The Cartographer's Mistake

The island appeared on charts for a hundred and fifty years, and it never existed.

It was entered in 1826 by a surveyor named Halloran, working the southern approaches in poor visibility, who recorded a landmass at a position he was reasonably confident of and a size he admitted he was not. He noted his uncertainty carefully in the margin of his log, in the way that careful people do, and the note travelled no further than the log.

What travelled was the island. It was copied onto the admiralty chart the following year, and from there onto the commercial charts, and from those onto the school atlases, each copyist reproducing the outline of the one before with the small unconscious improvements that copying introduces. By 1870 it had a coastline of some detail, two bays, and a name.

Three vessels searched for it in the 1880s and found open water. Their reports were filed, considered, and quietly set aside, because the chart was old and the ships were new, and a chart that has been right for fifty years acquires an authority that a single negative result does not easily overturn. Absence, after all, is what you would expect from a crew that had looked in slightly the wrong place.

It was removed in 1974, not because anyone finally searched properly, but because a survey being done for another purpose passed directly through the position and recorded four thousand metres of water.

The interesting part of the story is not the original error. Halloran did his job, saw something in bad weather, wrote down what he thought and how sure he was. The interesting part is what happened to his uncertainty. Every step in the chain preserved his observation perfectly and discarded his doubt, and by the end there was an island on the map with a name and two bays, and no trace anywhere of the man who had said he was not sure.