454 BELL SYSTEM TECllXICAL JOURNAL 



lluis culminates in a (iiagrain of stationary states, as for llu- ojilical 

 spectra. 



Such a diagram is shown in l"i^'. IH, wliich is for an t-k'nu-ni far up 

 in the periodic system, therefore with a rich s>'stcm of X-ray hnes 

 and stati<jnary states. In comparing it with one of the diagrams 

 made for optical spectra, it must be remembered that its scale is 

 enormously more compressed — the distance from top to bottom cor- 

 responds to about one hundred thousand equivalent volts. Each line 

 in the X-ray spectrum corresponds to an arrow between two of the 

 le\'els, but not every arrow- corresponds to a line. Again there is a 

 selection-principle, and this selection-principle is partly expressed by 

 attaching a double index to each of the levels. When the indices 

 are assigned as in Fig. 13, transitions between levels for which the 

 second numeral differs by one unit include the onh' ones which actualK' 

 occur. But this is not the complete selection-principle; it is neces- 

 sar>' to add that in any actually occurring transition, the first numeral 

 must change b\' one or more units; and further, that transitions may 

 occur only between levels to which ditTiTt-ni k'tiers are attached. 

 The first numeral is designated li\ n. tin- sicdnd !>> k; they are called 

 the total and the azimuthal quantiini-nunilicr. 



The levels are also frecjuently known li\' kilcrs with suliscript 

 numerals, as the diagram shows. The letters by now are i>retty 

 definitely fixed, but the subscripts are still being shuttled around. 

 The notation for the X-ray lines is in a terrible state. 



A curious and evidently important feature of the.se levels is, that 

 when an atom is put into any one of them — say into the K level, 

 or the Li level, or the L« le\el — it extrudes an electron. < )r. in other 

 words, each of these stationary states is a state in wliich the atom 

 lacks one of its electrons — like the "ionized-atom" state from which 

 we previously measured the energy-values in dealing with the optical 

 spectra. All of them, at least the highest ones, are in fact "ioni/ed- 

 atom states." Since, how'c\er, they are all different, it is natural to 

 suppose that a different electron is missing, or that an electron is 

 missing from a different place, in each of the different cases. .\|)- 

 |)arentl\' an atom cannot enter into a sl.iii(>nar\ si.ite with so liigh 

 an energy, and remain neutral. 



We must pause to consider from wji.u si.mdaiil slate the energy- 

 values of these stationary states are measuretl. In the prexious case 

 of the optical spectra, the energy-values of the stationary states were 

 measured, so to speak, downwards from the state of the ionized atom 

 to the normal state of the neutral atom; the energy of the ionized 

 atom was set equal to zero, that f)f the neutral atom in its normal state 



