I can second this recommendation - it took me a while to figure out the proofs, but it was wonderful to be exposed to a whole new style of reasoning that I had

I can second this recommendation - it took me a while to figure out the proofs, but it was wonderful to be exposed to a whole new style of reasoning that I had only seen filtered in a very indirect way in uni physics classes. His whole approach is regarded ultimately as a too-hard-to-imitate dead-end [ people regard it as having held back physics/maths research in Oxford/Cambridge for as many decades as his style was taught, as opposed to Leibniz's calculus ], but it's still very flashy and cool.