io6 



NATURE 



[March 24, 192 1 



metry as in Sommerf eld's or Land^'s work, but that 

 their motions are, on the contrary, linked to each 

 other in such a way that it is possible to remove any 

 one of the electrons from the group by a process 

 whereby the orbits of the remaining electrons are 

 altered in a continuous manner. 



These general remarks apply to the constitution and 

 stability of all the groups of electrons ih the atom. On 

 the other hand, the simple variations indicated above 

 of the number of electrons in the groups and sub- 

 groups of successive shells hold only for that region 

 in the atom where the attraction from the nucleus 

 compared with the repulsion from the electrons pos- 

 sesses a preponderant influence on the motion of 

 each electron. As regards the arrangements of the 

 electrons bound by the atom at a moment when the 

 charges of the previously bound electrons begin to 

 compensate the greater part of the positive charge 

 of the nucleus, we meet with new features, and a 

 consideration of the conditions for the binding 

 process forces us to assume that new, added electrons 

 are bound in orbits of a number of quanta equal to, 

 or fewer than, that of the electrons in groups pre- 

 viously bound, although during the greater part of 

 their revolution they will move outside the electrons 

 in these gVoups. Such a stop in the increase, or even 

 decrease, in the number of quanta characterising 

 the orbits corresponding to the motion of the elec- 

 trons in successive shells takes place, in general, 

 when somewhat more than half the total number of 

 electrons is bound. During the progress of the 

 binding process the electrons will at first still be 

 arranged in groups of the indicated constitution, so 

 that groups of three-quanta orbits will again contain 

 eighteen electrons and those of two-quanta orbits 

 eight electrons. In the neutral atom, however, the 

 electrons bound last and most loosely will, in general, 

 not be able to arrange themselves in such a regular 

 way. In fact, on the surface of the atom we meet 

 with groups of the described constitution only in the 

 elements which belong to the family of inactive 

 gases, the members of which from many points of 

 view have also been acknowledged to be a sort of 

 landmark within the natural system of the elements. 

 For the atoms of these elements we must expect the 

 constitutions indicated by the following symbols : 



Helium (2,), 

 Neon (2182), 

 Argon (2,8,8,), 



Krypton (2,8218382), 

 Xenon (2,8218,18382), 

 Niton (2,8218332,18382), 



where the large figures denote the number of elec- 

 trons in the groups starting from the innermost one, 

 and the small figures the total number of quanta 

 characterising the orbits of electrons within each 

 group. 



These configurations are distinguished by an in- 

 herent stability in the sense that it is especially difficult 

 to remove any of the electrons from such atoms so 

 as to form positive ions, and that there will be no 

 tendency for an electron to attach itself to the atom 

 and to form a negative ion. The first effect is due to 

 the large number of electrons in the outermost group; 

 hence the attraction from the nucleus is not com- 

 pensated to the same extent as in configurations where 

 the outer group consists only of a few electrons, as is 

 the case in those families of elements which in the 

 periodic table follow immediately after the elements 

 of the family of the inactive gases, and, as is well 

 known, possess a distinct electro-positive character. 

 The second effect is due to the regular constitution 

 of the outermost group, which prevents a new electron 

 from entering as a further member of this group. In 

 the elements belonging to the families which in the 

 periodic table precede the family of the inactive gases 



NO. 2682, VOL. 107] 



we meet in the neutral atom with configurations of 

 the outermost group of electrons which, on the other 

 hand, exhibit a great tendency to complete them- 

 selves by the binding of further electrons, resulting 

 in the formation of negative ions. 



The general lines of the latter considerations are 

 known from various recent theories of atomic con- 

 stitution, such as those of A. Kossel and G. Lewis, 

 based on a systematic discussion of chemical evidence. 

 In these theories the electro-positive and electro-nega- 

 tive characters of these families in the periodic table 

 are interpreted by the assumption that the outer elec- 

 trons in the atoms of the inactive gases are arranged 

 in especially regular and stable configurations, 

 without, however, any attempt to give a detailed 

 picture of the constitution and formation of these 

 groups. In this connection it may be of interest to 

 direct attention to the fundamental difference between 

 the picture of atomic constitution indicated in this 

 letter and that developed by Langmuir on the basis 

 of the assumption of stationary or oscillating elec- 

 trons in the atom, referred to in Dr. Campbell's letter. 

 Quite apart from the fact that in Langmuir 's theory 

 the stability of the configuration of the electrons is 

 considered rather as a postulated property of the 

 atom, ior which no detailed a priori interpretation is 

 offered, this difference discloses itself clearly by the 

 fact that in Langmuir 's theory a constitution of the 

 atoms of the inactive gases is assumed in which the 

 number of electrons is always largest in the outer- 

 most shell. Thus the sequence of the number of 

 electrons within the groups of a niton atom is, instead 

 of that indicated above, assumed to be 2, 8, 18, 18, 32, 

 such as the appearance of the periods in the sequence 

 of the elements might seem to claim at first sight. 



The assumption of the presence of the larger groups 

 in the interior of the atom, which is an immediate 

 consequence of the argument underlying the present 

 theory, appears, however, to offer not merely a more 

 suitable basis for the interpretation of the general pro- 

 perties of the elements, but especially an immediate 

 interpretation of the appearance of such families 

 of elements within the periodic table, where the 

 chemical properties of successive elements differ only 

 very slightly from each other. The existence of such 

 families appears, in fact, as a direct consequence of 

 the formation of groups containing a larger number 

 of electrons in the interior of the atom when proceed- 

 ing through the sequence of the elements. Thus in 

 the family of the rare earths we may be assumed to 

 be witnessing the successive formation of an inner 

 group of thirty-two electrons at that place in the atom 

 where formerly the corresponding group possessed 

 only eighteen electrons. In a similar way we may 

 suppose the appearance of the iron, palladium, and 

 platinum families to be witnessing stages of the 

 JFormation of groups of eighteen electrons. Compared 

 with the appearance of the family of the rare earths, 

 however, the conditions are here somewhat more 

 complicated, because we have to do with the forma- 

 tion of a Si'oup which lies closer to the surface of the 

 atom, and where, therefore, the rapid increase in the 

 compensation of the nuclear charge during the pro- 

 gress of the binding process plays a greater part. 

 In fact, we have to do in the cases in question, not, 

 as in the rare earths, with a transformation which in 

 its effects keeps inside one and the same group, and 

 where, therefore, the increase in the number in this 

 group is simply reflected in the number ^ of the 

 elements within the family under consideration, but 

 we are witnesses of a transformation which is accom- 

 panied by a confluence of several outer groups of 

 electrons. 



In a fuller account which will be published soon 



