50 SECTIONAL ADDRESSES 
in type with those of the neutral atoms of the next hghter element. 
The exceptional cases would include those elements whose atomic 
structure involved the commencement of the development of an inner 
system of orbits, such as those of the 3;, 4,, 4,, &c., groups. Subject 
to these limitations, we should expect to find that if the n last-bound 
electrons were removed from a neutral atom of an element the orbits 
that remained in this atom would be identical in type with those of 
the neutral atoms of the nth lighter element. This would mean that 
the arc spectrum of the monovalent positive ion of are element would 
be identical as to types of series involved with the arc spectrum of the 
neutral atoms of the next lighter element. ‘There would be this differ- 
ence, however, that in the series formule of the spectrum of the ion 
the Rydberg constant would be 4K, whereas in the series of the 
spectrum of the neutral atoms of the lighter element it would be K. 
Putting the matter as it is ordinarily stated, the spark spectrum of 
an element should be made up of series of the same type as those of the 
arc spectrum of the next lighter element. This is known as the Kossel- 
Sommerfeld Displacement Law. 
Numerous illustrations of this law might be cited. For example, 
the series in the spectrum of the monovalent positive helium ion are 
of the same type as those of the spectrum of atomic hydrogen. Again, 
the series in the spark spectra of the alkali elements have been shown 
to be similar in type to those of the are spectra of the rare gases. In 
the case of potassium,*’ it has been shown that in addition to its arc’ 
spectrum it can, under moderate excitation, emit a series spectrum 
identical in type with that of the red spectrum of argon, and under 
violent excitation a spectrum having all the characteristics of the blue 
spectrum of this inert gas. In the case of the alkaline earths, the 
series spark spectra have the same characteristics as the are spectra 
of the alkali elements. 
But perhaps the most striking confirmation of the correctness of 
Bohr’s view of the process of binding electrons to nuclei, and also at 
the same time of the validity of the Kossel-Sommerfeld Law, is found 
in recent work by Paschen ** and by Fowler” on the spectra of doubly 
ionised aluminium and trebly ionised silicon. 
It will be recalled that Fowler some years ago showed that the waye- 
lengths of the spark spectrum of magnesium could be organised into 
series having 4K for their Rydberg constant. Early this year Paschen 
carried the matter farther by showing that under strong excitation 
aluminium emits a spectrum that can be arranged into series with a 
Rydberg constant equal to 9K. Now Fowler has capped it all by 
showing in a brilliant piece of work that in the spark spectrum of 
silicon certain wave-lengths can be grouped into series with a Rydberg 
constant of 16K. With both elements the series referred to are doublet 
series of the type obtained in the are spectrum of sodium. 
27 McLennan, Proc. Hoy. Soc., vol. 100, p. 182, 1921, and Zeeman and Dik, 
Konin. Akad. Van Weten, Amsterdam, Proc., vol. xxv., p. 1, April 29, 1922, and 
Ann. der Phys., Bd. 71, Heft 1/4, p. 188, 1923. 
*8 Paschen, Ann. der Phys., Bd. 71, Heft 1/4, p. 142, Heft 8. p, 537, 1923. 
29 Fowler, Proc. Roy. Soc., vol. 103, No, A, 722. June 1923. 
| 
| 
