NavList:

I recently bought my first sextant, a second-hand Cassens & Plath 'Bobby
Schenk'. While trying to adjust the perpendicularity of the index mirror
using the usually recommended procedures, I found myself getting
increasingly confused. I think I have the correct procedure figured out,
but I hope someone can check my reasoning below, because it would be
surprising that all the well known manuals are incorrect or incomplete
on this common procedure...

As far as I understand (after much drawing of diagrams), the trouble is
with sextants whose index mirror is not centred on the index arm pivot
axis. On my sextant I'd estimate this offset to be 4 to 5 mm, in the
direction of the eye when looking into the mirror. This changes the view
enough that the most commonly referred to procedures don't describe what
I am seeing, and give confusing and/or wrong instructions (1).

Here's what I see when the mirror is misaligned (top angled forward
slightly, towards the eye).

When I look into the index mirror and angle the sextant in such a way
that the mirrored section of the arc appears in the lower right corner
of the mirror, the mirrored arc appears slightly _higher_ than its
continuation to the right of the mirror.

If I now tilt the sextant so that the mirrored section of the arc
appears in the top right corner, it appears _lower_ than the part of the
arc seen beside the mirror.

My first impression was that the mirror wasn't flat! I then figured that
I should probably aim for a good average, with the arcs lining up in the
centre of the mirror. Or perhaps I should align the arcs at a location
slightly higher on the mirror, along the reflected axis of the
telescope. All wrong of course.

If I am not mistaken this effect is not seen in sextants that have an
index mirror exactly on the centre of the arc. In these less exhausting
instruments, I imagine, a well-aligned index mirror will always show a
continuous arc, whatever the orientation the sextant is held in (at
least, I have convinced myself this is true for reasons of symmetry). A
misaligned index mirror will show a stair step effect commensurate with
the angle of misalignment, consistent whichever way you orient the
sextant. I hope this is correct so far, I haven't got such a sextant to
check this on.

On my sextant however, due to the offset of the mirror even a
_correctly_ aligned mirror will show a stair step effect growing in size
towards the top of the mirror. Only at the bottom of the mirror, in line
with the top plane of the arc, will the stair step effect disappear. But
this becomes hard to observe since the mirror doesn't extend down far
enough - hence the need for placing dice or some such on the arc. The
top plane of these objects being parallel to the plane of the arc, but
easier to view in the mirror.

When the index mirror on my sextant is misaligned with the top angled
backward, the mirrored part will always appear too low, though more so
at the top of the mirror.

The most confusing situation is when the index mirror is angled forward
just a little bit. This make the mirrored arc appear too high at the
bottom, but too low at the top of the mirror. When the angle of
misalignment becomes larger, one effect overtakes the other, and the
mirrored part appears always too high.

In summary, the correct procedure (in short form) according to me is:

1) place dice or some such on the arc (to construct a plane parallel to
the arc)
2) change your view so you are sighting along the top surface of the
unreflected die
3) adjust the mirror so the top surface of the reflected die lines up
with its sibling

* This procedure will work correctly for sextants with mirrors offset
any distance (including zero) from the index arm pivot
* Only not use dice if you're certain the mirror is not offset, or if
the sextant is constructed in such a way that you can sight in parallel
to the top plane of the arc.
* It's not correct to use the top _edge_ of the dice. One must be
looking along the top plane.
* A stair step effect between the unreflected and reflected arc does not
necessarily mean the mirror is out of alignment.

Did I get this all correct?

Luc


(1) I consulted:
* Bowditch, ed. 2002, p 267 (which is clearly wrong for this sextant),
* Bruce Bauer's often recommended 'The Sextant Handbook' (which led my
to wrongly try and line up the arcs in the center of the mirror, though
this is followed by the correct procedure using dominoes),
* Tom Cunliffe's 'Celestial Navigation', which only works for non-offset
index mirrors
* A Tamaya manual, not helpfull
* A Davis manual, same
* A SNO-T manual, which gives the correct procedure using the placement
of 'diopters' on the arc. But these diopters appear to be L-shaped (see
picture at
http://www.cruisersforum.com/forums/f121/questions-about-a-russian-sextant-112909.html
), with a sharp edge at the top instead of a flat surface, which now
makes me doubt again if I got everything correct.