Never mind!! Basic html sins being committed led to Chrome doing its best to correct the html. Have now fixed.
Hi,
Couch is concerned only with outputting the raw HTML that the web server hands over to the browser - when you do a 'view-source:', this is what you see. If that HTML appears properly formed, Couch's job is done.
If, as in your case, the outputted HTML appears to have been further processed upon display, that has definitely been done on the client-end where there is no Couch to affect anything at all (it runs entirely server-side, as you know).
The usual suspect is of course JS. Please try commenting out (using <cms:ignore></cms:ignore> tags) the parts that you think could be responsible to pin-point the culprit and rectify the problem.
Hope this helps.