NavList:

Josiah D you wrote: I am planning on acquiring a refurbished bubble sextant (i.e. MK IX or A10) to navigate simple cross countries in a Cessna 172.

I speak as someone who used bubble and pendulous reference sextants in the air professionally for 15 years between 1967 and 1982.

Firstly, a note on safety.  You don’t say whether you intend to complete these flights solo or accompanied.  If you intend using a bubble sextant, your attention will be away from the aircraft for at least a minute at a time.  Even if this Cessna 172 has an autopilot, you still need to maintain lookout unless you’re under radar advisory service at the very least.  If you can’t comply with these requirements, you might be better off using a box or marine sextant.  Such use was indeed possible and was used well into the 1920s and later.  The only reason for the introduction of bubble sextants was because as people extended the performance of aircraft, the natural or cloud horizon wasn’t always visible.  Calculating dip wasn’t a problem so long as you knew your true height above the sea or stratus cloud.  If you were alone, you could set Hc, take a quick peep holding the sextant with one hand, and estimate your intercept in Sun or Moon diameters.  If you chose the correct height to fly you could make dip and refraction cancel semi-diameter and have no corrections. 

Next, a note on dome/ windscreen refraction.  The best surface to shoot through is fresh air as with an airship or open cockpit aircraft.  The next best is totally flat glass or Perspex.  The standard dome refraction table has been around for so long that I’ve never been able to find out how it was constructed.  Was it calculated mathematically, or did somebody just take a spare dome off the shelf and calibrate it.  We don’t even know if this standard dome was blown or pulled over a grease covered pattern as they do with glider cockpits.  The 172 windscreen appears to be a complex 3D shape, which I assume is pulled.  In any case, refraction is going to vary considerably over its surface depending upon where the body to be observed lies.  You can’t always change course so you’re always pointing at it. 

Now a note on accuracy.  Don’t be too taken in by others claimed results.  People tend to quote their best and forget their worst.  During training 10-30nm is not uncommon.  Once trained and experienced consistently under 10nm ought to be possible apart from the occasional outlier.  Fixes consistently in the order of 3nm is considered very good indeed.  1nm or less is pure simple luck.  Why the poor accuracy compared to the marine sextant?  It’s mainly because of tiny acceleration on the bubble or pendulous reference cause by changes in speed or heading of the aircraft or change of position on the Globe (Corioli’s acceleration).  The effect of heading and airspeed changes can be reduced by averaging, but this takes time, which means there’s less time for lookout, and you only ever find out where you were several minutes previously.

You also wrote:   My concern is that the limited view out the front windshield would prevent me from shooting stars at high angles, unless I put the aircraft into a climbing pitch while taking a sight.  To see higher in a Cessna 152 you would need to either fly very slowly to increase the angle of attack (AoA) or put the aircraft into a climb.  At high AoA more concentration in flying would be required and you would be flying on the wrong side of the drag curve so speed stability would suffer.  Don’t ask me why, but for many aircraft longitudinal static stability reduces or sometimes disappears in the climb which also affects speed stability and hence acceleration errors.  You might perhaps us an aircraft with a Perspex roof or no roof at all.

To summarise, practice in the aircraft on the ground first, it's less expensivet; be accompanied; if you fly VMC consider using a box or marine sextant; start with the Sun, leave other stars and planets till later; keep on the right side of the drag curve; and don’t expect consistently brilliant results immediately (I really mean ever). DaveP