I contrast the profound differences between the ways in which algebra and analysis are used in physics. In particular I discuss the fascinating phenomenon that theoretical physicists devote almost all their efforts to algebraic problems even though all problems of experimental intrest require some methods of analysis.