I pretty much agree with all that Carl Meyer writes. Wonder how Stack Overflow handles "me too" responses? :-)
The original question doesn't give much specific information about the intended use - expected load / requests per second, hosting type, etc. Assuming that the load is small'ish, I would recommend using just Apache with mod-wsgi, or Apache with mod-wsgi and S3 for static file serving.
mod-wsgi recently became the recommended way to deploy Django, see http://docs.djangoproject.com/en/dev/howto/deployment/modwsgi/ . Given this status, you should expect this method to see a lot of usage, and therefore a lot of documentation, tips'n'tricks and blog posts. Graham Dumpleton also recently release mod-wsgi 2.4 which looks quite mature (I haven't used it yet).
You can put nginx in front of Apache to proxy dynamic requests, and serve static files. This setup serves many people well. The nginx wiki has good tutorials on setting this up, http://wiki.nginx.org/Main . But this also gives you two sets of config files to learn, and two HTTP servers to keep updated.
Using only Apache, you could use Apache to serve static files, use mod-wsgi in deamon mode to handle Django. Then split up the requests by URL so that static file requests do not invoke the WSGI / Django stack, but are handled by Apache alone. This is again well documented, see your favorite Apache book for location handlers, or see the Apache / mod-wsgi docs. This at least 'centralizes' your learning, so that you only need to learn Apache & mod-wsgi configuration, and to keep Apache & mod-wsgi updated.
Another simple way to go would be to use Apache + mod-wsgi for all dynamic Django content, and put all static content on f.x. Amazon S3. Depending on your expected use / frequency of updating static content, this might be the very best. Amazon lists a couple of freeware applications to make working with S3 simple, such as S3Fox, Bucket Explorer etc.
HTH,