This NATO Advanced Study Institute was concerned with modern ab initio methods for the determination of the electronic structure of molecules. Recent years have seen considerable progress in computer technology and computer science and these developments have had a very significant influence on computational molecular physics. Progress in computer technology has led to increasingly larger and faster systems as well as powerful minicomputers. Simultaneous research in computer science has explored new methods for the optimal use of these resources. To a large extent develop ments in computer technology, computer science and computational molecular physics have been mutually dependent. The availability of new computational resources, particularly minicomputers and, more recently, vector processors, has stimulat'ed a great deal of research in molecular physics. Well established techniques have been reformulated to make more efficient use of the new computer technology and algorithms which were previously computationally intractable have now been successfully implemented. This research has given a new and exciting insight into molecular structure and molecular processes by enabling smaller systems to be studied in greater detail and larger systems to be studied for the first time.

This volume records the lectures given at a NATO Advanced Study Institute on Methods in Computational Molecular Physics held in Bad Windsheim, Germany, from 22nd July until 2nd. August, 1991. This NATO Advanced Study Institute sought to bridge the quite considerable gap which exist between the presentation of molecular electronic structure theory found in contemporary monographs such as, for example, McWeeny's Methods 0/ Molecular Quantum Mechanics (Academic Press, London, 1989) or Wilson's Electron correlation in moleeules (Clarendon Press, Oxford, 1984) and the realization of the sophisticated computational algorithms required for their practical application. It sought to underline the relation between the electronic structure problem and the study of nuc1ear motion. Software for performing molecular electronic structure calculations is now being applied in an increasingly wide range of fields in both the academic and the commercial sectors. Numerous applications are reported in areas as diverse as catalysis and interstellar chernistry, drug design and environmental studies, molecular biology and solid state physics. The range of applications continues to increase as scientists recognize the importance of molecular structure studies to their research activities. Recent years have seen a growing dependence of these applications on program packages, which are often not in the public domain and which may have a somewhat lirnited range of applicability dicta ted by the particular interests and prejudices of the program author.

This book contains the transcripts of the lectures presented at the NATO Advanced study Institute on "Computational Techniques in Quantum Chemistry and Molecular Physics", held at Ramsau, Germany, 4th - 21st Sept. 1974. Quantum theory was developed in the early decades of this century and was first applied to problems in chemistry and molecular physics as early as 1927. It soon emerged however, that it was impossible to con sider any but the simplest systems in any quantita tive detail because of the complexity of Schrodinger's equation which is the basic equation for chemical and molecular physics applications. This remained the si tuation until the development, after 1950, of elec tronic digital computers. It then became possible to attempt approximate solutions of Schrodinger's equa tion for fairly complicated systems, to yield results which were sufficiently accurate to make comparison with experiment meaningful. Starting in the early nineteen sixties in the United States at a few centres with access to good computers an enormous amount of work went into the development and implementation of schemes for approximate solu tions of Schrodinger's equation, particularly the de velopment of the Hartree-Fock self-consistent-field scheme. But it was soon found that the integrals needed for application of the methods to molecular problems are far from trivial to evaluate and cannot be easily approximated.

Recent years have seen the proliferation of new computer designs that employ parallel processing in one form or another in order to achieve maximum performance. Although the idea of improving the performance of computing machines by carrying out parts of the computation concurrently is not new (indeed, the concept was known to Babbage ), such machines have, until fairly recently, been confined to a few specialist research laboratories. Nowadays, parallel computers are commercially available and they are finding a wide range of applications in chemical calculations. The purpose of this volume is to review the impact that the advent of concurrent computation is already having, and is likely to have in the future, on chemical calculations. Although the potential of concurrent computation is still far from its full realization, it is already clear that it may turn out to be second in importance only to the introduction of the electronic digital computer itself.

In a way the MOTECC-89 project started in the early sixties at the IBM Research Laboratory in San Jose, California. The six years of post-doctoral research, first with Giulio Natta on conductive polymers, with Michael Kasha on spin-orbit effects, with Kenneth S. Pitzer on high temperature molecules and thermo dynamics and with R. S. Mulliken in the quantum chemistry of small molecules had demonstrated pragmatically the importance of a broad-based research and also let me taste some of the excitement to be derived from interdisciplinarity. Thus when I started to gather a department in the newly opened IBM Research Laboratory in San Jose, California, I purposely named it "Large Scale Scientific Computation Department," avoiding reference to chemistry, physics, statistical mechanics or fluid dynamics, which were our main tasks. In the sixties interdisciplinarity was more and more recognized as a most important if not nec essary avenue to cope with the technical needs of our society. However, at that time interdisciplinarity was synonymous with "team work," and true interdisciplinarity was a formidably difficult objective. Although I headed an excellent group of scientists with different backgrounds and there was much progress and creativity, still each one of us was more or less conducting his own research in his own field with occasional cross-field partnerships and with some of the computational techniques as our common base. Later, in 1974, I made a second attempt, this time starting a new department at the Donegani Institute, Montedison, in Novara, Italy.

This book describes the use of modern computational methods in predicting high resolution molecular spectra, which allows the experimental spectroscopist to interpret and assign real spectra. * Offers a comprehensive treatment of modern computation techniques. * Provides a collection of material from different areas of theoretical chemistry and physics. * Bridges the gap between traditional quantum chemistry and experimental molecular spectroscopy.

Molecular descriptors are mathematical values that describe the structure or shape of molecules, helping predict the activity and properties of molecules in complex experiments. This book describes the equations known as QSAR (quantitative structure-activity relationships) and QSPR (quantitative structure-property relationships), showing how they can be used productively in a wide range of industries.

This text provides a summary of the key ideas, techniques and research results in the field of atomic, molecular and optical physics. It is intended to be of interest to all physicians (especially in condensed matter and chemical physics) and engineers, as