Virgil -- I don't see you contributing anything useful. Please use the wiki to help others along with the problems instead of complaining.
In addition, to Henry: I agree with your idea and your conclusion. We can expand a small amount from your idea by generalizing the forumla for $ X_i $:
$ X_i $ is geom$ (\frac{n - i + 1}{n}) $
E[# needed]=$ \sum_{i=1}^n E[Xi] = \sum_{i=1}^n \frac{n - i + 1}{n} $
I attempted to plug this into matlab's symbolic calculator, but the answer didn't come out to anything I understood.
Also, note for geom(p), variance is $ \frac{1-p}{p^2} $. I don't think we can simply add all the variances up; does anybody have a better idea?
