2 There is some equivocation in numbering the periods of the Periodic System. I shall call the First Period H and He; the Second Period running from Li to Ne, and the Third Period from Na to A. I shall speak therefore of C, N, and O as in the Second Period, and Si, P, and S as in the Third.
3 It should be added that the tendency to add electrons, carried to an extreme, ends in forming anions rather than molecules. Their strong tendency to form anions is one reason why fluorine - the most electronegative of the elements - plays no role in biochemistry, and why chlorine plays its principle role as an ion. This is also why I have paid so little attention to chlorine in the discussion of multiple bonding.
4 There is of course an example of what is called a nucleophilic attack, and the hydroxyl ion is here a typical nucleophilic reagent. I avoid this special terminology in the present paper in order to concentrate upon the phenomena themselves.
5 The large repulsion between nonbonding electrons has also been suggested to account for the abnormally low bond energies of N-N (38.4 kcal per mole), O-O (33.2), and F-F (36.6). N, O, and F form strong bonds with other elements in part because they are so highly electronegative; their bonds tend to have considerable ionic character. The bond energies of P-P (53.1 kcal per mole), S-S (50l9), and Cl-Cl (58.0) are relatively strong, because though the presence of more nonbonding electrons could lead to even stronger interatomic repulsions, these repulsions are relieved by the availability in these elements of 3d orbitals (cf. Pauling, 1960, p. 144).
6 See for example Huennekens and Whiteley (1960; footnote on p. 114): "The P= O double bond (i.e., of phosphoric acid) is more correctly written as a 'semi-polar' double bond: P -> O." Compare this with Pauling's discussion (1960, p. 320), concluding "that the available evidence indicates that the older valence-bond formulas . . . with the double bonds resonating among the oxygen atoms, making them equivalent, and with the bonds considered to have partial ionic character, represent the ions (of the oxygen acids) somewhat more satisfactorily . . ."
7 I am much indebted to Professor William Lipscomb of Harvard University for directing my attention to this paper.