I'm sure Tahoma isn't the font that Google uses to display search
results- so when Google puts your page into its own fonts to display
as a search result, it will replace characters that aren't in the font
set that it employs with markers to indicate that there was a
character there that could not be displayed in the new font- usually
the little square you are seeing. You could try using HTML character
entities to replace the characters that have diacritical markings.
You can use character entities to display correctly for most
computers, and can use them to reproduce any character in the Unicode
standard (including diacritical marks and characters in other
languages, like Thai or Sinhala).

This link provides information about using HTML entities to represent
Unicode; it is specific to a collaborative site that I visit a lot,
but the majority of the information is generally applicable to the
web.
http://www.everything2.com/index.pl?node=Using%20Unicode%20on%20E2

I've used the method of replacing diacriticals with Unicode placed
with HTML entities to write proper diacritical markings for Pali
previously; it displays properly for most browsers, and I think it
will work with Google's result display (I've seen some ampersands
encoded as HTML elements that display fine).

You might also consider using Meta tags to help indicate to Google
what to display as a search result, if you aren't already.

Does anyone on the list know of a good set of links for this sort of
information? If some exist, it would make an excellent addition to
the link list, or to our theoretical Pali Wiki.

Clay Collier