抄録
Einstein introduced the spontaneous process of "outgoing" radiation in 1916, corresponding to the radiation emitted from an oscillating Planck resonator. In 191& Bohr suggested that the probability coefficient of spontaneous emission could be determined by the amplitude of the multiply periodic system. The correspondence principle, however, enabled one to indicate only an asymptotic relation(13)(in the text) in the limit of high quantum numbers, and was powerless to fix the unique form of the amplitude valid for all quantum numbers. Kramers gave up determining the amplitude in terms of the multiply periodic system, and regarded Eq.(26)as quantum-theoretically valid relation. Then he reversed Eq.(26)and took the relation(27)as the definition of the characteristic amplitudes of the virtual oscillators. Eventually, Kramers' dispersion formula was the first fruitful attempt to embody Einstein's probability coefficient of the spontaneous emission in the form of the "strength"(f=A・γ) of the virtual oscillator.