One formative book for me was Ed Purcell's Electricity and Magnetism. Purcell was my early education in thinking like a physicist. It introduced me to thought experiments, simple models, and the usefulness of new mathematical tools. It's mathematically very clear, and physically insightful. The problems are extremely rich. It manages a huge deal of physics without much unnecessary computation. When I post replies to physics questions on this board, I sometimes wonder how Purcell would handle it - setting out the physical principles first, carrying out the calculations as clearly and succinctly as possible, and using physical insight for shortcuts and simplifications. I doubt I will ever live up to his model, though.
My entire freshman class read Purcell during second quarter. In conversation later, I heard over and over from people that after reading chapter 5, "The Fields Of Moving Charges", they were so awestruck by this 25-page illustrated introduction to relativistic E&M that when they finished, they stared at the wall for about ten minutes, then read the entire thing over again.
Also from the Berkeley physics series, Frank Crawford's Waves is charming, delving into many interesting wave phenomena with very nice explanations.
James Nearing's free book, Mathematical Methods for Physics on undergraduate math methods deserves more attention that it gets. It's clear, insightful, and packed with good problems.
This post imported from StackExchange Physics at 2014-05-04 15:36 (UCT), posted by SE-user Mark Eichenlaub