This java method cleans up common charset conversion issues. As I understand the issue, if you have utf-8 sequences and render them improperly as iso-8895-1, you’ll see these characters:  and ’. The utf-8 came from cut and paste from MS Word, in my case, so perhaps the form post wasn’t setting the charset properly? Unfortunately, IE renders these sequences properly even though the page’s charset is iso-8895-1, so the web site developers aren’t aware of the problem. I come in as a third-party consumer of the web site, so I have to live with their sins.