NavList:

I don't have very much to add to the very excellent advice given by the fine gentlemen before me but I do have a small thing or two to add.

First, as previously mentioned by the other fine gentlemen, finding a way to stabilize the bubble is very important to determining IC. I have tried bubble shots using the C+P bubble attachment while holding 1 of the sextant legs in a cradle supported by a tripod and it was not especially accurate - I could get LOPs randomly anywhere from perfectly thru my actual position or up to 7nm off.  Given that the C+P nautical sextant I had it on was a known entity, giving me perfect LOPs when using it with an actual horizon, I ascribed my difficulties to the vagaries of the bubble.

Secondly, take a number of shots and average them.  I would take 8-10 shots on an object (Sun or star at any time of day) and then average them.  The easiest way to average is, of course, to get some graph paper and plot time vs altitude then eyeball a straight line thru them - pick a time on the time axis and an altitude off the altitude axis and do a normal sight reduction using these results.  I have been told that these newfangled computer doodads can also do this automagically (something called a "spreader sheet") but others may know more about these infernal machines than I do.  After a number of these sessions, an IC starts to stand out.

To make the plot more accurate, you can compute the time-vs-altitude rate of change of the object rather than just eyeball a line thru them.  Using Pub229, this is relatively straightforward - LHA advances 1° in 4 minutes, flip the page from your computed LHA and read off the Hc and compute the difference of the Hc from the computed LHA page and the 1° advanced page and divide by 4.  To find the slope even faster, you can find Motion of the Object tables in the Air Almanac or use a handy Polhemus whiz wheel.

Note that averaging can also be done on something relatively stationary like Polaris, as well, if you are able to see it.  And, if you are quick (having a lovely assistant recording times and altitudes for you speeds things up considerably), it works at LAN, too.  It may not be as necessary when the object is not moving in altitude but it makes me feel better since my eyes and bubbles don't always agree on what is centered from shot to shot.

As an aside, I do find that averaging works best when the object is moving in altitude - I am in the Northern Hemishpere and averaging shots on near-S azimuths don't work as well but near-N azimuths do.  I have always interpreted that as a failing on my part rather than the method.

I use a Kollsman periscopic sextant most days when I can see the sky.  I asked a very talented friend to 3-D print a mount for me that I clamp onto a tripod and I find that the results taken on this very stable mount are excellent.  The nice things about the Kollsman's is that it averages for me greatly speeding up the task as my lovely assistant tends to get bored with my ceaseless efforts to determine where exactly our back porch is on the globe.

I hope this helps!  And good luck!

-- Arthur