Abstract
According to Planck's first theory (1900) the best expression for the average energy of an harmonic oscillator at high temperature is not kT but kT-1/2h nu . A second theory of Planck (1911) restores the expression kT, and also predicts zero-point energy. A discussion is given in terms of the correspondence principle, which, it is stressed, is active, rather than passive, in nature. Planck's second theory is then discussed in more detail, including the use made of it by Bohr (1918), both in the construction of his atomic theory, and in his development of the correspondence principle itself.