Robertson returned to the United States in 1927, and became an
assistant professor of mathematics at Caltech. In 1928, he accepted a position as an assistant professor of mathematical physics at
Princeton University, where he became an
associate professor in 1931, and a professor in 1938. He spent 1936 on sabbatical at Caltech. His interest in
general relativity and
differential geometry led to a series of papers in the 1920s that developed the subject. (translated from the second, revised German edition by Robertson) Robertson wrote three important papers on the mathematics of
quantum mechanics. In the first, written in German, he looked at the coordinate system required for the
Schrödinger equation to be solvable. The second examined the relationship between the
commutative property and Heisenberg's
uncertainty principle, generalizing the latter for any two
Hermitian operators. The third extended the second to the case of m observables. In 1931 he published a translation of Weyl's
The Theory of Groups and Quantum Mechanics. It was Robertson's anonymous 1936 critical peer review of a paper submitted by Albert Einstein to
Physical Review which caused Einstein to withdraw the paper from consideration. Yet perhaps Robertson's most notable achievements were in applying relativity to
cosmology. He independently developed the concept of an expanding universe, which would imply distant galaxies as seen from Earth would be
redshifted—a phenomenon previously confirmed by
Vesto Slipher . Robertson went on to apply the theory of continuous groups in
Riemann spaces to find all the solutions that describe the cosmological spaces. This was extended by
Arthur Geoffrey Walker in 1936, and is today widely known in the
United States as the
Robertson–Walker metric. One of Robertson's landmark papers, a brief note in
The Annals of Mathematics, entitled a "Note on the preceding paper: The two body problem in general relativity", solved that problem within a degree of approximation not improved on for several decades. Earlier work, such as the
Schwarzschild metric, were for a central body that did not move, while Robertson's solution considered two bodies orbiting each other. Nevertheless, his solution failed to include
gravitational radiation, so the bodies orbit forever, rather than approaching each other. Yet Robertson's name is most often associated with the
Poynting–Robertson effect, the process by which
solar radiation causes a dust mote orbiting a star to lose
angular momentum. This is related to
radiation pressure tangential to the grain's motion.
John Henry Poynting described it in 1903 based on the "luminiferous aether" theory, which was superseded by Einstein's theories of relativity. In 1937, Robertson described the effect in terms of general relativity. Robertson developed the theory of
invariants of tensors to derive the
Kármán–Howarth equation in 1940, which was later used by
George Batchelor and
Subrahmanyan Chandrasekhar in the theory of axisymmetric turbulence to derive
Batchelor–Chandrasekhar equation. == World War II ==