From the New York Times, October 19, 1983:
Professor Harish-Chandra began his work in the theory of elementary particles, but he turned in 1949 from physics to mathematics. His main interest was the study of infinite-dimensional representations. They first arose in quantum mechanics during the 1930's, when it was necessary to analyze the effect of symmetries on the motions of particles and waves. For these representations, Professor Harish-Chandra created a theory with implications for many domains, from geometry to number theory.