to be spent on the use of logarithm tables and manual square-root extraction can be prof­ itably turned to earlier studies of calculus or computer programming. Now that the calculator has been accepted into the classroom, we face a computer-software revolution which promises to be considerably more profound. Modern textbooks in the physical sciences routinely assume their readers have access not only to calculators, but often to home or even mainframe computers as well, and the problems teachers discuss and assign students can be more complex and often more realistic than in the days of only pad and pencil computations. As less effort is spent on numerical computation, more can be devoted to conceptual under­ standing and to applications of the increasingly sophisticated mathe­ matical methods needed for a real appreciation of recent advances in the discipline.