Given the time at which it was written, this book is still unsurpassed in its quality, conciseness, and clarity of style in the field of NMR. It was written by a physicist, contrary to Ernst's book, which is also excellent but focuses more on the interpretation of spectra rather than spin physics. (This was of course Ernst's intent anyways, there already was an excellent textbook on NMR by Abragam; he saw no need to reinvent the wheel.) While Abragam's book is quite a joy to read, it is also rather outdated. Nearly half the topics dealt with are no longer in use today. It needs to be supplemented by Ernst's book, as well as Callaghan's book. I also recommend Abragam & Goldman's book "Order and disorder" for a treatment of more modern NMR physics, if like me, you like eclectic topics. Abragam uses a purely quantum approach, and does not treat feynman path integral approach to relaxation. The chapter on relaxation is excellent, and has been praised by many, but he does not really explain much of the theory of random processes, so you'll need to learn this stuff elsewhere if you've never been exposed to stochastic processes before (well, Abragam makes very simple markov approximations, which is easy to understand as is- without too much extra background). Definitely the chapter on dipolar line width is worth mentioning.
Read more at http://ebookee.org/The-Principles-of-Nuclear-Magnetism_860132.html#Gf1JWQ8y0IFJSRcJ.99