Beginner question: applying Mercurial

[Jeffrey Barish]

I am a beginner not only at Mercurial, but at RCS in general.  I am starting 
to see how to apply Mercurial in my small, one-person project, but I am 
still unclear on one scenario.  I have my project in a repository.  I have 
tried making changes to files in the working directory and then committing 
them when I am satisfied that they are correct.  But now I anticipate adding 
a feature that is substantial.  I am thinking that there would be value in 
preserving a working version of the program while I add the new feature.  To 
that end, I thought that I would create a clone of the current repository 
and add the new code to the clone.  However, I am unclear on what I will do 
when I finish adding the new feature.  I know that I can pull the changes in 
the clone repository into the original repository.  Once I have done so, the 
two repositories are equivalent, so I don't need the clone anymore.  Do I 
delete it?  If not, it seems as if I will proliferate repositories, all of 
which are useless except one.  If I delete the clone, does Mercurial still 
know that it once existed?  Could I recreate it, if necessary?
-- 
Jeffrey Barish