Variable-G Gravitation

Abstract

Numerous theories have been proposed in which the Newtonian gravitational parameter G is variable, but progress in astrophysics and geophysics has winnowed out the field somewhat so that today there remain only three main competitors: (a) Dirac’s theory (Dirac, 1938, 1973); (b) the Hoyle/Narlikar theory (Hoyle and Narlikar, 1971a); and (c) the scale-covariant theory of Canuto et al. (1977a). In former years, much effort was expended on testing the consequences of the Brans/Dicke scalar-tensor theory (Brans and Dicke, 1961) but research in solar physics has shown that even if the theory is correct then it does not differ significantly from Einstein’s general relativity in its consequences (Chapter 8; Wesson, 1978a). Scepticism about the presence of a scalar interaction in gravitation has led also to a feeling that it is unlikely that there could be a significant vector interaction present either, so while it may be mathematically viable a vector-tensor theory of gravitation (Hellings and Nordtvedt, 1973; Ni, 1972) is not favoured as a possible alternative to the (pure tensor) theory of general relativity.