How is a 404 custom error page requested by browser?

view story

http://serverfault.com – I am trying to understand how the communication works between the browser and the server in the case of custom error message. I understand a 301/302 header which contains the URL of the page in question, HTTP/1.1 30x .... Location: http://www.example.org/ but the 404 header does not (as far as I can see). What tells the browser to go get the error page? I do not have a network sniffer and the question itself generates a huge amount of noise on google Here is what I see when I trace http: GET http: //myserver.com/unknownpage HTTP/1.1 Accept: image/gif, image/jpeg, image/pjpeg, image/pj (HowTos)