NavList:

Bruce you wrote:
"The index error was about zero so I don't intend to make any adjustment. Bauer calls this "side error" and it would require professional adjustment."

No, it does not require "professional adjustment" --really, nothing on a sextant requires a professional, except to the extent that someone with better than average skills chooses to label himself or herself as a "professional sextant adjuster" for marketing purposes. You, or any sextant owner, can remove 'side error' relatively easily if you would like to (and I go over this in all of my classes, including the one you were in many moons ago... forgotten already?? ...many moons do have that effect, don't they?). Do this adjustment after you check for index mirror perpendicularity (the order is: index mirror perpendicularity, then horizon mirror perpendicularity, and finally index error elimination). So-called "side error" is also not a source of actual "error" in normal practice. There's no need to reduce it to zero. Unless it's so large than it becomes inconvenient to make contact, it can safely be ignored. WARNING: Bauer apparently worked in a vacuum while writing his book --he did not have the advantage of a community like NavList to double-check his thinking. His book includes quite a few errors. 

You also wrote:
"I must say that checking the index against a star overlap seems easier than against a fuzzy horizon. Why is it not mentioned more often?"

Yes, it's a good way to check index error! There are lots of ways of checking the index error that are better than the horizon check. Why mention the horizon check so often? Because for a navigator at sea, it is the one reference that is always present, and sufficiently far away, for an index test, under circumstances when sights are possible. If you can't see the horizon, you can't take sights with a standard marine sextant.

Frank Reed