Though mostly about non-technical economics, the Stern Review also reviews much of the research on the impacts of climate change without going into any of the science. This is a long and important review, which only descends into economics jargon in a few places. The non-economist may be surprised to find that the case for action is essentially ethical; the figures for the impact on the world economy rely on treating people nearly equally now and in the future, independent of where they live. If you live in a wealthy country and don't care about other people, or people in the future, you will see no reason to take action on climate change. The review discusses how you can treat people more equitably to estimate GDP numbers for the world as a whole, accounting for discounting and growth (e.g. that we will be richer in the future, so the apparent cost - loss of utility - of taking action can seem less if we take it later). It then goes on to discuss in some detail possible mitigation strategies, accounting for potential economic pitfalls along the way. He makes a strong the case for aiming for 80%+ cuts in CO2 emissions as soon as possible, in an attempt to prevent temperature rises of more than 2C and reduce the risk of catastrophic feedbacks. But it is clear that many large uncertainties remain, and even this may not be enough to prevent very damaging consequences (at least if you are poor, or care about your environment). The case for carbon capture and storage is made clear: it allows demand, and hence the price, of carbon-rich fuels to remain high while not causing direct damage; simply switching away from carbon-rich fuels could collapse the price of the fuels and lead to a boom in consumption elsewhere. If you are a rich nation, perhaps the best thing you can do is buy as much fossil fuel as possible and burn it with carbon capture as soon as possible?
An appendix updates the review based on feedback on publication (for example criticism of his method of discounting the future). It doesn't directly address potentially interesting arguments for not taking action now; e.g. that by investing vast amounts of money on lifting people out of poverty in Africa now, then taking action about climate change later, the world as a whole might be better off. (Naturally this argument is academic; it is empirically proven to be very difficult to get rich nations to do anything significant for other countries without a strong element of direct self interest).
A very readable and enlightening discussion of what needs to be done to avoid catastrophic climate change. It seems to be well researched and balanced (even if it starts off a little polemically). Especially interesting for the economic aspects: paradoxically, increasing energy efficiency can actually lead to an increase in total emissions. Everyone should read it if they don't have the time to read something more comprehensive (shame it's not publicly available on his web site; also look out for upcoming book by David MacKay).
These books are highly recommended, well written and easy to read. Everyone should read them. They explain some of the subtleties of evolution very clearly and convincingly. Some of my first QBasic computer programs were geographical simulations of the evolution of strategies for playing iterated prisoner's dilemma - inspired by these books.
Further reading: The Extended Phenotype develops some of Dawkins ideas in a rather less readable way and is not really recommended. Darwin's Dangerous Idea by Daniel Dennett is a good (if long) look at the power of evolutionary thinking, with an interesting discussion of what exactly a species is. The evolutionary basis for human behaviour is discussed in the excellent The Red Queen and The Origins of Virtue by Matt Ridley. The Language Instinct by Steven Pinker explains Chomsky's ideas about universal grammars and the possible evolutionary basis.
Until I read this book in my early twenties I'd never really even thought about how to think about consciousness. I found it almost unputdownable - a very readable and engrossing story of what consciousness definitely isn't, and a good and thought-provoking stab at explaining what it is. Really gets you to think about how to think about thinking. You'll never look at the world quite the same way. Note: not everyone finds this book so readable or good, though I find it hard to identify with most of the negative reviews on amazon.
Further reading: The Conscious Mind by David Chalmers is much harder going than the Dennett book. He agrees with Dennett about a great many things, and only seems to differ in that he thinks consciousness is only physically supervenient on brain processes, whereas Dennet argues that is it logically equivalent. I find the simpler Dennett thesis much more convincing. Note: terms like supervenient are absent from Dennett, but all over the place in Chalmers, which is partly why it is much less readable. The Mind's I is an interesting collection of essays with some interesting thought experiments. Elbow Room and the more recent Freedom Evolves (also by Dennett) have considerable overlap and discuss free will in a deterministic universe. (Dennett includes a rather short discussion of hyperbolic discounting, and refers to it as being a glitch; for better ideas see e.g. ,  and references.) Are animals conscious? For a scientific analysis that quashes some naive expectations see Animal Suffering (though the main points are also in Consciousness Explained).
There is no doubt a great deal of philosophical literature about these issues. The great thing about Dennett is that he actually does philosophy, whereas I find that many 'philosophy' books appear to be almost entirely history of philosophy.
This book has an overall thesis about the nature of consciousness that I do not believe at all. However his exposition covers a large and fascinating area of physics and logic, and can be read entirely for that. I first read it when I was about 16. I found much of it inspiring and fascinating, but also some of it rather impenetrable. I re-read it a few years later after reading about some of the content elsewhere; some of it is much easier to read if you already know what he is talking about. In particular Gödel's theorem (Antony Lewis cannot prove this statement is true is a true statement that I can never prove to be true) is explained more clearly elsewhere. Other mathematical material is good though (Cantor's diagonal slash etc). On physics there is a very clear and illuminating discussion of entropy: this was where I first understood how the Sun drives life on earth even though energy is conserved. It clearly explains one of the key challenges for cosmology: how did the universe manage to start in what appears to be a very special and unlikely state? Most of the content of Penrose's talks is in this book.
Further reading: Shadows of the Mind is Penrose's follow up book, which I don't find any more convincing but does have some interesting material, for example a discussion of bomb testing.
Gödel, Escher, Bach by Douglas Hofstadter is the unashamedly pretentious classic that deals with Gödel's theorem (and much else). Some of it is quite hard going.
I came across this book quite recently, and found a very good lesson in how to think about time symmetries carefully. He presents a very interesting thesis that the nature of quantum mechanics is to a large extent determined by properties we should expect on general grounds from a time symmetric theory. He also clearly addresses the issue of entropy and initial conditions in cosmology; I now understand the idea of a Gold universe.
Further reading: Spontaneous Inflation and the Origin of the Arrow of Time by Carroll and Chen is an interesting (technical) paper that attempts to take some of these ideas seriously in a cosmological theory. It references may other interesting and related papers. Time's Arrow by Martin Amis is a rather fine short novel that can help you think backwards in time.
Good introduction to Bayesian methods and information theory, with some quite advanced topics. You can also download it for free.
Great books I've missed? Please let me know! Might add other sections...