The SW quadrant produced only large amounts of heavy slags, pits and hearths. The NE quadrant yielded all the glass, crucible and mould fragments. Burials were confined to the E side and mainly to the SE quadrant. Only in the NW quadrant were there any remains that might be interpreted as domestic. Thus it would appear that the SE quadrant formed a ritual or religious zone (though it did later spread a little into the NE quadrant). The SW quadrant seems to have been used purely for the production (smelting?) of metal. The NE quadrant seems to have been used for the production of ornate metal and/or glass artefacts. The NW quadrant may have housed the domestic quarters, or alternatively some other specialist workshop. If the latter is true then any associated residential area cannot be on the hilltop and the sheltered area to the S (where there are two post-medieval farm cottages) seems a likely location.