Fifty Years of Cosmology

The worrying situation at that time in cosmology, as it seemed, turned out to be a relatively minor matter, namely the choice of suitable coordinates. Even the best-known cosmologists – de Sitter, Eddington and Lemaitre – had chosen coordinates appropriate to localities in the universe rather than the whole. This produced a sense of mystery that was more apparent than real as to what happened at the boundary of a locality. It is one of the features of Einstein's general relativity that when you choose coordinate systems with special properties you can mistakenly come to think of the properties as physical instead of as mathematical artefacts. Early workers on gravitational waves thought they were investigating physical waves when in fact the waves were in their coordinate system, and a similar situation existed in cosmology. It was also in 1935–36 that this situation was put right, by H.P. Robertson in the United States and A.E. Walker in Britain and the resulting choice of coordinates later became known as the Robertson-Walker line element. Then in 1937 Robertson published an important article on cosmology in the Reviews of Modern Physics, which unfortunately I didn't read at that time because my research interests were in quantum mechanics and nuclear physics.