NavList:

Antoine, you wrote:
"On the other hand, when shooting Lunars, and by instinct I have always attempted to use a sextant with side error as small as possible, i.e. without accepting any visible side deviation."

And my point in my earlier post is that this is unnecessary. Even if you have 5 minutes of arc of side offset, for any lunars with distances greater than 10° the resulting "side error" is less than 0.05' of arc which is completely negligible even for lunars. And since the error is inversely proportional to the tangent of the observed distance, the potential error is less than half an arcsecond for all distances from 45°-120°. I realize that we get carried away with accuracy when playing around with lunars, and that's part of the fun. But when you get down to a single arcsecond, this is no longer a rational concern. Side error is not a real error.

And just for clarity, for anyone who avoids even the thought of lunars, these numbers also apply to ordinary altitudes. If you measure the altitude of the Sun to be 45°00.0' and if your sextant has a nice wide, plainly visible side offset of 5 minutes of arc, the so-called "side error" would be only 0.009 minutes of arc --well beyond the accuracy of manual sextant observations and well beyond the precision of the micrometer. Completely negligible. Imagine how long it would take for the Sun to change its altitude by that small amount... Suppose you're in mid-latitude and the Sun is rising at a rate of one minute of arc every six seconds. To time a sight expecting an altitude accurate to the nearest 0.01 minutes of arc (which would be a minimum requirement for such a tiny side error to be legitimately observable in any way), you would need to record your UT (and distinguish UT1 from UTC!) to the nearest 0.05 seconds. That's a twentieth of second, scarcely more than the time between two frames of a film or TV show (24 fps). That "flash of an eye"... that's how small the so-called "side error" is for this Sun altitude. And note that this is only one issue that affects altitudes at this ultra-fine level. How far would your sextant have to be out of vertical to yield an error of 0.01 minutes of arc?? 

But what about the books? The textbooks of navigation seem worried about side error. Some even say it's critical! Books can'te be wrong, can they? Not so many of them collectively, right?? And of course they can be wrong, because they copy from each other and from the collective common knowledge that is the culture of navigation. Celestial navigation has many myths and also bits of bad advice that have been collected over the decades which have become "common" knowledge. Common ≠ True.

Frank Reed