1915. Where Newton had used the motion of the moon around the earth as a guiding example in his work, Einstein used this deviation of Mercury. The central ideas of general relativity were already in place; at issue was the exact form of the curvature term G in the Einstein equation G=kT. A precise relativistic model of the sun's gravitational field was not needed--Einstein used a simple polynomial approximation. Late in this year he suceeded, and the 43 second lag was eliminated.
1916. A few weeks later, Einstein, working in Berlin, received a paper from Karl Schwarzschild, an astronomer who, though no longer young, was serving in the German army in Russia. Hospitalized by an illness that soon proved mortal, Schwarzschild had time to discover the desired precise relativistic model, and Schwarzschild spacetime replaced the Newtonian model as the best description of the gravitational field of an isolated spherically symmetrical star. But only a few theorists were familiar with relativity, and significant experimental tests were not possible in earth-borne laboratories.
1920s. With the end of the World War, further astronomical tests of general relativity were begun, notably by an expedition led by the British physicist Arthur Eddington. The goal was to compare two observations of a star, one near the sun, the other far from it, to see if gravity could in fact "bend" light as Einstein had predicted. The successful result led to enormous popular and scientific interest in relativity, and the cooperative development of astronomy and relativity in this decade was explosive.
A crucial area of study was stellar evolution. The gravitational collapse of a normal star such as our sun is prevented by nuclear burning. As its fuel is used up, the star must contract, and many end as white dwarfs. These are about the size of the earth but with masses comparable to that of the sun--thus with densities of thousands of tons per cubic inch.
1931. The first relativistic model of the interior of a white dwarf, by the astrophysicist S. Chandrasekhar, produced a simple curve relating its mass and radius. Surprisingly, the larger the mass the smaller the radius. In fact, if the mass is more than about 1.2 solar masses the radius is so small that the star cannot stabilize: further collapse is inevitable.
1934. W. Baade and F. Zwicky predicted that this collapse strips the atoms of their electrons, packing the nucleii together as a neutron star. These are only 10-15 miles in diameter, with densities on the order of a billion tons per cubic inch. While general relativity is useful in the study of white dwarfs, for the superdense neutron stars it is a necessity--Newtonian physics no longer applies.
1939. The first theoretical appearance of black holes, in a paper by Robert Openheimer and H. Snyder: "When all thermonuclear sources of energy are exhausted, a sufficiently heavy star will collapse. Unless [something can somehow] reduce the star's mass to the order of that of the sun, this contraction will continue indefinitely"...past white dwarfs, past neutron stars, to an object cut off from communication with the rest of the universe.
Such discoveries redirected attention to the Schwarzschild model of the exterior of a star. Heretofore it had generally been assumed that the model becomes singular at the Schwarzschild radius r*=2m. This seemed of no great significance since in the case of our sun, for example, whose radius is about 700,000 km, the Schwarzschild model's presumed singularity is buried in at r*=3 km. But for black holes it was gradually realized that radius r*=2m is not a singularity but rather the "horizon" from which nothing, not even light, can emerge.
1954. The Reports of the State University of Kazan (a city 300 miles east of Moscow in the then Soviet Union) contained this year a classification of spacetimes given by the young physicist A.Z. Petrov. The families of radially ingoing and outgoing light rays in the Schwarzschild model show that it has what is now called Petrov type D. Petrov's classification was slow in making its way into the mainstream, but was crucial to the next major development.
All stars rotate. For a very slowly turning star like our sun this is not very important, but when a star collapses, conservation of angular momentum implies that its rate of rotation increases. Thus a neutron star, for example, can be expected to rotate at a fantastic rate. It is believed that pulsars are rapidly rotating neutron stars.
But the star producing Schwarzschild spacetime does not rotate. In l915 this static model had been found in only a few weeks, so it must have been considered fairly easy to set it spinning. However, years passed without success.
1963.. The British-educated New Zealand physicist Roy Kerr, working at the University of Texas, adopted a shrewd strategy: Bearing in mind that Schwarzschild spacetime has Petrov type D, he did not aim directly at the elusive rotating model, but instead examined an algebraically simple class of type D metric tensors. The long-sought metric appeared.
Kerr's minimal one-and-a-half page announcement of his discovery was followed two years later by elaborate detailed calculations.
1967-68. R.H. Boyer and R.W. Lindquist (1967) made Kerr spacetime more accessible by introducing the elegant coordinate system that now bears their names. In the same paper they found the maximally extended Kerr spacetimes and investigated their geodesics. However, a full analysis of Kerr geodesics become possible only with the discovery of a fourth geodesic first-integral by Brandon Carter, a student of Stephen Hawking at the University of Cambridge. Carter's paper (1968) remains the best brief exposition of the global properties of Kerr spacetime.
1968-1978. From the 1960s theoretical and astronomical evidence mounted for the existence of black holes, including very massive ones formed by the collapse of great clusters of stars at the center of galaxies. Speculatively, the high densities following the big bang may have formed primordial holes, including tiny ones. But even for single collapsing stars their wide variety of physical properties might be expected to produce a diversity of black holes. However, a series of results principally by Werner Israel, Brandon Carter, Stephen Hawking, and David Robinson leads to the conclusion that a collapsing star loses its individual characteristics, settling down to a final state uniquely determined by mass and rate of rotation--thus leaving the Kerr model as the prime black hole of nature.
[By the last year (1978) mentioned above, the essentials of black hole theory were in place, and emphasis shifted to the more practical problem of showing beyond all reasonable doubt that black holes exist and determining their role in the organization of the universe. The remarkable results to date are outlined in The Search for Black Holes.]