110 BELL SYSTEM TECHNICAL JOURNAL 



the first /-State of sodium has very nearly the same energy-value 

 as the 4^ State of hydrogen; the second /-State of sodium nearly 

 coincides with the 54 State of hydrogen, and so forth along the se- 

 ciucnce. Hence to the successive States of the /-sequence of the 

 sodium atom one attaches with confidence the symbols 44, 64, 64, and 

 so onward. In some cases this is practicable for the terms of the 

 ^/-sequence also; but ne\'er for those of the 5-sequence. The Stationary 

 States of the s-sequence depart so far from those of hydrogen, that 

 one cannot with any security conclude what values of the Total 

 Quantum Number should be assigned to them. It used to be assumed 

 that ;/ = l for the first term of the 5-sequence and 71 =2 for the first 

 term of the ^-sequence, and the usual notation for the Stationary 

 States reflects this supposition; which however is neither necessary 

 nor probable. 



All of the foregoing interpretations are based upon a theory of the 

 alkali-metal atoms which may be summarized in this way : as the 

 hydrogen atom is supposed to consist of a nucleus surrounded by an 

 inverse-square field through which an electron travels always in one 

 or another of certain orbits determined by quantum-conditions, so 

 also the alkali-metal atom is supposed to consist of a kernel sur- 

 rounded by a not-inverse-square field through which an electron 

 travels always in one or another of certain orbits determined by 

 identical quantum-conditions. As the Stationary States of the 

 hydrogen atom correspond each to a certain orbit and are designated 

 each by certain values of two quantities n and k, or for short by a 

 symbol ;?/, indicating the features of that orbit, so also the Stationary 

 States of the alkali-metal atom correspond each to a certain orbit and 

 are designated each by a symbol tik. For the hydrogen atom we recog- 

 nize the proper nk for each Stationary State because of the wonderful 

 numerical agreement between Bohr's theory and the experimental 

 values for the energy of each State. For the alkali-metal atom we 

 can only guess the proper nk for each Stationary States from indi- 

 cations of much lesser evidential value. We suppose, however, that 

 k = 1,2,3,4 for the various States of the s, p, d and / sequences, respec- 

 tively; so that the 5-sequence is like the «i sequence of hydrogen, the 

 /^-sequence like the Wa sequence, and so on. Of the values of n we 

 are moderately sure for the / and d sequences, quite uncertain for 

 the terms of the 5 and p sequences. 



One may now wonder whether it is possible to invent a central 

 field, such that the orbits traced in it according to the quantum- 

 conditions (2) and (3) would yield a series of energy-values : greeing 

 with the obser\-ed energy-values of the Stationary States of (let me 



