Best Science Books

I posted about this earlier today on Critical Mass, but had to post it here too, because it's a subject near to my heart: Discover Magazine has just published their list of the 25 Greatest Science Books of All Time. They picked Darwin's Voyage of the Beagle and Origin of Species as number one. If you disagree with them, go vote on a different one ... you can choose one from their list, or add one you think they missed. (I can't believe Lewis Thomas didn't make the top 25!)

Clearly, their list is more about scientific umph than readability (with a couple exceptions, like The Man Who Mistook His Wife for a Hat). If we're talking about books that made great scientific contributions, I'd say their top 25 is right on, but if we're talking literary contributions, that's another story. Because unfortunately, so many important science books are unreadable. For their next list, I'd love to see Discover do the 25 best science reads of all time, because I think the best science writing conveys important and complicated information to the general public through storytelling. But unfortunately, those books can be hard to find.

I'd start with this Lewis Thomas book and Randy Shilts, an under-read writer responsible for one of the best and most important science books ever. Then off the top of my head, I'd turn to NBCC winner Jonathan Weiner, Oliver Sacks, Deborah Blum, Tracy Kidder, NBCC winner Anne Fadiman, Richard Rhodes, Paul Hoffman, Michael Pollan. I loved Gay and Laney Salisbury's The Cruelest Miles: The Heroic Story of Dogs and Men in a Race Against an Epidemic, but I'm a sucker for a dog story. Burkhard Bilger is doing some of the best new science writing at the New Yorker, while Ted Conover's eerie essay Trucking Through the AIDS Belt has stuck with me for more than a decade. I think all science writers should read John McPhee's Travels in Georgia, then spend several days thinking about its structure, and several more thinking about its character development.

I'm curious to hear what others might add ...