In the misty past I tried a hacksaw. This was very laborious and yielded
disappointingly messy results.

Shells would be too hard for most bandsaw blades, except perhaps for a
metal bandsaw. I wouldn't want to hold the shell while the blade whizzed by
my fingers, and I bet you couldn't get the saw's clamp to hold the shell
securely.

My own suggestions are: for large solid shells try a lapidary saw. For
small or delicate shells a lapidary saw is too coarse, and will tend to
catch and chip or break the shell. All of the lapidary saws I have used run
with water, either a flow, or a bath if it's a big saw.

I have had good results using a diamond lap. With this you can wet-grind
one side of the shell away bit by bit until you are satisfied. At all
stages you can see where you are going and stop just where you want. This
method is exact and also quick. You can then reverse the specimen and lap
the other side off, leaving an outline of outer wall with the columella up
the center, if desired. Such laps run with a slow water flow.

A thin-section saw such as a Leco VariCut will also give good results with
delicate specimens. This runs in a kerosene bath. The shell can be clamped
or glued to a microscope slide. The thin-bladed saw works on a specimen
held by a counterbalanced damped arm and can be set to almost zero
pressure.

Very delicate shells such as Spirula must be permanently set in some solid
medium eg polyester or plaster of paris. even with a lap, expect some
chipping of septal walls. If you find a solid that can be dissolved to
release the specimen afterward, great; let me know what solid it is.

The specimen should always be small enough to be cut/lapped in one go.
Otherwise you will end up with a step where the cuts mismatch.

If you have a university with a geology department handy, the responsible
technician there may be happy to help. I have had good results with all of
these machines. I'm happy to help when peolpe approach me for such things,
and i would hope that other geological technicians would be likewise.

If desired, finish with a wet-sand on wet and dry sandpaper set on a piece
of plate-glass. Water will hold the paper on the glass. 1200-grit (or 2
pieces of 600-grit that have been rubbed together) will give the cut edges
a good polish. Careful not to press too hard and break the specimen you
worked so hard to cut.

Andrew