Bug: a88/376

This one might require a little thought. Anonymous 
reviewer writes:
‘The major flaw that was found is in the process of defining the set of
samples in each site, making sure the PCA analyses ended up in that site,
and then figuring out a simple way to view and analyze the site mean
results. This part of the analysis process is very cryptic, and not easy
to examine or modify. It would be a significant improvement if a data
table with sample name, PCA results, and the Fisher mean and stats were
also included in a window- preferable an interactive window (so data can
be coded, deleted, etc) along with the Site Equal Area plot. Although the
process would become more familiar with practice, I was never very
confident that the right results of the PCA analyses were being used for
the site mean (it took a bit of very close examination to determine this
was in fact working). The great-circle analysis was very good at the
individual sample level, but was also a bit cryptic when examining the
site-level or suite-level results.’
Since this is described as ‘the major flaw’ I should be thorough
in addressing it.

I have now added a table which shows sample directions within the
current site. It is slightly interactive, in that clicking on a line
will jump to the appropriate sample.

I have added a site parameter table to complement the sample
parameter table. Like its predecessor, it can change the current 
sample when the user clicks on it. I think that this should be
enough to address the concerns of the reviewer.