D O C . 3 1 8 C R I S I S O F T H E O R E T I C A L P H Y S I C S 2 5 3 total chemical action of radiation that is permeated by matter is only dependent on its total energy but not at all on its spatial energy density. Experiments by E. War- burg have furthermore demonstrated that the energy absorbed per chemical ele- mentary process is always equal to hv, independent of the spatial energy of the radiation.[14] This result also follows from experiments on the photoelectric effect and on the generation of cathode rays from X-rays. We know today that this energy really does originate from the radiation and that it is not, for example, gradually accumulated. The absorption of light constitutes indivisible elementary processes, during each one of which the energy hv is com- pletely transformed. We know nothing about the details of such an elementary pro- cess. If just the energetic properties of radiation were known, we would then see ourselves obliged to postulate a kind of molecular theory of radiation of the kind of Newton’s emission theory of light.[15] But finding explanations for diffraction and interference processes poses insurmountable obstacles. Moreover, it should probably be kept in mind that the field theory of radiation is no more false than the theory of elastic waves in solids, which establishes their thermal content for, both theories collide to the same degree with the quantum relation and must be com- bined with the latter in the same way in order to arrive at an appropriate interpreta- tion of the results of experience. The grand development of our knowledge of the composition of atoms, which we essentially owe to the great masters Rutherford and Bohr, has led to a highly significant generalization of the quantum rule, which we now want to consider. Even prior to the Rutherford-Bohr theory,[16] the assumption was that absorption or emission of a spectral line would have to correspond to a transition of an atom or molecule from one preferred state into another. As the states of the elementary forms were certainly not interpretable as sine-type oscillations, the problem arose of how to extend quantum theory to mechanical systems of more general character, which Bohr’s, Sommerfeld’s, Epstein’s, and Schwarzschild’s investigations have succeeded in doing, step by step.[17] The results obtained by these researchers are being elevated to secure findings through precise confirmation in the field of spec- troscopy. If a mechanical system is describable by coordinates , which experience cyclic changes over time, and if for each degree of freedom v the momentum belonging to is describable as a function of exclusively the one coordinate , then for any v the integral over one cycle is an integral mul- tiple of Planck’s constant h. Thus the “permissible” states according to quantum [p. 7] qv pv = ∂L(q p) ⋅ ∂qv --------------------- - qv qv pvdqv