The pair-defect-sum approximation to the bond strength of a hybrid orbital (angular wave functions only) is compared to the rigorous value as a function of bond angle for seven types of bonding situations, with between three and eight bond directions equivalent by geometrical symmetry operations and with only one independent bond angle. The approximation is seen to be an excellent one in all cases, and the results provide a rationale for the application of this approximation to a variety of problems.The quantum mechanical treatment of the electronic structure of molecules has great value. For all except the simplest molecules, it is so difficult to carry out a thorough calculation as essentially to prevent its application. One simplification in the theoretical discussion of the nature of the bonds formed by an atom is to assume that the bond-forming strength of an atomic orbital is proportional to the amount of overlap of the orbital with a bond orbital of an adjacent atom. The amount of overlap increases with increase in the concentration of the bond orbital in the bond direction. A second approximation was made by one ofus in 1931. This approximation is based on the knowledge that the radial parts of the atomic orbitals involved in the formation of hybrid bond orbitals are not much different for the different atomic orbitals, so that the assumption that they are the same does not lead to great error. Accordingly attention is, focused on the angular distribution of the bond orbitals. The strength, S, of a bond orbital was defined as the value in the bond direction of the angular part of the bond orbital, normalized to 4ir over the surface of a sphere (1, 2). A number of sets of bond orbitals were discussed in this way (1, 2). Hultgren then attacked the general problem of the best possible sets of sp3d5 orbitals (3). The mathematical problem was, however, too difficult to handle before computers had been developed, and Hultgren made a simplifying assumption, the assumption of cylindrical symmetry of the bond orbitals. This assumption led to the erroneous conclusion that sets of more than six bond orbitals would not be well suited to bond formation. Nearly 40 years later the general problem of the best set of nine spd bond orbitals was attacked by McClure (4). He found one excellent set of nine hybrid bond orbitals, all with strength close to the maximum possible, 3. Because of the difficulty in carrying out the calculations rigorously for large sets oforbitals, an approximation was then suggested. This approximation is based on the equations giving the strength SO of two equivalent bond orbitals that have the maximum strength in directions making the angle a with one another. For sp3, sp3d5, and sp3d5f7 hybrid orbitals the equations for the bond strength SO (a) have been derived (5, 6)-Eqs. 1, 2, and 3, respectively. [2] [3] where x = cos2 (a/2). For sp3 hybrid orbitals, the maximum strength Sma=, 2, is found at the tetrahedral bond angle 109.470; for sp3d5 hybrid orbitals, Smax = 3 at a = 73.15°and a...