Marked as a holy well on one edition of the Ordnance Survey map, it left no visible trace on the ground and, when enquiries were made locally, no one remembered it at all. No spring, no stone, no worn path through the grass, nothing.

The puzzle deepens when the maps are set side by side. The 1837 edition of the OS six-inch map does not mark the feature at all, yet a later edition, published in 1922, labels what appears to be the same location not as a holy well but as an altar. That discrepancy matters, because the altar in question turned out to be something quite specific: the remains of a cross-shaft on a pedestal and base. A cross-shaft is exactly what the name suggests, the upright section of a stone cross, and when one survives with its pedestal and base still present it can indicate an early medieval monument of some significance. The suspicion, then, is that the name "Holy Well" was simply an error, a misreading or mislabelling that sent the feature down one classificatory path when it belonged on another entirely.

What this leaves behind is less a place to visit than a lesson in how landscape history can unravel. A name on a map generates an expectation, the expectation generates a category, and the category persists even when the ground itself offers no corroboration. The actual stonework, the cross-shaft on its base, survives as a separate recorded monument. The well, if it ever existed, does not.