When archaeologist F. Reilly opened up a roughly 50 by 30 metre strip of ground at Carmanhall under licence 02E0074, what emerged was the ghost of a ditch that had already been forgotten for at least a century before the first Ordnance Survey teams arrived in the early nineteenth century. The OS maps, both first edition and later revisions, show no trace of it. Whatever this boundary once divided, it had been erased from the landscape long before it could be recorded.

The ditch itself, designated C55 in the excavation record, meandered northward through the site for approximately 40 metres. It was generally U-shaped in cross-section, averaging about a metre wide, though it varied considerably in both depth and width along its length. Scattered through the fill were large stones, angular to sub-rounded, ranging from 0.1 to 0.4 metres long, interpreted as the collapsed or deliberately dismantled remains of a wall that once ran along the ditch's western side. A single stone still stood upright at the surface midway along that side, possibly the only original wall stone surviving in place. Among the fill material, excavators recovered a possible medieval pottery sherd and three pieces of flint, cautious evidence for the boundary's age. The northern end of the ditch told a slightly different story: its fills contained more charcoal, a detail that appears connected to nearby burning features, including an irregular cut whose fills had been burnt in situ and a circular fire-pit depression showing clear scorching. The excavators suggest the ditch may have been a medieval field boundary, infilled during eighteenth-century land works, perhaps associated with the creation of parkland around Rocklands or Leopardstown houses.

Carmanhall today is a built-up area on the southern edge of Dublin city, close to Sandyford and the Leopardstown road. The excavation site is not publicly accessible as a heritage feature, and there is nothing to see above ground. The value here is in the record rather than the visit: the excavation report, compiled by Geraldine Stout and Padraig Clancy, sits within the Irish national archaeological database and offers one of those quietly absorbing cases where a routine development excavation recovers something genuinely ambiguous, a boundary whose purpose, ownership, and end we can only sketch around the edges.