relate the curvature of the space-time manifold M* to the distribution of matter. In these equations R~j is the Ricci tensor of M 4, R=g~JRij=R~ is the scalar curvature, T~i is the stress-energy-momentum tensor, T=T~ ~ is the trace of Tij, x is the gravitational constant, and we have chosen units so that c, the speed of light, is 1. There have been a number of attempts at heuristic derivations of these equations, starting of course with EINSTEIN. An especially simple and geometric one is given in WHEELER'S paper [6]. WHEELER'S derivation is admittedly not completely satisfactory, except perhaps in the case of empty space (T~j= 0) in which case the equations reduce to R~=0. WHEELER has expressed the desire for a heuristic derivation that will lead to the full equations, including the important terms 89 R or 89 If such a derivation is given, then the divergence condition T~j = 0 is automatically satisfied, and these equations can then be interpreted as giving the equations of motion. In EINSTEIN'S original derivation he did not obtain the full equations; he later added the term 89 in order to achieve the divergence condition. In this note I propose to give a heuristic derivation of the full equations in the simplest physically realistic situation possible, namely a small ball of fluid sitting in an otherwise empty universe. In this derivation, no use will be made of the famous "geodesic hypothesis". There is much contact here with the 60 years of existing literature of general relativity; the starting point will be a paper of LEVI-CIVITA written in the year 1917. My object has been to follow the most direct route from Newtonian gravitation and special relativity to general relativity. Of course I will be guided by many things that came to light only after EINSTEIN discovered his remarkable equations.
Read full abstract