Reading non-ascii characters, recognise an encoding

[Andrew Torda <torda@zbh.uni-hamburg.de>]

I sometimes have trouble (maybe always) reading mail from some
Europeans with their representation of non-ascii characters.
Typically, a word looks like
   f=FCr
whereas they intend it to look like 
   fur
with two dots (umlaut) over the "u".
Their mail header says that the message is iso-8859-1 encoding,
so I read all the info pages and tried every command I could find
which would force the buffer to that encoding.
It does not seem to make any difference.
I found an info page which suggested I should add a line at the
top of the file, like
-*-coding: iso-8859-1-*-
but closing and re-visiting the file did not seem to help.

Less important is that their mailer puts an equals "=" sign at
the end of every line. I suspect this is part of their machine
trying to encode the message.

Is there any correct command that can persuade emacs to eat all
the =XX type representations and convert them to something more
readable ?
This is happening with emacs 21.1.
Many thanks
Andrew Torda