The foregoing arguments represent the foundation of Einstein's theory of gravity. In the Newtonian view the sun produces in the space around it a field of force that makes the planets move along curved trajectories instead of straight lines. In Einstein's picture space itself becomes curved and the planets move along the straightest (geodesic) lines in that curved space. Here we are speaking of geodesics in the four-dimensional space-time continuum. It would, of course, be wrong to say that the orbits themselves are geodesic lines in three-dimensional space.
Einstein's interpretation of gravity as the curvature of space-time does not lead to exactly the same results as those of the classical Newtonian theory. We have already mentioned the bending of light. The relativistic theory also gives slightly different answers for the motions of material bodies. For example, it explained the difference between the calculated and observed rates of precession of the major axis of Mercury's orbit, which represented a long-standing mystery of classical celestial mechanics.
Newton's law of gravitational interaction between masses is quite similar to the law of electrostatic interaction between charges, and Einstein's theory of the gravitational field has many common elements with James Clerk Maxwell's theory of the electromagnetic field. So it is natural to expect that an oscillating mass should give rise to gravitational waves just as an oscillating electric charge produces electromagnetic waves. In a famous article published in 1918 Einstein indeed obtained solutions of his basic equation of general relativity that represent such gravitational disturbances propagating through space with the velocity of light. If they exist, gravitational waves must carry energy; but their intensity, or the amount of energy they transport, is extremely small. For example, the earth, in its orbital motion around the sun, should emit about .001 watt, which would result in its falling a millionth of a centimeter toward the sun in a billion years!
No one has yet thought of a way to detect waves so weak. In fact, some theorists, among them Sir Arthur Eddington, have suggested that gravitational waves do not represent any physical reality at all but are simply a mathematical fiction that can be eliminated from the equation by a suitable choice of space-time co-ordinates. More thorough analysis indicates, however, that this is not the case and that gravitational waves, weak though they may be, are real.
Are gravitational waves divided into discrete energy packets, or quanta, as electromagnetic waves are? This question, which is as old as the quantum theory, was finally answered two years ago by the British physicist P. A. M. Dirac. He succeeded in quantizing the gravitational-field equation and showed that the energy of gravity quanta, or "gravitons," is equal to Planck's constant, h, times their frequency-the same expression that gives the energy of light quanta or photons. The spin of the graviton, however, is twice the spin of the photon.
Because of their weakness gravitational waves are of no importance in celestial mechanics. But might not gravitons play some role in the physics of elementary particles? These ultimate bits of matter interact in a variety of ways, by means of the emission or absorption of appropriate "field quanta." Thus electromagnetic interactions (for example the attraction of oppositely charged bodies) involve the emission or absorption of photons; presumably gravitational interactions are similarly related to gravitons. In the past few years it has become clear that the interactions of matter fall into distinct classes: (1) strong interactions, which include electromagnetic forces; (2) weak interactions such as the "beta decay" of a radioactive nucleus, in which an electron and a neutrino are emitted; (3) gravitational interactions, which are vastly weaker than the ones called "weak."