29^ REPORTS ON THE STATE OF SCIENCE, ETC. 



exact multiple of a smaller central frequency. This must not be interpreted to 

 mean that the absorption bands characteristic of a substance extend to an 

 unlimited distance in the infra-red. The frequencies in the very long wave 

 infra-red. Ai etc., are the smallest which characterise a molecule or its parts. 

 This is proved by the fact that these frequencies exhibit themselves as single 

 narrow absorption lines, and if there existed smaller frequencies still they would 

 be associated with subsidiary frequencies due to their combination with those 

 smaller frequencies in the manner already described. When a molecule absorbs 

 one quantum of energy at one of its smallest frequencies, Ai etc., it radiates 

 this again as one quantum at that frequency. 



This deduction from the Planck theory agrees, therefore, with the observed 

 facts as regards the integral relations between the central frequencies charac- 

 teristic of any one substance. The Planck theory, as it stands, however, does 

 not offer an explanation of the second relation between the infra-red frequencies, 

 namely, the principle of the least common integral multiple. There is no doubt 

 that this principle must be connected in some way with energy quanta, and the 

 question arises as to what this connection is. It has already been stated that 

 Coblentz discovered the fact that the long wave infra-red frequencies, previously 

 designated by Ri, are due to definite groups of atoms, and emphasis may be 

 laid on the fact that a specific group of atoms exhibits its characteristic fre- 

 quency whatever may be the composition of the rest of the molecule. This 

 very clearly establishes the individuality of these frequencies and their origin 

 in the relevant atomic groups. Then, again, more recent work shows that the 

 same very long wave infra-red frequencies are common to two or more com- 

 pounds which have the same atoms in common; and, further, these frequencies 

 are combined in their least common integral multiple to give the frequencies 

 of groups of those atoms. Since also the molecular frequency in the short wave 

 infra-red is based on the least common integral multiple of all the very long 

 wave infra-red frequencies present, the conclusion was drawn that these last 

 are characteristic of the atoms themselves. If, as observation would prove, an 

 atom is always characterised by the same frequency, of whatever molecule it 

 forms a part, it must always absorb the same quantum of energy, and the 

 natural deduction may be made that the atom is characterised by that quantum 

 of energy. 



This deduction that an atom is characterised by a definite quantity of energy 

 gives a basis on which to formulate a theory of absorption which has certain 

 material advantages over the Planck theory. The Planck theory starts from 

 the assumption that molecules are characterised by certain vibration frequencies, 

 and states that absorption and emission of energy at those frequencies are not 

 continuous but discontinuous, and consists of the absorption or emission of 

 energy quanta, each quantum being defined by the product of the frequency 

 into the constant 6.57x10"-'. No attempt is made to explain why a molecule 

 is characterised by specific frequencies, and for this reason the theory loses 

 somewhat and lacks completeness. The theory also takes no account of the 

 relationships between tlie various frequencies characteristic of a molecule, and 

 hence becomes a theory of monochromatic absorption or radiation, in that these 

 processes are assumed to be due to independent oscillators with a specific vibra- 

 tion frequency. Very important applications of this theory have, moreover, 

 been made to the energy changes involved in chemical reaction, and in the 

 great majority of cases the lesults experimentally obtained differ enormously 

 from those calculated on this theory. The alternative theory, however, not 

 only gives an explanation of these discrepancies, but also accounts for all 

 absorption spectra observations. 



This theory was fully given, up to the stage of development then reached, in 

 the last report of tlie Committee {B. A. Annual lioporifi, 1920, p. 222). but in 

 order that the recent evidence in its favour can be understood it is advisable 

 briefiv to re-state it. Four initial assumptions are made, which are quite 

 straightforward and in each case the simplest possible The first assumption 

 is that every elementary atom is characterised by a definite quantity of energy 

 associated with a definite physical process taking place within itself, such as 

 the shift of one electron from one orbit to another. This quantity of energy is 

 the same for all atoms of the same element, but is different for' the atoms of 



