NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Star to Star Distances taken on a Second Hand Sextant
From: Peter Monta
Date: 2019 Dec 7, 23:58 -0800
From: Peter Monta
Date: 2019 Dec 7, 23:58 -0800
Paul Hirose writes:
That's what I did, the few times I have tried to observe star separation angles. A steady superimposition of both bodies doesn't work for me. To my eye it makes the error more difficult to perceive. I prefer to repeatedly sweep one star past the other.
Yes, same here. It also helps if the star images have low contrast, with a background field full of light, either from daylight, twilight, or artificial light. Then just the central cores of the stars' light can be swept past one another, with no distraction from close-in aberrated or diffracted light. With a bright star on a black background, the mess of faint spikes obscures the core and just kills the precision.
Simms mentions stars for determining index error in "The Sextant and its Applications". He says:
Coincidence of images of a star, however, may be observed with very great nicety; and if we select for the purpose a bright one, and make our observations of coincidence when this first becomes visible in the twilight, we shall find a number of readings rarely exhibiting a difference of more than 3 or 4 seconds between the extremes. This then is the method I am inclined to recommend; and on repeating the operation in the course or at the conclusion of a series of observations, should the sky have darkened in the meantime, a star of inferior magnitude may be taken for the purpose, but still not so small as to occasion any unpleasant exertion of the eye to perceive clearly the coincidence of images. The coincidences should in all cases be made by alternately elevating and depressing the reflected image, so that, the first being obtained by elevating the reflected image, in making the second we should depress it, and so on. The mean of an even number of readings, not fewer than eight, will, I believe, seldom differ 1" from the truth.
Since he's talking about index error, there's no question of sweeping: the two stars are stationary relative to each other. (Presumably the side error is adjusted to zero in advance.) But his remarks at least point to the possibility of precision of a few arcseconds for star-star distances under ideal conditions.
With a camera capturing the small streaks of the two stars during sweeping, the separation normal to the streaks could be measured with a precision far better than by eye.
Cheers,
Peter
