D.J.Woolliscroft and M.H.Davies
Glenbank was discovered from the air by G.S.Maxwell (1) in 1983 and is currently the southernmost Gask installation known. It lies just south of the Roman road and, at the time, the fortlet itself was estimated to measure c.36m (NW-SE) x 27m (NE-SW) over the ramparts. It is surrounded by a double ditch and has a single entrance facing north-west towards the road. The site sits on a very slight mound with excellent views to the North, East and West and a somewhat poorer view to the south, where it faces gently rising ground. To the north-west, it has all of the know Gask installations in sight as far as Kaims Castle, including the fort of Ardoch.
Excavations conducted by Maxwell, soon after the site’s discovery, located gate posts flanking a 3m wide entrance and suggested that the fortlet (like other installations on the Gask line) had been deliberately demolished at the end of its service life. The work also produced a number of amphora fragments(2). These excavations have yet to be published, however, and so both the full structural data and the date and context of the finds remains unknown.
In an attempt to acquire additional information, the Gask Project conducted a resistivity survey of the site in 1998. The fortlet showed well on the resulting plot with both ditch lines being visible as clearly separated, narrow, parallel bands of low resistance around their entire circuits, although the north-western corner was somewhat obscured by a separate north-south running band of low resistance, probably a modern drain or water pipe. The south-eastern corner was particularly distinct.
The outer ditch measures c.51m NW-SE x c.49m NE-SW externally. The inner ditch has an external diameter of c.41m (NW-SE) x 39m (NE-SW), with an inter-ditch separation of 3-4m. Both ditches appear unusually slight for such a site, with the inner ditch around 2 – 2.5m wide and the outer ditch noticeably smaller, perhaps as narrow as 1m wide in places. Nevertheless, comparable ditches were found at the nearby double ditched towers of Greenloaning and Shielhill South and so this is probably normal for the area. The inner ditch entrance gap, at 3-4m, corresponds closely with the 3m entrance found by Maxwell in the fortlet itself, but the outer ditch entrance may be considerably wider, perhaps as much as 10m. This too follows the pattern of Greenloaning and Shielhill South, but there are a number of difficulties here. Firstly, the outer ditch appears to become even slighter on the north-western side of the site, and particularly around the entrance. This again follows the precedents set by the towers in the area, but it does make the ditch particularly difficult to see, and so measure accurately, on the resistivity plots. Secondly, there is a hint in the resistivity image that the inner and outer ditches might join on both sides of the entrance, rather than coming to separate butt ends, as is more usual. This impression is reinforced by air photographs of the site, which appear to show the same phenomenon. Caution is needed here, in the absence of excavated confirmation, however, for Greenloaning produced exactly the same effect on both air photographs and resistivity plots and yet, on excavation, the ditch ends proved to be perfectly normal, with the apparent joining an illusion caused by later drainage slots. The resistivity effect (although not that seen on air photographs) at Glenbank may even have been caused by Maxwell’s excavation trenches, whose locations remain unpublished, but as these are not otherwise visible on the plot, they may have been too shallow to produce a resistivity signature.
Unlike the neighbouring tower of Greenloaning, no sign of an external upcast mound could be detected in the resistivity image, but the ditch spoil may well have been ploughed away, or it may have been used to backfill the ditches from which it originally came.
Two parallel areas of high resistance are visible on the eastern and western sides of the interior which might represent the remains of a rampart or, less probably, of internal buildings. Neither interpretation can be guaranteed, however, as the features are on the line of a general band of high readings, which runs diagonally from north to south across the whole grid, and for which the cause is probably geological. Nevertheless, the entire fortlet interior appears to show a slightly higher resistance than the area outside the site. This could indicate that the interior was surfaced with gravel or cobbles, like those of Kaims Castle fortlet and Greenloaning tower, but again it is impossible to be certain without excavation data.
The site’s dimensions are interesting. The double ditched towers in this same southern sector have outer ditch external diameters hardly bigger than those of the single ditched examples further north. Even allowing for these sites’ unusually narrow ditches, this obviously leaves them with much smaller interiors, and suggests that it is the inner ditch, and not the outer which is the additional feature. We were, therefore, curious to see whether the same would be true of a double ditched fortlet. In fact, although the inner ditch dimensions at Glenbank are slightly smaller than those of the more northerly single ditched fortlets at Kaims Castle and Midgate (which are both around 44m x 40m externally), the whole site, including the outer ditch, is considerably larger. It was not possible to measure the exact dimensions of the fortlet itself from the resistivity data, but it must be smaller than the c.36m (NW-SE) x 34m (NE-SW) area enclosed by the inner ditch, as there will inevitably have been a berm between the rampart and ditch lip. Maxwell’s stated size of 36m x 27m is thus probably a slight overestimate, at least on the north-south axis. Kaims Castle and Midgate are both around 30m x 28m over the ramparts (22m x 20m internally) and, given slightly narrower berms, Glenbank may well have been much the same.
1. Maxwell, G. S., and Wilson, D. R., Air reconnaissance in Britain 1977-1984, Britannia XVII,1987
2. Maxwell, G. S., Flavian Frontiers in Caledonia in Vetters, H., and Kandler, M., Internationalen Limeskongresses 1986, Vienna, 1990, p354