What does the rel = "index" link do?
<link rel='index' title='mysite' href='http://mysite.com/' />
It states that the linked document contains an index for the current document ... which is probably incorrect.
<link rel='canonical' title='mysite' href='http://mysite.com/' />
It states that http://mysite.com/ is the same page as the current document, and the correct URL is http://mysite.com/ .
From the HTML4 specification :
Index
Refers to a document containing an index for the current document.
So this will be your main page. (This is the same reason we call "index.html".)
And the Google Webmasters blog says "canonical" provides the preferred URL for the page.
The rel attribute in the link element simply says: "there is this other document whose url is the value of the href attribute associated with the current document.
There are many types of values โโused for rel , the most common of which is stylesheet . In the case of "index" this means that this other document is somehow indexing this document.
There is a list of common rel values, from all places, w3schools (http://www.w3schools.com/tags/att_link_rel.asp). (I'm not a fan of w3schools, but I thought you might like the list of rel values โโthey provided.)
Here is information about Google and rel = canonical: http://www.google.com/support/webmasters/bin/answer.py?answer=139394