The possible enclosure at Fanningstown, in the barony of Coshma in County Limerick, belongs firmly to this category. It is the kind of site that archaeology acknowledges without quite being able to confirm, a placeholder in the record for something that might be there, or might not.

The story of how this site came to attention is instructive. During excavations carried out in connection with the Croom bypass, a possible enclosure was logged as Site 35 in the surrounding area. When the Archaeological Survey of Ireland visited and surveyed the location in 2000, they found no surface remains whatsoever. The 1840 Ordnance Survey six-inch map, one of the most detailed early cartographic records of rural Ireland, depicted the area not as an enclosure but as a long trapezoidal-shaped marshy area, roughly 48 metres north to south and 29 metres east to west. No antiquity was marked. The site sits on a slight west-facing slope in pasture, about 30 metres south of the townland boundary with Garranroe, with moderate to good views in most directions. What gave the site renewed interest were aerial and satellite images: orthophotos taken by Digital Globe between 2011 and 2013, and Google Earth imagery from March 2017, both showing a faint oval-shaped cropmark measuring approximately 15 metres east to west and 12 metres north to south. A cropmark, for the uninitiated, appears when buried features affect the growth rate of surface vegetation, producing subtle differences in colour or height that only become legible from above. The record compiled by Alison McQueen and Vera Rahilly, uploaded in August 2020, concludes that the monument is of doubtful antiquity.

For anyone curious enough to seek it out, the site lies in ordinary farmland and there is nothing to see at ground level, which is rather the point. The interest here is methodological as much as archaeological: this is what the early stages of potential discovery look like in practice, long before any trench is opened or any firm conclusion reached. The cropmark is visible on freely available satellite imagery, and comparing the 1840 OS mapping with the modern orthoimage gives a reasonable sense of how the ground has been interpreted across different periods. Access to the field itself would require landowner permission, and there is no formal public access or signage.