NavList:

Douglas Denny, you wrote:
"I should have said: Measure the star distance as well as the altitudes; compare with calculated star distance and you have index error."

Not quite. Maybe you just mis-spoke again --happens to the best of us-- but the goal is not to get "index error". For index error, you would want the star-star distance equal to zero (in other words, superimpose a star's image upon itself in the sextant).

The whole point of measuring (non-zero) star-star distances is to get the "arc error". This is variable along the sextant arc. It's a function of the angle. The idea is to generate an approximate calibration table showing errors at every thirty or ten or even five degrees along the sextant's arc. Note that any measured star-star distance error can either be considered a combined error = arc error + index error, or a separate pure arc error, which by definition would be zero at zero degrees. Since the pure arc error is probably much more stable, it's probably better to correct each observation for index error as usual before comparing observed star-star distances with true star-star distances. But in the end, the arc error is just like an index error that depends on the measured angle.

And again, this is just ONE WAY of measuring arc error. It can be very effective for people with good eyes, high-power telescopes on their sextants, and maybe some means of stabilizing the instrument. My results with star-star distances have generally been disappointing compared to lunar distance observations which are consistently more accurate. Of course different optical tasks involve different capabilities in the human vision system, and it may be nothing more than that.

-FER

----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------