240 REPORTS ON THE STATE OP SCIENCE. — 1920. 



solution. The reaction with oxygen, therefore, is characteristic of one or both 

 of the two molecular phases present in alcoholic solution. If henzaldehyde is 

 dissolved in concentrated sulphuric acid it exhibits two new molecular fre- 

 quencies, one in the visible and the other in the ultra-violet region. Two 

 further molecular phases, therefore, exist in solution in sulphuric acid. In this 

 case the benzaldchyde is no longer oxidised by oxygen, but is readily converted 

 to the sulphonic acid. 



Now the question arises as to the amount of energy necessaiy to convert 

 one molecular phase into another and the mechanism whereby this energy is 

 supplied. The amount of energy required per molecule is readily calculated, 

 and is equal to one or more quanta measured at the infra-red fundamental of 

 that moljcule. If the frequency characteristic of the first phase is x times 

 the infra-red fundamental and the required phase is characterised by a frequencv 

 which is y times the infra-red fundamental, then the energy required for each 

 molecule is x—y quanta at the infra-red fundamental. Obviously the molecular 

 system can absorb this energy when exposed to radiation of a frequency equal 

 to its infra-red fundamental, or, as explained above, it may absorb it at any of 

 the frequencies characteristic of its component atoms. Lastly, the molecule 

 mav absorb one quantum at its characteristic phase frequency, and under 

 ordinary circimistances this energy will again be entirelv radiated as quanta 

 at a lower phase frequency, the infra-red fundamental, or the atomic frequencies. 

 If there is present a substance capable of reacting with a less condensed phase, 

 then the molecule is converted into that phase and reacts, the balance of 

 energy being evolved as infra-red radiation. The essential point is that the 

 necessary amount of energy to change the molecular phase is x — y quanta at 

 the infra-red fundamental, and that when one quantum is absorbed at the phase 

 frequency the excess energy over and above that required is radiated. The 

 change of molecules from one phase to another under the influence of light is 

 readily enough shown experimentally, but it is necessary to stabilise the second 

 phase in some way, since otherwise it returns instantaneously to the first phase. 

 An interesting example is furnished by trinitrobenzene. an alcoholic solution 

 of which contains a molecular phase characterised by a frequency in the ultra- 

 violet. A piperidine solution contains a molecular phase of trinitrobenzene 

 which is characterised by a frequency in the blue and the solution is deep red 

 in colour. This second phase, therefore, is favoured by piperidine. If to an 

 alcoholic solution of trinitrobenzene a small nuantity of piperidine is added, not 

 more than one molecule of piperidine to 10 molecules of trinitrobenzene, the 

 solution remains perfectly colourless. On exposure to light of the frequency 

 characteristic of the phase in alcohol the solution turns red, owing to the 

 formation of the second molecular phase, and the solution slowly becomes 

 colourless again when placed in the dark. 



There is no need to enter into a discussion of the anplication in detail of 

 this theory to the quantitative relations involved in the energy changes of 

 chemical reaction. It is obvious that the theory renders possible the calcula- 

 tion of the complete energy changes, and this aspect of the phenomena may be 

 left on one side. From the point of view of absorption snectra the essential 

 fact is that the theory leads to the conclusion that a molecule must exist in one 

 of a number of possible phases, each of which is characterised by its own 

 absorption band in the visible or iiltra-violet region of the spectrum. It has 

 been proved that a molecule can be brought from one phase to another by the 

 gain of a whole number of fundamental infiia-red quanta and that this can be 

 brought about by exposure to radiant energy at a frequency characteristic of the 

 molecule. Reference has already been made to the fact that it is nossible to 

 chancre the nhase in which a molecule exists by the use of a suitable solvent, 

 and indeed it is to this effect of a solvent that the variation in the absorption 

 spectra of many compounds is due. 



In order to understand this effect of a solvent it is necessary to consider the 

 condensation of the molecular force fields a little more in detail. From what 

 has already been said it is clear that this condensation will proceed to the 

 farthest possible extent. In the case of a molecule in which the external force 

 fields of the atoms are well balanced the condensation will proceed far with 

 the establishment of a highly condensed field characterised by an absorption 

 band in th? ^xtreine iiltra^violpt. On, the pt-hpr h^nd, if the external force fields 



