NavList:

Wow. I want to imagine you recording all this with an ink quill in a great parchment book, because that is a work of devotion! Very impressive. I especially like that you investigated micrometer eccentricity. I have tested for it on a few sextants. I haven't yet found one with convincing micrometer error, but I'm convinced it's more common than most navigators would expect.

I suspect that a navigator interested in doing a calibration like you describe could get by with far fewer observations, right? If we suppose we want a reasonable value every 15° from 15 through 120 that's eight entries. To get a stable value by lunars, we would at least five sights at each angle, so that would be a minimum of forty actual sights. I suppose there's a number somewhere between 40 and 3000 observations that would be an ideal choice. :)

Frank Reed