NavList:
A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
Re: Perpendicularity and other qstns.
From: Alexandre Eremenko
Date: 2004 Oct 13, 16:07 -0500
From: Alexandre Eremenko
Date: 2004 Oct 13, 16:07 -0500
Dear Herbert, I already admitted that my message you refer to was incorrect. Maybe worse, it was arrogant, and I apologize for this. And indeed, I should have thought more about messages on this subject before I wrote my own. I found my mistake as soon as I strated writing mathematical theory of this experiment, on Frank's request. In the process of writing, I realized that if the axis of rotation of the index arm passes through the plane of the index error (which is probably the case for many (most?) sextants) then the effect I described will be be absent. I repeat that I have never seen another sextant except my SNO-T. Then I saw the message with Frank's experiment on the kitchen floor. By that time I already did not need to stage this experiment myself because mathematical conclusions were in complete agreement with the experiment described by Frank. Only then I realized that your suggestions made before Frank's experiment were right. And what I was saying about "all sextants" was wrong: what I was saying only applies to some sextants, like SNO-T. So next time I will try to be more cautious in my statements. The mathematical demonstrations I have at this moment are of the following: 1. If the axis of rotation of the index arm passes through the reflecting surface of the index mirror, then the perpendicularity test as described in most books is correct, and it does not matter from what height you look. 2. If this axis does not pass through the reflecting surface, the test should be modified. This is the case with SNO-T and maybe with some other sextants. I am still not sure that this mathematical "theory" has to be posted on the list, because its conclusions apparently coinside with the "common knowledge" of the members of this list, and the "theory" adds nothing new to it. But of course I will post if there is still any need. So far I only have one practical recommendation for SNO-T and other sextants with this similar property (that the axis of rotation does not pass through the mirror surface). Put your sextant on the table horizontally. Set the index arm on 35 degrees. Put one visor on 0 on the arc, so that its front surface is aligned with the inner edge of the arc. Put the second visor on 120 deg on the arc, so that its front surface is 5mm BEHIND (that is "to the outside" of the arc) this inner edge of the arc. Now the test can be done in the usual manner. That is the height of the eye becomes irrelevant. I find this easier to do than to keep the height of the eye precisely. Which I tried for some time yesterday. (Using standard SNO-T visors, it could be easier with Celestaire cylinders. Piles of coins did not work: two piles of 16 dimes were found of unequal height, apparently because some dimes were bent, I could not determine which ones, and thus abandonned the dimes. I don't have dominoes or dice at home, so I cannot say anything about them). The recommended 5mm distance should be determined for each sextant experimentally. I found 5mm appropriate for my SNO-T. Alex.