Glacial update

Should be 'quiet' not 'silent'

Rather than calculating the isotropic velcoty
from the isotropic component of the poly-xtal
anisotropic tensor (the norm of which changes
with texture) use simple VRH mean of the single
xtal tensor. This gives values that are just a
function of the temperature and do not vary with
texture.

Also split the velocity graph into two graphs.

Choose the format based on the first word
in the first line of the EBSD files. Data
is the same in both, but colomn order
changes.

These are still straight but have a inclination
that is not 90 degrees. Plot as a function of
source polrization (otherwise we end up with
chaos as we sweep around the azimuth) and
inclnation for different azimuths.

Some cleanup is needed.

If 'fname' is a real (rather than a string) write
to that assuming it is an fid (and avoid opening a
new file). In this case the file is not closed
prior to exiting MS_save. This can be used when
either several Cijs must be written to the same
file or to create parts of larger reports.

Note that 'fname' = 1 will write data to the
terminal.

Create an input file for MODSIMPLE - a model builder
for the ATRAK anisotropic ray tracer.

Aparanlty we had the wrong units for modsimple.
According to Mike this is correct.

Add these to the summary plot. We should
probably refactor MS_plot and make this an
MS_summary or something.

These should correspond to SKS, local events and
icequakes.

We do not need a depth for the top of the model -
so just print the depth, not zero.