Hi Paul
When you write that it "serves the html source code" do you mean that the content is seen like HTML in a text editor rather than being displayed as a browser would normally see it?
If this is the case try different browsers to see if it is restricted to some of them or whether the behaviour is general.
When the browser requests a page like index.htm it know that it expects the content to be HTML. When, however, no file is given it doesn't know what content to expect and it may be using the mata data returned in the header of the HTML in identify the content. Check the header details (eg. <meta http-equiv="content-type"content="no-cache"content="text/html"> in the head of your index.htm to ensure that it is not confusing the browser, or is missing.
Regards
Mark