What books are recommended for an advanced undergraduate course in electrodynamics?
D.J. Griffith's Introduction to Electrodynamics must be mentioned. To my knowledge this text is ubiquitous in junior-level E&M courses.
- The writing is extremely friendly and is excellent for self-study.
- The author frequently tells you what he is doing and provides motivation, unlike the ubiquitous graduate-level text by Jackson.
- Equations often use a convenient notation (you know that script $r$?) that makes them appear less complicated, yet is straightforward to expand.
As for required background, I would say the only thing really required is a thorough understanding of multi-variable calculus. The physics content is self-contained, so I'd argue even freshmen level E&M knowledge isn't necessary, though it would only help a learner in thinking "like a physicist" to help solve problems.
Being a junior-level undergraduate text, it is not thorough nor does it go into much depth, at least compared to graduate level texts.
- You won't get a mathematically complete understanding of using Green's functions to solve boundary condition problems (e.g., Dirichlet conditions).
- Some results are simply stated rather than worked out due to their complexity, though the author is up front about this.
(This is a community response; feel free to add additional items)
Purcell is a good non-Griffiths option. I would judge the completeness of the material between Griffiths and Jackson, but with an intuitive level of understanding close to Griffiths. I used it to study for graduate qual exams when Jackson was making me feel particularly obtuse.
- Touches more ideas than Griffiths
- Uses some real-world examples
- Sometimes sacrifices full mathematical details to aid conceptual understanding
Let me give you one example. Electric quadrupole radiation. From my memory, Purcell starts with the already-derived expression for dipole radiation, then adds a second inverted dipole in the appropriate location to get the time-dependent electric field (by superposition) for a quadrupole source. How intuitive! (My memory is a bit hazy here, so somewhat can edit and correct if need be.)
As for required background, I would say it's pretty self-contained as most intro E&M books are, but multivariable calculus is a definite must.
Besides Purcell I really like Feynman Vol. II. I finally could understand magnetic materials and electromagnets. (Warning, Feynman uses his own notation for B,H and M.)
The lectures are available online and for free, as the New Millenium Edition, at
in a nice re-mastered edition with re-drawn vector-graphics figures and equations.
The lectures themselves do not contain any exercises, so they are not as good, on their own, for self study. This is partially solved by Feynman's tips on physics, which contains three problem-solving lectures by Feynman and a set of exercises and answers assembled by R.B. Leighton and R. Vogt (cf. the new preface to the lectures).*
*Hopefully someone familiar with that book can weigh in on how much that helps?
W. K. H. Panofsky and M. Phillips, Classical electricity and magnetism, Addison Wesley, 2nd ed., 1962
Especially the first 14 chapters are very enjoyable yet carefully written study text about both basic and more advanced topics in macroscopic EM theory (including discussion of EM energy from more experimental angle than is usual and of density of force inside matter - much omitted yet interesting topic), while miracuously still remaining concise (267 pages !). As a basic textbook, great life saver from the bulky compilations on EM theory.
Of course everybody told you about Griffiths and Jackson. And they are right about them, but let me add some other options:
As mentioned, a first book would be Purcell and David Morin's "Electricity and magnetism": this book worth reading page by page, word by word, don't miss anything, even if you already had a course on electro... this book is a classic made modern when Morin took the job of updating it. Read it.
After that you will like to advance to more formal books. The next step is Griffiths: is a canonical undergraduate book and fill the gaps between Purcell and Jackson quite well. But there are others: If you do not want to read Griffiths try Andrew Zangwill's "Modern electrodynamics". I am sorry Griffiths, you know I love you, but Zangwill is better.
Another great text is Charles Brau "Modern problems in classical electrodynamics". Ready it with Zangwill and you will never need anything else in that level.
Ohanian's "Classical electrodynamics" is a good book too, but I prefer Zangwill and Brau.
There is a book written by Julian Schwinger (God bless Julian Schwinger) called "Classical electrodynamics" and there is "Principles of electrodynamics" by Melvin Schwartz. I never really read them except for a few excerpts and I like what I saw (there is a topic on Schwartz's book where he discuss that once you have the electric field a magnetic field necessarily has to exist that is great and Schwinger shows that the least action principle for electromagnetic energy implies Maxwell's equations...). Schwinger's book is divides into little chapters that actually are lectures he gave so is a book of topics not a textbook.
There are Josif Frenkel's "Princípios de eletrodinâmica clássica" by a brazilian professor that is quite good too.
Professor Florian Scheck has a theoretical physics series where there is a book on classical electrodynamics that is very, very good, but Scheck writes with some very heavy and formal mathematics so is a future reference.
Another one is Greiner's "classical electrodynamics". All of Greiner's series on theoreticla physics are excellent.
When you are ready there is Jackson's "Classical electrodynamics" and William Smythe "Static and dynamics electricity". Jackson covers everything on the subject and with rigour and profound insights, but is a heavy book, no easy and is famous for treating some passages of some equations as trivial and when you try to calculate you see that you actually nedd 7 billion pages of calculations to reach the next step. I never read Smythe but my professor once told me that his problems are epic. And, of course, Landau. Those three are for graduate level. Mature audience only.
So, in suma: I would start with Purcel and Morin, then I will get Greiner and/or Zangwill and/or Charles Brau, then I would go to Scheck and Landau, and finally Jackson and/or Smythe.
PS: If you like to see how physics evolved and how it was done some 150 years ago, there is Maxwell's book; of course is not a textbook nowadays but you can always learn something when you read good books, and this is a classic writen by some of the greatest so it worthy some time.
The links provided in
Gerard 't Hooft's webpage (how to become a good theoretical physicist) can be quite nice:
There is one thing I recommend though: learn it backwards compared to Griffiths; i.e. read the chapters on electromagnetic dynamics first, then take a look at the static cases. Otherwise you might end up as confused as me, feeling lied to. I recommend this generally: try to learn the non-simplified cases first.
- Fulvio Melia's Electrodynamics.
My graduate course on E&M used this text as a basis for the lectures (subsequently changed to the aforementioned Jackson). This book is very short (246 pages as compared to say Griffiths at 624 pages!), but covers all the relevant topics of E&M (Electrostatics, Magnetostatics, etc) before smoothly transitioning into more advanced material such as special relativity and the Lagrangian formulation of electrodynamics. While there are a few examples throughout the book, there are no in-chapter or end-of-chapter problems to assign to students.
NB: Melia uses Gaussian units in the text.
- Jerrold Franklin's Classical Electromagnetism.
This book is what I used as a supplement for my graduate work (though not as a required supplemental book). I basically view this text as a more-thorough extension of Melia's work; it covers much of the same topics in a similar order, but does it at a slower pace with more examples and homework problems.
NB: Franklin also uses Gaussian units in the text.
Since there are several mentions of Jackson, let me (somewhat self-promotingly:) add a link to my online notes based on the 1st-year graduate E&M course I took at CCNY, based on that book: http://www.forkosh.com/u715.html and http://www.forkosh.com/u716.html It was a two-semester course, and these are my first- and second-semester notes, respectively. Sorry that a few of the lectures towards the end of each semester are missing (especially in u716) -- I was cramming for finals and never got around to transcribing my hastily-written class notes into the notebook.
Boyer was a great teacher, and very well-liked by all his students. And my notes based on his course contain many supplementary examples and discussions not in the text. Indeed, his lectures were mostly supplementary material (you're pretty much expected to already know the material in the text), and I hope my notes convey some of Boyer's exceptional ability and enthusiasm for the subject.
Any comments, suggestions, questions re the notes? Though I warn you, it's been quite some time since I was intimately familiar with the material. (P.S. But, yes, I got two A's:)
One more option to consider is Wangsness. It is a bit more solid on developing vector calculus muscles and covers most of the classic areas. Not as hard or intricate or terse as Jackson, but good trainer before moving on to Jackson.
There are mixed opinions on the very long vector calc section at the beginning but I really think this is a good idea. If you don't get that down you are going to have a hard time later on.
The relativity section within there is not a be all treatment either. But I think students should learn the classical theory first. Having relativistic treatment as a bit of an end of the course digression is probably best for undergrad. You need to be solid on things like method of images first.
P.s. FWIW, I don't think Jackson is appropriate for an undergrad class (nor Stratton or Panovsky/Phillips or similar texts). These answers almost read like the person did not read the entire question. [I guess you can make an argument for using a grad text in UG. I would still disagree with it as it is not good pedagogy. But if you did have that point of view it would be a good idea to explain why you feel this contrary opinion to be true. Just to omit it though smacks of not reading the question. Like when someone says he wants a relatively easy real analysis text and baby Rudin is tersely suggested.