114 BELL SYSTEM TECHNICAL JOURNAL 



three or four or five or six or seven. The preceding notion might be 

 brought under it as an especial case. If it is accepted the theory of 

 atoms other than the alkah-metal atoms will inevitably be more 

 complex than the theory mentioned for these in section P. 



An interesting feature of some of these spectra discloses that the 

 residue of the atom may exist in either of two distinct states. It will 

 be recalled that the energy-values of the Stationary States have been 

 measured from the state of the ionized atom, to which the energy- 

 value zero is assigned. In this fundamental state, one electron and 

 the residue of the atom are completely sundered; and the energy- 

 value of any other Stationary State is the energy required to tear the 

 electron completely out of the atom when the latter is initially in that 

 Stationary State. This definition implies that the state attained when 

 the electron is completely separated from the rest of the atom is de- 

 terminate and unique. Such must be the case if the atom consists of 

 an invariable nucleus and one electron, as in hydrogen; but if the 

 atom contains several electrons, there is no a priori reason for ex- 

 cluding the possibility that there may be several "states of the ionized 

 atom"; in each of these states one electron will be far away, but the 

 residue will have as many different arrangements as there are different 

 states. Extending this idea, one infers that there may be two or 

 more distinct sets of Stationary States for certain elements, each set 

 culminating in a different final configuration of the residue, — that 

 is to say, of the ionized atom. 



Several instances of atoms possessing two such distinct families of 

 Stationary States are known; the most noted is probably that of neon, 

 but I will describe the case of calcium, lately interpreted by Russell 

 and Saunders and independently by Wentzel. Two families of terms 

 "primed" and "unprimed," had been identified in the spectrum of this 

 element, and important sequences of each could be followed suffi- 

 ciently far to make the extrapolation to the limit not too daring. The 

 limits were different, showing that the amount of energy required to 

 separate an electron from an atom initially in its normal state had 

 two values differing from one another by 1.72 equivalent volts. Con- 

 sequently the residue may remain (it is not necessary to assume that 

 it can long remain) in either of two States differing from one another 

 (when the extra electron is far away) by this amount. 



At this point a very significant numerical agreement enters upon 

 the scene. The residue of the calcium atom, the ionized-calcium atom, 

 has itself a spectrum which is known, and from which its system of 

 Stationary States has been learned and mapped. Like the systems of 

 Stationary States possessed by neutral atoms, this one includes 



