The vector-tensor (VT) theory of gravitation revisited in this article was studied in previous papers, where it was proved that VT works and deserves attention. New observational data and numerical codes have motivated further development which is presented here. New research has been planed with the essential aim of proving that current cosmological observations, including Planck data, baryon acoustic oscillations (BAO), and so on, may be explained with VT, a theory which accounts for a kind of dark energy which has the same equation of state as vacuum. New versions of the codes CAMB and COSMOMC have been designed for applications to VT, and the resulting versions have been used to get the cosmological parameters of the VT model at suitable confidence levels. The parameters to be estimated are the same as ingeneral relativity (GR), plus a new parameter $D$. For $D=0$, VT linear cosmological perturbations reduces to those of GR, but the VT background may explain dark energy. The fits between observations and VT predictions lead to non vanishing $|D|$ upper limits at the $1\sigma $ confidence level. The value $D=0$ is admissible at this level, but this value is not that of the best fit in any case. Results strongly suggest that VT may explain current observations, at least, as well as GR; with the advantage that, as it is proved in this paper, VT has an additional parameter which facilitates adjustments to current observational data.}

We study relevant cosmological topics in the framework of a certain vector-tensor theory of gravitation (hereafter VT). This theory is first compared with the so-called extended electromagnetism (EE). These theories have a notable resemblance and both explain the existence of a cosmological constant. It is shown that, in EE, a positive dark energy density requires a Lagrangian leading to quantum ghosts, whereas VT is free from these ghosts. On account of this fact, the remainder of the paper is devoted to study cosmology in the framework of VT. Initial conditions, at high redshift, are used to solve the evolution equations of all the VT scalar modes. In particular, a certain scalar mode characteristic of VT -which does not appear in general relativity (GR)- is chosen in such a way that it evolves separately. In other words, the scalar modes of the standard model based on GR do not affect the evolution of the VT characteristic mode; however, this scalar mode influences the evolution of the standard GR ones. Some well known suitable codes (CMBFAST and COSMOMC) have been modified to include our VT initial conditions and evolution equations, which are fully general. One of the resulting codes -based on standard statistical methods- has been used to fit VT predictions and observational evidences about both Ia supernovae and cosmic microwave background anisotropy. Seven free parameters are used in this fit. Six of them are often used in GR cosmology and the seventh one is characteristic of VT. From the statistical analysis it follows that VT seems to be advantageous against GR in order to explain cosmological observational evidences.

A certain vector-tensor (VT) theory is revisited. It was proposed and analyzed as a theory of electromagnetism without the standard gauge invariance. Our attention is first focused on a detailed variational formulation of the theory, which leads to both a modified Lorentz force and the true energy momentum tensor of the vector field. The theory is then applied to cosmology. A complete gauge invariant treatment of the scalar perturbations is presented. For appropriate gauge invariant variables describing the scalar modes of the vector field (A-modes), it is proved that the evolution equations of these modes do not involve the scalar modes appearing in General Relativity (GR-modes), which are associated to the metric and the energy momentum tensor of the cosmological fluids. However, the A-modes modify the standard gauge invariant equations describing the GR-modes. By using the new formalism, the evolution equations of the A-perturbations are derived and separately solved and, then, the correction terms --due to the A-perturbations-- appearing in the evolution equations of the GR-modes are estimated. The evolution of these correction terms is studied for an appropriate scale. The relevance of these terms depends on both the spectra and the values of the normalization constants involved in extended electromagnetism. Further applications of the new formalism will be presented elsewhere.

A certain vector-tensor theory is revisited. Our attention is focused on cosmology. Against previous suggestions based on preliminary studies, it is shown that, if the energy density of the vector field is large enough to play the role of the dark energy and its fluctuations are negligible, the theory is not simultaneously compatible with current observations on: supernovae, the cosmic microwave background (CMB) anisotropy, and the power spectrum of the energy density fluctuations. However, for small enough energy densities of the vector field, the theory becomes compatible with all the above observations and, moreover, it leads to an interesting evolution of the so-called vector cosmological modes. This evolution appears to be different from that of general relativity, and the difference might be useful to explain the anomalies in the low order CMB multipoles.

We work in the framework of a simple vector-tensor theory. The parametrized post-Newtonian approximation of this theory is identical to that of general relativity. Our attention is focused on cosmology. In an homogeneous isotropic universe, it is proved that the energy density, $\rho_{A}$, of the vector field $A$, and its pressure, $p_{A}$, do not depend on time, and also that the equation of state is $\rho_{A} = -p_{A}$. This means that, in the theory under consideration, there is a cosmological constant, which is not vacuum energy, but the dark energy of the cosmic vector field $A$, whose evolution is classical.