The bond diagrammatic representation of molecules is the foundation of MOVB theory. To a certain extent, this kind of representation is analogous to the one on which "resonance theory" is based and this fact can be projected by a comparison of the various ways in which MOVB theory depicts a species made up of three core and two ligand MO's which define two subsystems containing a total of six electrons and the ways in which "resonance theory" (i. e. , qualitative VB theory) depicts a six-electron-six-AO species such as the pi system of CH =CH-CH=CH-CH=O. The 2 different pictorial representations are shown in Scheme 1 so that the analogies are made evident. First of all, the total MOVB diagrammatic representation of the 6/5 species is obtained by a linear combination of three complete bond diagrams, as in Al, which describe the optimal linear combination of!l! MOVB Configuration Wavefunctions (CW's). By the same token, a total VB diagrammatic representation of the 6/6 species can be obtained by writing a "dot structure", as in Bl, and taking this to mean the optimal linear combination of all VB CW's. Next, we can approxi mate the MOVB wavefunction of the 6/5 species by one complete (or detailed) bond dia gram" (A2). No simple VB representation analogy can be given in this case. Alterna tively, we can approximate the MOVB wavefunction by a linear combination of compact bond diagrams, as in A3, in the way described before.

In the last fifty years. computational chemistry has made impressive strides. Huckel NO computations were rapidly succeeded by semiempirical monodeterminantal Self Consistent Field (SCF) MO calculations which now give way to high quality ab initio calculations of the poly-determinantal SCF-MO and Generalized VB variety. By contrast. no analogous progress has been made in the area of the qualitative theo~ of chemical bonding. In fact. more than a half-centu~ after the exposition of HUckel MO theory the conceptual superstructure of chemist~ is still founded on it. This is made glaringly evident by the fact that highly sophisticated computations are still interpreted with primitive HUckel MO theory. despite the fact that most chemists are well aware of its formal deficiencies. The current popularity Qf qual1tati. ve MO theory among experimental i sts is not the resul t ~f fonnai -advances ~Wt, rather the consequence of stimulating application of old MO theoreti~a. 1 ~oncepts. : . . " This work attemps to improve this situation by outlining a~t. iJlitative theory of chemical bonding which operates at a high level of theoretical sophistication. It was first presented at the NATO Advanced Study Institute on "Topics in Theoretical Organic Chemistry" in Gargnano. Italy. in June 1979. and in other international meetings and conferences. colloquia. and informal gatherings in the period of time follOWing the Gargnano meeting. It was also presented in a seminar given at the University of Washington in October 1980.