We first investigate the form the General Relativity Theory would have taken had the gravitational mass and the inertial mass of material objects been different. We then extend this analysis to electromagnetism and postulate an equivalence principle for the electromagnetic field. We argue that to each particle with a different electric charge-to-mass ratio in superimposed gravitational and electromagnetic fields there corresponds a spacetime manifold whose metric tensor $g_{\mu\nu}$ describes the dynamical actions of gravitation and electromagnetism. The electric field outside a charged sphere asserts itself independently rather than contributing to the gravitational field. The contribution of the electric field to the spacetime metric outside the charged sphere is shown to be similar to the gravitational one in the Schwartzschild metric but with a charge-to-mass ratio dependence of the test particle instead of the Reissner - Nordstr\"om metric, resulting in a unified description of gravitation and electromagnetism. We point out that there are existing experiments whose results can be explained by the equivalence principle for the electromagnetic field presented here. Additional experimental predictions of the theory are mentioned.

We first review the cosmological constant problem, and then mention a conjecture of Feynman according to which the general relativistic theory of gravity should be reformulated in such a way that energy does not couple to gravity. We point out that our recent unification of gravitation and electromagnetism through a symmetric tensor has the property that the free electromagnetic energy and the vacuum energy do not contribute explicitly to the curvature of spacetime just like the free gravitational energy. Therefore in this formulation of general relativity, the vacuum energy density has its very large value today as in the early universe, while the cosmological constant does not exist at all.

We point out in this work that if our recently proposed unified description of gravitation and electromagnetism through a symmetric metric tensor is true, then building in the laboratory black holes for electrons with radii r_E\ge 0.5m in air and with much smaller radii in a vacuum should be possible.

It is shown that in the cosmological models based on a vacuum energy decaying as a^{-2}, where a is the scale factor of the universe, the fate of the universe in regard to whether it will collapse in future or expand forever is determined not by the curvature constant k but by an effective curvature constant k_{eff}. It is argued that a closed universe with k=1 may expand forever, in other words simulate the expansion dynamics of a flat or an open universe because of the possibility that k_{eff}=0 or -1, respectively. Two such models, in one of which the vacuum does not interact with matter and in another of which it does, are studied. It is shown that the vacuum equation of state p_{vac}= -\rho_{vac} may be realized in a decaying vacuum cosmology provided the vacuum interacts wuth matter. The optical depths for gravitational lensing as a function of the matter density and other parameters in the models are calculated at a source redshift of 2. The age of the universe is discussed and shown to be compatible with the new Hipparcos lower limit of 11Gyr. The possibility that a time-varying vacuum energy may serve as dark matter is suggested.

We have reanalyzed the age of the universe problem under the assumption that the lower limit on the age of the globular clusters is 11Gyr, as predicted by the recent Hipparcos data. We find that the globular cluster and the expansion ages in a \lambda=0 universe are consistent only if the present value H_0 of the Hubble constant is \le 60kms^{-1}Mpc^{-1}. If H_0\ge 60kms^{-1}Mpc^{-1} some kind of modification of the standard \lambda=0 model is required. Invoking a (time-independent) cosmological term \lambda in the Einstein field equations, as has been done frequently before, we have found that due to the gravitational lensing restrictions a fla universe with the present matter density parameter \Omega_M <0.5 is not problem free. A nonflat universe with \Omega_M\le 1 does not suffer from the age problem if H_0\le 75kms^{-1}Mpc^{-1}.

If gravitation and electromagnetism are both described in terms of a symmetric metric tensor, then the deflection of an electron beam by a charged sphere should be different from its deflection according to the Reissner-Nordstr\"om solution of General Relativity. If such a unified description is true, the equivalence principle for the electric field implies that the photon has a nonzero effective electric charge-to-mass ratio and should be redshifted in an electric field and be deflected in a magnetic field. Experiments to test these predictions are proposed.

We consider the low-energy effective string action in four dimensions including the leading order-$\alpha'$ terms. An exact homogeneous solution is obtained. It represents a non-singular expanding cosmological model in which the tensor fields tend to vanish as $t\to \infty$. The scale factor $a(t)$ of the very early universe in this model has the time dependence $a(t)^2=a_0^2+t^2$. The violation of the strong energy condition of classical General Relativity to avoid the initial singularity requires that the central charge deficit of the theory be larger than a certain value. The significance of this solution is discussed.

We discuss the notion that quantum fields may induce an effective time-dependent cosmological constant which decays from a large initial value. It is shown that such cosmological models are viable in a non-de Sitter spacetime.

We attempt to treat the very early Universe according to quantum mechanics. Identifying the scale factor of the Universe with the width of the wave packet associated with it, we show that there cannot be an initial singularity and that the Universe expands. Invoking the correspondence principle, we obtain the scale factor of the Universe and demonstrate that the causality problem of the standard model is solved.