Void models provide a possible explanation of the "accelerated expansion" of the Universe without dark energy. To make the conventional void models more realistic, we allow the void, an underdense region around us, to be anisotropic and consider an average of the distance-redshift relations over the solid angle subtended at the observer. We first show that after taking the average of a form of the optical scalar equation (distance equation), the effective distance equation we obtain coincides with the one for the Lemaitre-Tolman-Bondi universe with a Dyer-Roeder-like extension. We then numerically solve the equation to compare with observational data of Type Ia supernovae. We find that anisotropy allows smaller size of void and larger Omega_m.

Motivated by the inverse problem for the Lemaitre-Tolman-Bondi dust solution, in which problem the luminosity distance function $D_L(z)$ is taken as an input to select a specific model, we compute the function $D_L(z)$ of the LTB solution up to the third order of $z$. To perform the otherwise cumbersome computation, we introduce a new convenient form of the LTB solution, in which the solution is explicit and unified. With this form of the LTB solution we obtain the luminosity distance function with the full generality. We in particular find that the function exactly coincides with that of a homogeneous and isotropic dust solution up to second order, if we demand that the solution be regular at the center.

We discuss about the speed of sound squared in the Mass Varying Neutrinos scenario (MaVaNs). Recently, it was argued that the MaVaNs has a catastrophic instability which is the emergence of an imaginary speed of sound at the non-relativistic limit of neutrinos. As the result of this instability, the neutrino-acceleron fluid cannot act as the dark energy. However, it is found that the speed of sound squared in the neutrino-acceleron fluid could be positive in our model. We examine the speed of sound in two cases of the scalar potential. One is the small fractional power-law potential and another is the logarithmic one. The power-law potential model with the right-handed neutrinos gives a stable one.

We present a model of an inhomogeneous universe that leads to accelerated expansion after taking spatial averaging. The model universe is the Tolman-Bondi solution of the Einstein equation and contains both a region with positive spatial curvature and a region with negative spatial curvature. We find that after the region with positive spatial curvature begins to re-collapse, the deceleration parameter of the spatially averaged universe becomes negative and the averaged universe starts accelerated expansion. We also discuss the generality of the condition for accelerated expansion of the spatially averaged universe.

We have measured thermoelectric power (TEP) as a function of hole concentration per CuO2 layer, Ppl, in Y1-xCaxBa2Cu3O6 (Ppl = x/2) with no oxygen in the Cu-O chain layer. The room-temperature TEP as a function of Ppl, S290(Ppl), of Y1-xCaxBa2Cu3O6 behaves identically to that of La2-zSrzCuO4 (Ppl = z). We argue that S290(Ppl) represents a measure of the intrinsic equilibrium electronic states of doped holes and, therefore, can be used as a common scale for the carrier concentrations of layered cuprates. We shows that the Ppl determined by this new universal scale is consistent with both hole concentration microscopically determined by NQR and the hole concentration macroscopically determined by the Cu valency. We find two characteristic scaling temperatures, TS* and TS2*, in the TEP vs. temperature curves that change systematically with doping. Based on the universal scale, we uncover a universal phase diagram in which almost all the experimentally determined pseudogap temperatures as a function of Ppl fall on two common curves; upper pseudogap temperature defined by the TS* versus Ppl curve and lower pseudogap temperature defined by the TS2* versus Ppl curve. We find that while pseudogaps are intrinsic properties of doped holes of a single CuO2 layer for all high-Tc cuprates, Tc depends on the number of layers, therefore the inter-layer coupling, in each individual system.