Joseph Silk's "Horizons in Cosmology" isn't just a great non-mathematical introduction to cosmology and galactic astronomy: it's a fascinating portrait of how science develops by what our theories *don't* explain. It represents a brilliant snapshot of Astronomy at the beginning of the 21st century. In this respect, Horizons is quite different from Silk's textbook "The Big Bang".
This book contains the clearest explanation of galaxies I've ever seen: Silk's work in galaxy formation gives him an insider's view of what we know (quite a bit), what we don't (a whole lot more), and what we really *need* to know. At this point, the "big picture" of Cosmology is pretty well established: the Universe began in a big bang about 14 billion years ago, then slowly developed into a web of galaxies influenced by dark matter and energy. However, the details of how galaxies formed and how they develop through time is much less clear. Supercomputers are not yet much help: the physical processes at work in the growth of galaxies so complex we can't yet build realistic simulations. More critically, scientists haven't yet figured out how the many processes of feedback and environmental interactions fit together (you can't simulate what you don't understand.)
Again and again, Silk's explanations show the bleeding edge of theory: each successful model somehow raises as many questions as it answers. Though the book is non-mathematical, Silk's exposition is rigorous and detailed - this is not in any sense an easy-reading book. However, persistence brings real rewards.
For the last several years, I've been involved in the Galaxy Zoo project [...], where volunteers classify images of galaxies from the Sloan Digital Sky Survey and the Hubble space telescope. This book is a MUST for zooites: I know of no clearer explanation of how galaxies work. Great job, Joe!