Subtitled "Understanding Einstein's Relativity," David Mermin's It's About Time is another book (like An Illustrated Guide to Relativity) that grew out of a non-majors course on physics that Mermin offers at Cornell. It's also an almost-forty-years-later update of an earlier book he wrote on the same subject. And it's been a really good resource for writing the book-in-progress, which I ought to repay by reviewing it here.
Like the Illustrated Guide, this is a book that aims to teach students something about how relativistic kinematics actually works. Unlike the Illustrated Guide though, this does not stint on the math. Not at all. While the book is liberally illustrated, it also contains lots of equations, and expects the reader to work through a good bit of algebra. As such, it's not for the faint of heart-- it wouldn't be a good option for, say, an English major with a passing interest in the subject.
If you can make it through the math, though, Mermin offers a good deal of insight into the workings of the theory. He builds up the whole strange edifice of relativistic kinematics very rigorously, from first principles, and with an emphasis on the importance of time to the whole enterprise. Even if you're already an old hand with the Lorentz transformation formulas, this book is worth a look, especially Chapter 9, which provides an extended demonstration of how everything weird about the theory can be reproduced at low speeds using deliberately unsynchronized clocks. It's a great demonstration of the importance of the relativity of simultaneity for understanding how the whole business works. If it were possible to do it justice in less space and without equations, I totally would've copied it for the book-in-progress.
Of course, like the Illustrated Guide it does a great job with the kinematic aspects of the theory, up to and including the notion of the spacetime interval and spacetime diagrams (Mermin's diagrams aren't as charming as Takeuchi's, but his descriptions were a huge help to me), but sort of peters out when it gets to dynamics. The chapter on momentum and E=mc2 involves a great deal of hand-waving and approximation that I half suspect comes in because the detailed treatment of kinematics has eaten up basically the whole semester by that point, and he needed to wrap his course up.
It's kind of a good thing from my point of view, though, because if it had a treatment of momentum and energy that was as good as the earlier stuff, it would've provoked a massive crisis of faith for me. As it is, I feel like I have something to bring to the project of writing about relativity for a non-scientific audience other than, you know, talking dog jokes. I can take some of this, and cut out most of the equations, lift some diagrams from Takeuchi's book, take the very good description of energy and momentum from Cox and Forshaw and get rid of the cringe-inducing running joke about Thales of Miletus (which, seriously, sounds creepier every time I re-read it, something I've done a lot in order to get the problematic Chapter 7 into shape. I have no idea what they were thinking with that.), and so on.
I don't know that I would attempt to teach a course out of this-- if Cornell's non-majors are down with this, they're several cuts above our students. That, or the bigger population just means the small fraction of non-scientists with the right mix of skills and interests for this is big enough to justify a course like this. It's also a little too long and involved to really work for a major-track course, though I'll be sure to recommend it for students to read on their own.
I guess the real audience for this book is other physicists who find themselves needing to know more about relativity in order to teach it (or write books about it). And for that audience, I recommend it very highly-- if you know the equations, but want a better feel for what's behind them, pick this up and work through it.
In my opinion, he makes the spacetime diagram chapter too difficult by making it purely abstract. That is, he's so caught up in the frame independence of it that he doesn't draw perpendicular "lab frame" axes the way most people do when dealing with spacetime diagrams. As a result, the whole discussion is less approachable than it would have been with that concession to practicality.