107 



one electron; it will therefore be quite analogous to the hydrogen spectriiin and will 

 be given by the formula (35), if A' is replaced by 9 A'. 



If next we proceed to beryllium which is the fourth element in the itcriodic 

 table and the neutral atom of which will therefore possess four electrons, we may 

 expect that in the normal state of the atom the two electrons move in orbits out- 

 side the two others. The beryllium spectrum has not been investigated in great 

 detail, but from the observations on the spectrum of magnesium, which is the next 

 element in the Be group of the periodic table, we shall expect that beryllium will 

 possess a series spectrum of the first order of a new type, different from that of 

 the first order spectra of helium and lithium. This spectrum may be assumed to 

 originate from transitions between states in which one electron moves at distances 

 from the nucleus large compared with the three inner electrons, of which the one 

 again moves in an orbit outside the two others. The essential difference between the 

 spectrum under consideration and the helium spectra of the first order may be ex- 

 plained by the fact, that the outermost of the inner electrons in the beryllium atom 

 moves in a field which differs considerably from that of a single nucleus of double 

 charge, and its orbit will therefoi-e not undergo perturbations under the inlhience 

 of the outer electron of the same order of magnitude as that of the inner electron 

 in the helium atom. In a discharge of greater intensity beryllium may next be ex- 

 pected to show a spectrum of the second order of the same type as the lithium 

 spectrum of the first order. This is supported in an interesting way by the close 

 analogy of the structure of the magnesium spark spectrum mentioned above with 

 that of the ordinary arc spectra of the alkali metals. When exposed to sufficiently 

 intense discharge we maj' further expect that beryllium will show two separate 

 series spectra of the third order and one of the fourth order. Of these the two former 

 spectra will be analogous to the helium series spectrum of the first order, while 

 the latter will be of the same type as the hydrogen spectrum. In this way we may 

 proceed to elements of higher atomic numbers. For each new element we shall 

 expect, that there will appear a first order series spectrum of a new tj'pe together 

 with spectra of the same types as those of the preceding elements but of higher 

 orders. In this connection it will be observed that the well known similarity, referred 

 to above, between the spectra of low orders of the different elements in the same 

 group of the periodic table may be ascribed to the similar arrangement of the outer 

 electrons in the atoms of these elements, indicated by the general similarity of the 

 physical and chemical properties. The increasing number of inner electrons in the 

 atom with increasing atomic weight of the elements weithin each group, however, 

 may not only be expected to produce the gradual change of the position of the 

 lines of these spectra, especially through the etTect on the outermost of the inner 

 electrons, but it would also seem likely that the presence of the inner electrons in 

 some more direct way may be responsible for the rapidly increasing separation of 

 the components of the lines (doublets, triplets etc.) with increasing atomic weight. 



Even if we cannot account in detail for the eflect of the inner electrons on 



