The Chart Was Not Wrong
We weighed anchor at 04:50 and cleared the Napuka shelf with the first of the dawn in the east and Rahiti asleep on the port cockpit cushion with the trusting thoroughness of a young man who had been at sea since he could walk.
The bearing Teva had given me was 082 true, and the distance, by his estimate, was a fraction over seventy-three nautical miles. I had plotted it on the paper chart the night before, alongside the uncorrected bearing from Manning's 1790s sheet, which put the atoll about four miles further east and two miles further north. The difference was not large — it was, in fact, a testament to Juan de Aguilar's competence — but it was a difference, and I intended to steer to Teva's bearing, not Aguilar's.
The wind filled to fourteen, the *Maeva* settled onto the course under reefed main and staysail, and we ran. Rahiti woke around eight, drank coffee, watched the horizon for a long while in silence, and then — to my considerable interest — went forward and sat cross-legged on the foredeck with his back against the mast. He sat there for most of the morning. Eloa explained, when I asked, that Rahiti was reading the swell. His grandfather had taught him. There are in the open ocean between atolls three or four dominant swells running from different quarters, and their interference patterns — where they meet, where they are masked by a rise in the seabed, where a