The word "bit" is a portmanteau word for Binary digIT
en.wikipedia.org
en.wikipedia.org
The paper is well worth reading if you haven't already done so.
A Mathematical Theory of Communication
(http://cm.bell-labs.com/cm/ms/what/shannonday/shannon1948.pd...)Especially page 5, which is a very easy to understand description of EDIT markov chains. But you can use that letter frequency thing to do a really nice easy description of compression.
Obligatory http://xkcd.com/739/