On The Nature Of Light

Voi. 12, 1926 PHYSICS: TOLMA N A ND SMITH 343 in absorption in Mg vapor.11 The fact that in the 2p states of HgH one electron interacts with the H electron, while in the Mg(QP)H(2S) state of MgH the two Mg electrons interact with each other and not with the H electron, may be correlated with the fact that Hlg shows chemical valences of both one and two, while Mg shows only the valence two; similar differences also exist in the spectroscopic behavior of the two atoms. A more complete discussion will be given in later papers. R. S. Mulliken, these PRO-CEEDINGS, March, 1926. W. Pauli, Jr., Zeit. Physik, 31, 765 (1925). W. Heisenberg, Ibid., 32, 841 (1925). F. Hund, Ibid., 33, 345 (1925). 3R. Mecke, Naturwissenschaften, Sept. 4, 1925. 4 E. E. Witmer, these PROCUEDINGS, April, 1926. 5 J. C. Slater, these PROCEEDINGs, Dec., 1925& 6 W. Lenz, Verhandl. D. Phys. Ges., 21, 632 (1919). 7L. Nordheim, Zeit. Physik, 19, 69 (1923); cf. also L. Mensing, Zeit. Physik, 34, 602 I 2 (1925). 8 A. Sommerfeld, Atombau und Spektrallinien, 4th edition, p. 203-6. 9 R. T. Birge, Nature, Feb. 13 and Feb. 27, 1926. 10 Cf. H. Sponer and J. J. Hopfield, Physic. Rev., May (?), 1926 (Abstract). S. Barratt, Proc. Roy. Soc., 109A, 194 (1925). ON THE NATURE OF LIGHT By RICHARD C. TOLMAN AND SINCLAIR SMITH NORMAN BRIDGE LABORATORY OF PHYSICS AND MOUNT WILSON OBSERVATORY Communicated March 18, 1926 In a recent article in these PROCBEDINGS1 Professor G. N. Lewis has suggested that the conflict between the quantum theory of light and the known facts of interference might be resolved by assuming- that an atom never emits a quantum of light except to another atom, the possibility of transmission of the quantum being determined by the nature of the paths connecting the two atoms in such a* manner that the transmission will not occur if forbidden by the known laws of interference. The idea is made more plausible by describing such transmissions with the help of Minkowski's four-dimensional geometry of relativity. In the language of this geometry the path of the quantum passing between the two atoms would lie along a singular line, and since intervals along such singular lines have zero length in the geometry used, the atoms may be regarded as in virtual contact, thus making it less surprising that the emitting atom should "know" the existence of a receiving atom in such a condition and location as to permit a transfer. The apparent contradiction to our usual ideas as to the sequence of cause and effect also becomes less serious when viewed 344 PHYSICS: TOLMA N A ND SMITH PROC. N. A. S. from the point of view of this four-dimensional geometry. To common sense it seems absurd that the emission of light from a distant star should be determined by the condition, thousands of years later, of a chloryphyll molecule on the earth which had not even been synthesized at the time of emission. Nevertheless by a proper choice of space-time axes in the fourdimensional geometry, both the distance between the source and recipient and the time of transmission may be made as small as desired. Views, in some ways similar to those of Lewis, have previously been presented by other authors.2 Thus as early as 1921-22 Schottky discussed in considerable detail the difficulties encountered by the quantum theory as to the sequence of cause and effect, and indeed speaks as follows: "....die gegenseitige Bedingtheit von Emission und Absorption nicht verstiindlich ware, wenn man die Emission in normaler Weise als die zeitlich yorangehende Ursache der Absorption ansahe; vielmehr miisse die Wechselbeziehung zwischen Emission und Absorption so aufgefasst werden, dass beide Vorgange sich in vollig symmetrischer Weise gegenseitig bedingen." In addition Smekal in 1922 formulated the general thesis that absorbing and emitting s