Which?

I doubt. They might be shown as *you intended* in certain broken
browsers but that does not mean "correctly". View the pages in Mozilla
http://www.mozilla/org/ . Give us a URL if possible.

Probably because it *is* rubbish. Please read carefully
http://ppewww.ph.gla.ac.uk/~flavell/charset/fontface-harmful.html
about the abuse of the infamous Symbol font.

I'm afraid, no. You did the wrong thing by using the Symbol font.

Again: Internet Explorer might show them *as you intended* but this is
not correct, it's wrong.

Which?

I doubt. They might be shown as *you intended* in certain broken
browsers but that does not mean "correctly". View the pages in Mozilla
http://www.mozilla.org/ . Give us a URL if possible.

Probably because it *is* rubbish. Please read carefully
http://ppewww.ph.gla.ac.uk/~flavell/charset/fontface-harmful.html
about the abuse of the infamous Symbol font.

I'm afraid, no. You did the wrong thing by using the Symbol font.

Again: Internet Explorer might show them *as you intended* but this is
not correct, it's wrong.

I'm a bit confused by what you report here.  Are you saying you cannot
view a *web* version of the document?  - then that's not surprising,
for the reason that Andreas has explained.  Or are you saying that you
can no longer successfully *print the Word document*?  I would find
that rather surprising, since this technique, although foreign to the
WWW specifications, is one that has been used quite deliberately in
Word for many years.

But the web doesn't work like that (at least, it might have previously
given an impression of doing what you intended, but that could
justifiably be described as a bug, and browser developers are tidying
that up now.)

Possibly...

Background: I've had a long-standing argument with the author of the
"tth" program, on what was effectively the same topic.

At http://hutchinson.belmont.ma.us/tth/manual/sec12.html#tth_sEc12
you can find the author's perverse presentation of how to go about
breaking web browsers so that they do what he intended, rather than
what they're specified to do.

Some years ago I lost patience with that, and wrote a Perl script to
read-in his <font face=Symbol> rubbish and turn it into proper
Unicode references.  This is specifically for HTML, you understand:
the script won't do anything for Word or RTF documents as such.

I should stress that the script was quite simple-minded, and tuned
rather exactly to the syntax that his program produced at that time
(as you can see from his updated documentation, the program now has
an option for doing that by itself, so my script is now redundant).

The script definitely needs a workover (it dates from 1997, the
comments say), and would need to be adapted to the syntax which MS
produce as web format; but if you're interested you'd be welcome to a
copy, e.g per email.

I should also mention that several of the Symbol font characters
formerly assigned code points in the PUA (Private Use Area) have been
given their own official Unicode points in later Unicode versions.  
There are code tables online which can help with that transition:

http://www.unicode.org/Public/MAPPINGS/VENDORS/ADOBE/symbol.txt
http://www.unicode.org/Public/MAPPINGS/VENDORS/APPLE/SYMBOL.TXT

hope this may be useful