NavList:

In all my years of NavList, I have never seen anyone ever finish the calibration of their arc by this method.  That's a pretty large number of years now.  

We have discussed this topic, ad infinitum.  The concept appears in navigation tome after navigation tome.  The consensus (from what I gather) is that the method may be useful after you drop your only sextant when out to sea to confirm that you are getting reasonable results.  That is, the star to star distances are in general agreement to the computed result, so it is okay to trust the altitudes for navigation.  But for calibrating the arc?  I am afraid I must burst your bubble.

Here is an interesting challenge for anyone, but in particular, you to try.  Sometime over the next week, pick out any pair of stars and measure the distance.  Do it as many times as you like.  What you will find is "general" agreement with the calculated distance, but a distribution that shows large variation.  In other words, good enough to show the sextant functional but simply not accurate enough to "calibrate the arc".  Certainly not to 12" (0.2').  

In all my years in industry in which I provided NIST traceable calibration of rotary devices, I have found that you need roughly 10× the accurate resolution in the calibrating device in order to prove the calibrated rotary device to a given value.  A sextant is a rotary device, in which we attempt to measure the part of the circle, subtended.  In the example, you state that star to star distances will provide 0.2 minutes or 12 arc seconds.  10× means your calibration measurements need to be on the order of 0.02 minutes of arc or 1.2 arc seconds.  That simply isn't practical with a sextant.  A modern sextant is graduated to 6", the antiques to 10".  The measurement of calibration cannot suffice under such a regime to provide the requisite accuracy or resolution.  That's my hard won opinion, built over years of calibration for industry.  Your opinion may vary.

Elsewhere, you mention the cost of having a laboratory performing the calibration.  May I ask which laboratory you are referring to?  To my knowledge, they are all gone, long out of business.  The manufacturers no longer provide any arc calibration, but merely state the arc is sufficient for navigating.  I agree with that statement, the error due to arc will be small compared to all the other vagaries CN is subject to.  So the sextant you buy new today is good enough.  But it is simply not calibrated.

That's enough for now.  Pick out a pair of stars and do the experiment.

Brad