The architecture of the argon atom is in a certain sense less complete than that of the neon atom. In argon there are indeed four orbits of the 3₁ type and four of the 3₂ type, but the third kind of 3-quantum orbit, the circular 3₃ one, is missing. Nor does it appear in the next element, potassium (19). The electron No. 19 prefers, instead of the 3₃ orbit, the 4₁ orbit, which consists of oblong loops and which gives a firmer binding because it dives in among the electrons bound earlier, while the circle 3₃ would lie outside them all. We thus obtain an atom of type similar to the lithium and sodium atoms. But the slighted 3₃ path lies, so to speak, on the watch to steal a place for itself in the neutral atom, and this has grave results for the subsequent development. Even in calcium (20), after the first eighteen electrons are bound in the argon architecture, both the nineteenth and the twentieth go into a 4₁ orbit, and the behaviour of calcium is like that of magnesium. But since the increasing nuclear charge means for the electron No. 19 a decrease in the dimensions and an increase in the binding of the orbits corresponding to the quantum number 3₃, a point will finally be attained where the 3₃ orbit of the nineteenth electron lies within the boundaries of what may be called the argon system, i.e., the architecture corresponding to the first eighteen electrons, and corresponds to a stronger binding than a 4₁ orbit would do. In scandium (21) the 3₃-type orbit occurs for the first time in the neutral atom and will not only come into competition with the 4₁ type, but will also cause a disturbance in the 3-quantum groups, which in the following elements must undergo reconstruction. As long as this lasts the situation is very complicated and uncertain. When the reorganization is almost completed, we come to the blotting out of chemical differences, particularly known from the triad, iron, cobalt and nickel. Moreover, there comes a fluctuation in the valency of the elements. Iron can, as has been said, be divalent, trivalent or hexavalent. This oscillation in valency begins in titanium.
We should perhaps expect that the reconstruction would be completed long before nickel (28) is reached, because even with twenty-two electrons we could get four orbits of each of the 3-quantum types (3₁, 3₂ and 3₃); but from the chemical facts we are led to assume that in a completed group of 3-quantum orbits there can be room for six electrons in each sub-group. At first sight we should, then, expect the end of the reconstruction with nickel, which has indeed eighteen electrons more than neon where the group of 2-quantum orbits was completed. We might expect that nickel would be an inert element in the series with helium, neon, and argon. On the contrary, nickel merely imitates cobalt. This is explained by the fact that the group of eighteen 3-quantum orbits, although it has a symmetric architecture, is weakly constructed if the nuclear charge is not sufficiently large. The binding of this group is too weak for it to exist as the outer group in a neutral atom. In nickel the electrons, in a less symmetrical manner, will probably arrange themselves with seventeen 3-quantum orbits and one 4-quantum orbit.
The group of eighteen 3-quantum orbits becomes stable, however, when the nuclear charge is equal to or larger than 29, in which case it can become the outer group in a positive ion. In this we find the explanation of the properties of the atom of copper. The neutral copper atom has its twenty-ninth electron bound in a 4₁ orbit consisting of oblong loops (cf. diagram at the end); this electron can easily be freed and leaves a positive monovalent copper ion with a symmetrical architecture. Even under these circumstances, although possessing a certain stability, the ion is not very firmly constructed. Thus the fact that copper can be both monovalent and divalent, must be explained by the assumption that for a nuclear charge 29, the 3-quantum group still easily loses an electron.
When we come to zinc (30) the group of eighteen is more firmly bound; zinc is a pronounced divalent metal which in its properties reminds us of calcium and magnesium. From zinc (30) to krypton (36) we have a series of elements which in a certain way repeat the series from magnesium (12) to argon (18).
In [Fig. 34] is shown Bohr’s arrangement of the periodic system in which the systematic correlation of the properties of the element appears somewhat clearer than in the usual plan ([cf. p. 23]). It shows great similarity with an arrangement proposed nearly thirty years ago by the Danish chemist, Julius Thomsen. The elements from scandium to nickel, where, in the neutral atom, the electron group of 3-quantum orbits is in a state of reconstruction, are placed in a frame; the neutral oblique lines connect elements which are “homologous,” i.e., similar in chemical and physical (spectral) respect.
Fig. 34.—The periodic system of the elements. The elements where an inner group of orbits is in a stage of reconstruction are framed. The oblique lines connect elements which in physical and chemical respects have similar properties.
In krypton (36) we again have a stable architecture with an outer group of eight electrons, four in 4₁ orbits and four in 4₂ orbits. Owing to the appearance of 4₃ orbits in the normal state of the atoms of elements with atomic number higher than 38, there follows in the fifth period of the natural system a reconstruction and provisional completion of the 4-quantum orbits to a group of eighteen electrons, which shows a great simplicity with the completion of the 3-quantum group in the fourth period. In [Fig. 34] the elements where the 4-quantum group is in a state of reconstruction are framed. The 4-quantum group with eighteen electrons is of more stable construction than the group of eighteen 3-quantum orbits in the elements with an atomic number lower by eighteen. This is due to the fact that all the orbits in the first-mentioned group are oblong and therefore moored, so to say, in the inner groups, while in the complete group of 3-quantum orbits there are six circular orbits. This is the reason why silver, in contrast to copper, is monovalent.
The next inactive gas is xenon (54), which outside of the 4-quantum group has a group of eight electrons in 5-quantum orbits, four in 5₁ orbits and four in 5₂ orbits. We notice that in xenon the group of 4-quantum orbits still lacks the 4₄ orbits. On the theory we must, therefore, expect to meet a new process of completion and reconstruction when proceeding in the system of the elements. The theoretical argument is similar to that which applies in the case of the completion of the 3-quantum group which takes place in the fourth period of the natural system. In fact, in the formation of the normal atoms of the elements next after xenon, caesium, 55, and barium, 56, the fifty-four electrons first captured will form a xenon configuration, while the fifty-fifth electron will be bound in a 6₁ orbit, consisting of very oblong loops, which represents a much stronger binding than a circular 4₄ orbit. Calculation shows, however, that with increasing nuclear charge there must soon appear an element for which a 4₄ orbit will represent a stronger binding than any other orbit. This is actually the case in cerium (58), and starting from this element we meet a series of elements where, in the normal neutral atom, the 4-quantum group is in a state of reconstruction. This reconstruction must occur far within the atom, since the group of eighteen 4-quantum orbits in xenon is already covered by an outer group of eight 5-quantum orbits. The result is a whole series of elements with very slight outward differences between their neutral atoms, and therefore with very similar properties. This is the rare earths group, which in such a strange way seemed to break down the order of the natural system ([cf. p. 21]), but which thus finds its natural explanation in the quantum theory of the structure of the atom.
The elements in which the 4-quantum group is in a state of reconstruction are, in [Fig. 34], enclosed in the inner frame in the sixth period. Moreover, in the outer frame all elements are enclosed where the group of 5-quantum orbits is in a state of reconstruction, which started, even before cerium in lanthanum (57), where the fifty-fifth electron in the normal state is bound in a 5₃ orbit. The element cassiopeium, with atomic number 71, which is the last of the rare earths, stands just outside the inner frame, because in the normal neutral atom of this element the 4-quantum group is just completed; this group, instead of eighteen electrons with six electrons in each sub-group, consists now of thirty-two electrons with eight electrons in each sub-group. The theory was able to predict that the element with atomic number 72, which until a short time ago had never been found, and the properties of which had been the subject of some discussion, must in its chemical properties differ considerably from the trivalent rare earths and show a resemblance to the tetravalent elements zirconium (40), and thorium (90). This expectation has recently been confirmed by the work of Hevesy and Coster in Copenhagen, who have observed, by means of X-ray investigations, that most zircon minerals contain considerable quantities (1 to 10 per cent.) of an element of atomic number 72, which has chemical properties resembling very much those of zirconium, and which on this account had hitherto not been detected by chemical investigation. A preliminary investigation of the atomic weight of this new element, for which its discoverers have proposed the name hafnium (Hafnia = Copenhagen), gave values lying between 178-180, in accordance with what might be expected from the atomic weight of the elements (71) and (73). ([Cf. p. 23].)