A.— JLVIHEJIATICAL AND PHYSICAL SCIENCES. 19 



in the production of lines are restricted in accordance with selection rules 

 which have since been extended to more complex spectra, as will appear 

 later. 



From the theoretical point of view the terms have a more direct physical 

 significance than the actual lines, and it is therefore important to determine 

 them with the greatest possible accuracy. Actual values of the terms can 

 be determined only from spectra in which series have been identified. The 

 calculated limit for one of the series then gives one of the terms of a system, 

 and all the rest follow from the interrelations by subtracting the observed 

 wave-numbers of lines from the limits of the series to which they belong. 

 Since the same terms may enter into several combinations, it is obvious 

 that the representation of a spectrum by terms is a great simplification. 



It should be understood that these studies of the structure of spectra 

 were pursued with the clear conviction that they would ultimately reveal 

 the secrets of atomic structure, and the analysis of spectra, as distinct 

 from spectrum analysis, gradually became one of the principal objects 

 of spectroscopic research. 



Notwithstanding the absence of a guiding theory of the origin of 

 spectra, experimental spectroscopy continued to attract a number of 

 workers, who approached the subject from various points of view and 

 accumulated a vast store of observational data. 



With the advent of Bohr's theory of spectra in 1913, spectroscopy 

 entered on a new phase of activity. The theory and its immediate 

 explanation of the spectra of hydrogen and ionised helium are now so 

 well known as to call for little more than mention. Adopting the Ruther- 

 ford conception of a neutral atom (namely, a positively charged nucleus 

 with sufficient electrons in orbital motion around it to neutralise the 

 positive charge of the nucleus) and restricting the possible orbits by 

 quantum considerations, Bohr was able to extend the theory to account 

 in a general way for the series spectra of other elements. Spectroscopic 

 terms were translated by the theory into ' energy levels ' of the atom, so 

 that a spectrum line is considered to represent the energy emitted by 

 an excited atom when it passes from a non-radiating state of a certain 

 energy to another of lesser energy. The terms are, in fact, proportional 

 to the energies of the corresponding 'stationary' states. 



The spectrum of a neutral atom is supposed to be generated, one 

 line at a time, by the transitions from one possible orbit to another of 

 the most loosely bound of the outer electrons (the ' series electron ' or 

 ' light electron '), while the whole spectrum represents the integration 

 of the various transitions taking place in different atoms. In partial 

 justification of this view, one might point to the spectra of the alkali 

 metals, which are very closely similar, and thus indicate that the general 

 type of spectrum is independent of the total number of electrons present. 

 The influence of the underlpng electrons, or of the nuclear charge, however, 

 becomes apparent in the increase of doublet separations with atomic 

 number, and in the displacement of corresponding lines to different parts 

 of the spectrum. 



The theory in its first form also gave a definite significance to the 

 enhanced lines occurring in the spectra of other elements besides helium. 



2 



