DENNISON, DAVID MATHIAS

(b. Oberlin, Ohio, 26 April 1900; d. Ann Arbor, Michigan, 3 April 1976)

physics.

Dennison was the son of Walter Dennison, a professor of classics at Oberlin College, and of the former Anna L. Green. From 1902 to 1910 the Dennisons lived in Ann Arbor, Michigan, where David’s father taught Latin at the University of Michigan. The family then moved to Swarthmore, Pennsylvania, where in 1921 David received his A.B. from Swarthmore College. Dennison was interested in science from an early age, but his first course in physics did not arouse his interest and he majored in mathematics. It was a casual conversation with H. C. Hayes, a Swarthmore physicist, and a summer job in 1920 at General Electric, working with Irving Langmuir, that drew Dennison to physics. He enrolled at the University of Michigan, where he was the first student to submit a theoretical dissertation to the department of physics. He received his Ph.D. in 1924 and, on 18 August of that year, married Helen Lenette Johnson. They had two sons.

Dennison’s doctoral research, on the structure of the methane molecule, was directed by Oskar Klein, who had earlier been assistant to Niels Bohr. When Bohr came to the United States in the fall of 1923, he visited Klein at Ann Arbor and met Dennison. Bohr’s impression was favorable, and he invited Dennison to Copenhagen. Dennison was awarded a two-year fellowship by the National Education Board and with it went to Bohr’s Institute for The oretical Physics in Copenhagen.

Dennison was in Europe at a most opportune time. He arrived in Copenhagen in late summer 1924, during the waning months of the old quantum theory. During his three years in Europe, Dennison witnessed the creation of quantum mechanics, starting with Werner Heisenberg’s version (June 1925), through Erwin Schrödinger’s and P. A. M. Dirac’s versions, and ending in the spring of 1927, when he attended Heisenberg’s first seminar on the uncertainty principle. “To all of us who heard him,” Dennison recalled, “there was never any question but that this was a really wonderful development.”

Dennison’s first work at Copenhagen continued the analysis of molecular spectra along the lines he had developed in his doctoral dissertation. Although some of his interpretations—made prior to the creation of quantum mechanics—were wrong, the methods Dennison established set the pattern of the subsequent research on molecular structure. After Heisenberg created quantum mechanics, Dennison applied the new mechanics to a symmetric-top molecule, using matrix methods to calculate the rotational energy states, selection rules, and intensities of a symmetric-top rotator. This work, published in 1926, was the first application of matrix mechanics to appear in Physical Review.

A fellowship from the University of Michigan allowed Dennison to stay in Europe a third year; he went to Zurich, where Schrödinger was the attraction. There, Dennison worked on homopolar diatomic molecules; later, at Cambridge, during the spring of 1927, this line of work led to Dennison’s most noteworthy contribution. Ralph Fowler had invited Dennison to give three lectures to his graduate class, and in preparing the third of these lectures, Dennison resolved the disparity between calculated and measured values for the specific heat of the hydrogen molecule. In this work he was influenced by Heisenberg’s 1926 quantum mechanical treatment of the helium atom. Heisenberg explained the puzzling spectrum of helium by recognizing that the wave function for helium can be either symmetric or antisymmetric with respect to the exchange of electrons. In a similar fashion Dennison recognized that the rotational states can be either symmetric or antisymmetric with respect to an exchange of nuclei. Transitions between these symmetric and antisymmetric states are forbidden unless the proton has a spin.

Dennison assumed that the proton, like the electron, had a spin; however, he recognized that the magnetic moment of the proton was so small that transitions between the symmetric and antisymmetric states were so slow that hydrogen gas at low temperatures could be regarded as a mixture of a gas with symmetric rotational states and a gas with antisymmetric ones. When hydrogen was considered as a mixture of two gases, each with its own specific heat, theory agreed with experiment. Dennison’s work on the specific heat of hydrogen, published in Proceedings of the Royal Society of London, provided the first quantitative evidence for the spin of the proton.

In 1927 Dennison returned to Ann Arbor, where he had received an instructorship in physics at the University of Michigan. He rose rapidly through the ranks and became a full professor in 1935. During this period the University of Michigan became a center for theoretical physics in the United States. In addition to Dennison, S. A. Goudsmit. G. E. Uhlenbeck, and Otto Laporte were full-time members of the faculty. The influence of the University of Michigan physics department was further enhanced by the Summer Symposia in The oretical Physics, which began in 1928 and continued until 1940. These seminars attracted theoretical physicists from both the United States and leading centers of physics in Europe.

Throughout his career Dennison’s primary interest was the application of quantum mechanics to the structure of molecules and the interpretation of molecular spectra. In large part owing to Dennison, the University of Michigan became a world center for both theoretical and experimental molecular studies. An early example of the fruitful interaction between Dennison and the experimentalists occurred in 1932 when Dennison, with George Uhlenbeck, solved the quantum mechanical two-minimum problem that involves the quantum mechanical effect of tunneling. This theory, when applied to the pyramid-shaped ammonia molecule (NH₃) predicted that the nitrogen atom, at the peak of the pyramid, could “tunnel” through the plane of the three hydrogen atoms, thereby inverting the pyramid. This nonclassical effect had spectroscopic consequences, and Dennison persuaded a colleague to perform the experiment. Theresult agreed with Dennison’s predictions; the experiment was the first in what became microwave spectroscopy. Throughout the 1930’s Dennison studied the vibrational and rotational behavior of molecular systems and brought his theoretical studies to bear on infrared spectroscopic data.

During World War II, Dennison worked on proximity fuses and was cited for this work by the U.S. Navy. After the war an electron accelerator was built at the University of Michigan. Dennison’s colleague H. R. Crane designed the accelerator to have both curved and straight portions. Dennison, with Theodore Berlin, established the general conditions for the stability of such orbits; their paper became a basic reference for accelerator builders. But the accelerator, the glamour tool of postwar physics, could not keep Dennison from his first love: molecules.

Dennison’s fascination with molecules was shared by a number of first-rate American physicists and chemists: E. U. Condon, Robert Mulliken, J. Robert Oppenheimer, Linus Pauling, John Slater, J. H. Van Vleck, and H. C. Urey. These scientists did not participate in the creation of quantum mechanics, but with the new theory in place, they immediately started to apply the new theoretical formalism to atoms, molecules, and solids. Their work played a significant part in the emergence of the United States as a leading center for physics during the 1930’s.

The field of molecular physics, due in large part to Dennison’s contributions, became a highly refined area of research after the war. Until he retired as Harrison M. Randall professor of physics in 1971. Dennison worked on such problems as vibrationalrotational interactions, centrifugal distortion effects, and hindered rotations in molecules. Among his honors were election to the National Academy of Sciences in 1953 and fellowship in the American Physical Society.

BIBLIOGRAPHY

I. Original Works. A complete list of Dennison’s publications follows the article by H. Richard Crane in Biographical Memoirs. National Academy of Sciences, 52 (1980), 139–159. A personal account of his early years in physics is given in “Recollections of Physics and Physicists During the 1920s,” in American Journal of Physics, 42 (1974), 1051–1056.

Dennison’s papers are deposited in the Bentley Library, University of Michigan.

II. Secondary Literature. Dennison’s physical research is described in many treatises on molecular spectroscopy. See, for example, Walter Gordy and Robert L. Cook. Microwave Molecular Spectra (New York, 1970); and Gerhard Herzberg, Molecular Spectra and Molecular Structure, J. W. T. Spinks, trans., 2 vols., 2nd ed. (Princeton, 1950).

John S. Rigden