Works better here ... although I'm getting a message of "Wide character in print at z line 10." Because perl is converting this to UTF8 successfully, the ord is coming out as 8217. The 226 you're getting is simply the first character in a multi-byte character because your perl is treating the multibyte character as a series of single byte characters.

You may try something like this:

C:\> set LANG=en_US.UTF-8
C:\> perl html_entities_test.pl
[download]

I'm not sure how well it'll work for you, though.