can multi-content-type in a message?

I want to program a cgi to transfer a piece of audio together with a text/html
back to the browser. The audio can be a explanation to the file or even a some
music. I need it because the file is in English and the audio explanation is
in Chinese, and the audio should be immediately played without a user to click.

So, can I transfer different types of contents in a single message? From the 
http draft I can see no hint whether it is possible or impossible.

Thanks!

Robert Wang