A systematic account of a new dualistic approach to gravitation developed over the years 1979-83 by the author is presented. The main idea of this article is to fill up the gaps in the exposition and'physical interpretation rather than any important extension. We propose here a dual theory for gravitation. I n the exact form it is supposed to be appropriate for microscopic (subnuclmr) phenomena and in an averaged out form (may be more than once averaged) appropriate for a description of the usual gravitation of macroscopic bod&. A pamllel can be found in the case of the Lorentz and Maxwell theories of electromagnetic phenomena. When sources are present the macroscopic theory is the same as the conventional Einstein theory but in the absence of sources space-time is flat ruling out the existence of gravitational waves in tbis approach. Experimental verification of this should be a test case for the validity of the theory.
Ann. Physik Leipzig 41 (1984) 6
for a geometrical behaviour of bodies large compared with molecules. Only experimental success may justify such an attempt to ascribe physical reality to the basic ideas of Riemannian geometry outside regions of their proven validity. However, it may also so happen, that an extrapolation of this nature does not have any more basis than an extension of the idea of temperature to parts of a body of molecular dimensions".
A. EINSTEIN, Geometrie und Erfahrung; Sitzungsber preuss. Akad. Wiss. 1921, T. 1, p. 123.