A Community Devoted to the Preservation and Practice of Celestial Navigation and Other Methods of Traditional Wayfinding
From: Frank Reed
Date: 2011 Oct 3, 04:46 -0700
Perfect early fall weather here in Connecticut yesterday afternoon. There were sailboats by the dozen, folks fishing on the rocks, people strolling on the beach, and me, calibrating a sextant.
I've described this method over the years to a number of people privately. You find one good sextant with an established calibration (good luck on that, but once you've got one, the rest is easy). You set that sextant to some angle, for example thirty degrees. You remove its scope and place it on a flat surface, like a table with the normal "eye" side aimed at some suitably distant marker. It is now a source of thirty degree deflected light. Then you place the sextant you want to calibrate on the same surface with its index mirror facing the index mirror of the good sextant. Note: the sextants have to be leveled with a little shimming, but this is not difficult in practice. You set the second sextant to the same angle as the good one. It sees the deflected light from the first sextant and also sees the direct view of the distant marker. So the view through the second sextant is exactly like a standard index correction test of a sextant set to zero. You see two images of the same object nearly superimposed. Now adjust the second sextant until the images are exactly superimposed. The reading on the micrometer is the arc error at that angle (really, the difference from the good sextant). Finally, to make the whole process really accurate, instead of using the second sextant's own telescope, remove it and place a medium-power spotting scope in line with the normal position of the sextant telescope. A 20x telescope yields good repeatability to a tenth of a minute of arc.
My thought process getting to this arrangement might entertain some of you. Years ago, Alex Eremenko (if I remember correctly) posted a link to a patent from about 1905 for a kit for testing sextant calibration. It consisted of a series of standard prisms which would produce light deflected by a particular angle. The prisms were meant to be placed above the index mirror of a sextant, and then the calibration proceeded like a standard index correction. I hunted around for prisms for a little while, but eventually decided it wasn't worth the time or the money. Perhaps a year later, it dawned on me that we could do the same thing with a pair of mirrors mounted perpendicular to the face of a metal plate. They would be fixed permanently in place and each such mirror device would be a standard for a particular angle. I made some sketches and set that aside for a while. Then some months later, I looked at my drawings and thought, "why not mount one of the mirrors on a rotating bearing so that we wouldn't need a whole set...? and then we would need some sort of scale to read off the angles... Yes, that's it!" I reasoned that I could build a metal frame with two mirrors mounted perpendicular to is face, one of which could turn, and the whole thing would have a calibrated scale of angles! And, of course, right about there, I looked at a sextant on a nearby table and realized that a good sextant with its telescope removed IS ITSELF this tool for calibrating another sextant. This all happened three or four years ago. I've experimented with it many times since, and it really works remarkably well.
I'm attaching some photos of this arrangement taken yesterday. My "distant marker" in this case is a water tower just about 9.9 nautical miles away. At that distance the error from parallax between light paths is about two seconds of arc, so below the measurement threshold of the sextant. The water tower, by the way, is located on Plum Island off of Long Island, New York. If you visit the Wikipedia article on Plum Island, you can see the water tower in the photo. You can also read about the interesting history of this island: it's the "Area 51" of animal diseases...
The short calibration table for the sextant I was testing (one of those "Tamaya-likes" from the late 1970s) reads as follows: 0d: 0.0', 15d: +1.0', 30d +1.0', 45d: +0.2', 60d: +0.1', 75d: -0.3', 90d: -0.4' (these are the corrections to be applied, not the error). The printed calibration included by the manufacturer was a string of zeros at all angles --I never believed that! The nice thing about arc error is that it is a completely correctable error. For this sextant, I will now add a minute of arc for any angle measured around 15 degrees or 30 degrees. Note that if you have a lot of spare time and patience, you can easily do this at a finer resolution (every five degrees? every two degrees??). All of this assumes that we start with a "gold standard" for the other sextant. The one I was using here is a Plath from the 1950s. It's very good.
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