Best Abstracts Ever

Michael Berry (of levitating frog fame) has recently been lauded as writing the best scientific abstract ever:

It's short, attention grabbing and conveys results, all of which are good things in an abstract. And I think it's better than many other claims for "best abstract". Avron and coworkers published in Communications in Mathematical Physics and decided "Abstract" was a good description of their paper (although some parts are concrete).

And Spiegelman's contribution to EOS-Transactions seems to indicate that anything can get published. I'm not sure if that was meant as a commentary on the deluge of fractal publications or just silliness.

My favorite abstract isn't ever mentioned in the corresponding paper. I'm not sure if the dry wit of the article and Dr. Strangelove reference in the title resonate with today's students. But examples the point out that a correlation doesn't mean a causation don't get better than:

Let me know if you know of any other candidates for "best abstract ever".