Using the effective equations in our Loop Quantum Cosmology (LQC) -- inspired models, we resolve the $(n + 2)$ dimensional black hole singularity. This resolution in the four dimensional case is same as that given recently by Ashtekar, Olmedo, and Singh. We then study the fate of a massive collapsing star in $(n + 2)$ dimensions, focussing on a singularity resolved black hole, a string theoretic fuzz ball, and a Loop quantum star -- which is a solution to the the effective modified equations for static spherically symmetric stars we obtained recently using LQC ideas. We find qualitatively that a massive collapsing star is likely to become a Loop quantum star : it is likely to have a macroscopic core region with Planckian densities and pressures which is surrounded by a low density corona region extending to the Schwarzschild radius.

We rewrite the Tolman -- Oppenheimer -- Volkoff (TOV) equations for four and higher dimensional static spherically symmetric stars so that they resemble the equations for anisotropic cosmology. This becomes possible by treating the curvature terms of the sphere as part of the matter sector. Comparing then with the effective equations in the Loop Quantum Cosmology -- inspired models, we propose analogous modifications to the TOV equations which contain one arbitrary function. A linear function gives the TOV equations. If the function and all its derivatives are finite then the solutions are non singular. For $sin \; x \;$, an `image star' appears beyond the central region of maximal pressures in the original star. This is evocative of a bouncing universe in Loop Quantum Cosmology. We discuss possible observational consequences of these features.

The dynamics of a (3 + 1) dimensional homogeneous anisotropic universe is modified by Loop Quantum Cosmology and, consequently, it has generically a big bounce in the past instead of a big-bang singularity. This modified dynamics can be well described by effective equations of motion. We generalise these effective equations of motion empirically to (d + 1) dimensions. The generalised equations involve two functions and may be considered as a class of LQC -- inspired models for (d + 1) dimensional early universe cosmology. As a special case, one can now obtain a universe which has neither a big bang singularity nor a big bounce but approaches asymptotically a `Hagedorn like' phase in the past where its density and volume remain constant. In a few special cases, we also obtain explicit solutions.

Unitarity of evolution in gravitational collapses implies existence of macroscopic stable horizonless objects. With such objects in mind, we study the effects of anisotropy of pressures on the stability of stars. We consider stars in four or higher dimensions and also stars in M theory made up of (intersecting) branes. Taking the stars to be static, spherically symmetric and the equations of state to be linear, we study `singular solutions' and the asymptotic perturbations around them. Oscillatory perturbations are likely to imply instability. We find that non oscillatory perturbations, which may imply stability, are possible if an appropriate amount of anisotropy is present. This result suggests that it may be possible to have stable horizonless objects in four or any higher dimensions, and that anisotropic pressures may play a crucial role in ensuring their stability.

A massive compact object is that which forms when a sufficiently massive star collapses. This is commonly taken to be a black hole with a singularity surrounded by a horizon and which evolves by emitting Hawking radiation. In a quantum theory of gravity, singularities are expected to be resolved and the evolutions are expected to be unitary. Assuming that such a theory with these properties exists, and with a few more physically motivated assumptions, we argue that a massive compact object has no singularity (by assumption) and must also have no horizon; otherwise, there may be a loss of predictability in the case of a black hole candidate observed today. With no singularity and also with no horizon, the massive compact object will then evolve as a standard quantum system with large number of interacting degrees of freedom.

We study stars in M theory. First, we obtain the analog of Oppenheimer -- Volkoff equations in a suitably general set up. We obtain analytically the asymptotic solutions to these equations when the equations of state are linear. We study perturbations around such solutions in several examples and, following a standard method, use their behaviour to determine whether an instability is present or not. In this way, we obtain a generalisation of the corresponding results of Chavanis. We also find that stars in M theory have instabilities. Therefore, if sufficiently massive, such stars will collapse. We discuss the significance of these (in)stabilities within the context of Mathur's fuzz ball proposal.

We describe static, brane--like, solutions to vacuum Einstein's equations in D = n + m + 2 dimensional spacetime with m \ge 2 and n \ge 1. These solutions have positive ADM mass but no horizon. The curvature invariants are finite everywhere except at r = 0 where r is the radial coordinate in the m + 1 dimensional space. The presence of n \ge 1 extra dimensions is crucial for these properties. Such solutions may be naturally anticipated if Mathur's fuzzball proposal for black holes is correct.

We assume that the early universe is homogeneous, anisotropic, and is dominated by the mutually BPS 22'55' intersecting branes of M theory. The spatial directions are all taken to be toroidal. Using analytical and numerical methods, we study the evolution of such an universe. We find that, asymptotically, three spatial directions expand to infinity and the remaining spatial directions reach stabilised values. Any stabilised values can be obtained by a fine tuning of initial brane densities. We give a physical description of the stabilisation mechanism. Also, from the perspective of four dimensional spacetime, the effective four dimensional Newton's constant G_4 is now time varying. Its time dependence will follow from explicit solutions. We find in the present case that, asymptotically, G_4 exhibits characteristic log periodic oscillations.

Using super Hamiltonian formalism, we study the motion of particles whose dispersion relations are modified to incorporate Ho\v{r}ava -- Lifshitz type anisotropic scaling symmetry. We find the following as consequences of this modified dispersion relation: (i) The speed of a charged particle under a constant electric field grows without bound and diverges. (ii) The speed of a particle falling towards the horizon also grows without bound and diverges as the particle approaches the horizon. (iii) This particle reaches the horizon in a finite coordinate time, in contrast to the standard case where it requires infinite time.

We study anisotropic cosmology in Ho\v{r}ava's gravity theory and obtain Kasner type solutions, valid for any number d of spatial dimensions. The corresponding exponents satisfy two relations, one involving the marginal coupling \lambda. Also, Ho\v{r}ava's (super)renormalisable theory predicts (super)stiff matter whose equation of state is p = w \rho with w \ge 1. We discuss briefly the implications of these results for the nature of cosmological collapse.

We consider N -- charge, intersecting brane antibrane configurations in M theory which are smeared uniformly in the common transverse space and may describe our universe. We study the consequences of U dualities and find that they imply relations among the scale factors. We find using Einstein's equations that U dualities also imply a relation among the density \rho and the pressure p_i for the single charge case. We present an ansatz for \rho and p_i for the N -- charge case which yields all the U duality relations among the scale factors. We then study configurations with identical charges, and also with net charges vanishing. We find among other things that, independent of the details of the brane antibrane dynamics, such four charge configurations lead asymptotically to an effective (3 + 1) -- dimensional expanding universe.

We obtain the entropy of a homogeneous anisotropic universe applicable, by assumption, to the fractional branes in the universe in the model of Chowdhury and Mathur. The entropy for the 3 or 4 charge fractional branes thus obtained is not of the expected form E^{{3/2}} or E^2. One way the expected form is realised is if p \to \rho for the transverse directions and if the compact directions remain constant in size. These conditions are likely to be enforced by brane decay and annihilation, and by the S, T, U dualities. T duality is also likely to exclude high entropic cases, found in the examples, which arise due to the compact space contracting to zero size. Then the 4 charge fractional branes may indeed provide a detailed realisation of the maximum entropic principle we proposed recently to determine the number (3 + 1) of large spacetime dimensions.

We study a d-dimensional FRW universe, containing a perfect fluid with p = w \rho and \frac{1} {d - 1} \le w \le 1, and find a correspondence principle similar to that of Horowitz and Polchinski in the black hole case. This principle follows quite generally from thermodynamics and the conservation of energy momentum tensor, and can be stated along similar lines as in the black hole case: ``When the temperature T of the universe becomes of order string scale the universe state becomes a highly excited string state. At the transition, the entropies and energies of the universe and strings differ by factors of {\cal O}(1).'' Such a matching is absent for w \ne 1 if the transition is assumed to be when the curvature or the horizon length is of order string scale.

Long ago, McVittie had found a class of solutions which can be thought of as Schwarzschild black holes in an FRW universe. In recent years they have been studied extensively and generalised to charged and uncharged black holes in D \ge 4 dimensions also. Here, assuming an ansatz similar to McVittie's, we present solutions for uncharged branes which can be thought of as branes in a time dependent universe. We consider their application to the brane antibrane decay process, also referred to as tachyon condensation, and discuss the necessary generalisations required for our ansatz to describe such a process.

We consider scalar tensor theories in D-dimensional spacetime, D \ge 4. They consist of metric and a non minimally coupled scalar field, with its non minimal coupling characterised by a function. The probes couple minimally to the metric only. We obtain vacuum solutions - both cosmological and static spherically symmetric ones - and study their properties. We find that, as seen by the probes, there is no singularity in the cosmological solutions for a class of functions which obey certain constraints. It turns out that for the same class of functions, there are static spherically symmetric solutions which exhibit novel properties: {\em e.g.} near the ``horizon'', the gravitational force as seen by the probe becomes repulsive.

We study the dynamical features of Maggiore's generalised commutation relations. We focus on their generality and, in particular, their dependence on the Hamiltonian H. We derive the generalisation of the Planck's law for black body spectrum, study the statistical mechanics of free particles, and study the early universe evolution which now exhibits non trivial features. We find that the dynamical features, found here and in our earlier work, are all generic and vary systematically with respect to the asymptotic growth of the Hamiltonian H.

We study the dynamical consequences of Maggiore's unique generalised uncertainty principle (GUP). We find that it leads naturally, and generically, to novel consequences. In the high temperature limit, there is a drastic reduction in the degrees of freedom, of the type found, for example, in strings far above the Hagedorn temperature. In view of this, the present GUP may perhaps be taken as the new version of the Heisenberg uncertainty principle, conjectured by Atick and Witten to be responsible for such reduction. Also, the present GUP leads naturally to varying speed of light and modified dispersion relations. They are likely to have novel implications for cosmology and black hole physics, a few of which we discuss qualitatively.

We present a simple derivation of an upper bound on the average size of the true vacuum bubbles at the end of inflation, in models of extended inflation type. The derivation uses the inequality that the total energy inside a given volume must be less than its linear dimensions. The above bound is the same as that obtained earlier, by applying the holographic principle according to Fischler-Susskind prescription. Such a bound leads to a lower bound on the denisty fluctuations.

A closed universe containing pressureless dust, more generally perfect fluid matter with pressure-to-density ratio w in the range (1/3, - 1/3), violates holographic principle applied according to the Fischler-Susskind proposal. We show, first for a class of two-fluid solutions and then for the general multifluid case, that the closed universe will obey the holographic principle if it also contains matter with w < - 1/3, and if the present value of its total density is sufficiently close to the critical density. It is possible that such matter can be realised by some form of `quintessence', much studied recently.

We apply the holographic principle during the inflationary stage of our universe. Where necessary, we illustrate the analysis in the case of new and extended inflation which, together, typify generic models of inflation. We find that in the models of extended inflation type, and perhaps of new inflation type also, the holographic principle leads to a lower bound on the density fluctuations.

We study homogeneous isotropic universe in a graviton-dilaton theory obtained, in a previous paper, by a simple requirement that the theory be able to predict non trivial values for $\beta$ and/or $\gamma$ for a charge neutral point star, without any naked singularities. We find that in this universe the physical time can be continued indefinitely into the past or future, and that all the physical curvature invariants are always finite, showing the absence of big bang singularity. Adding a dilaton potential, we find again the same features. As a surprising bonus, there emerges naturally a Brans-Dicke function, which has precisely the kind of behaviour needed to make $\omega_{bd} ({\rm today}) > 500$ in hyperextended inflation.

We look for a graviton-dilaton theory which can predict non trivial values of the PPN parameters $\beta$ and/or $\gamma$ for a charge neutral point star, without any naked singularity. With the potential for dilaton $\phi$ set to zero, it contains one arbitrary function $\psi(\phi)$. Our requirements impose certain constraints on $\psi$, which lead to the following generic and model independent novel results: For a charge neutral point star, the gravitational force becomes repulsive at distances of the order of, but greater than, the Schwarzschild radius $r_0$. There is also no horizon for $r > r_0$. These results suggest that black holes are unlikely to form in a stellar collapse in this theory.

Analysing the static, spherically symmetric graviton-dilaton solutions in low energy string and Brans-Dicke theory, we find the following. For a charge neutral point star, these theories cannot predict non trivial PPN parameters, $\beta$ and $\gamma$, without introducing naked singularities. We then couple a cosmological constant $\Lambda$ as in low energy string theory. We find that only in low energy string theory, a non zero $\Lambda$ leads to a curvature singularity, which is much worse than a naked singularity. Requiring its absence upto a distance $r_*$ implies a bound $| \Lambda | < 10^{- 102} (\frac{r_*}{{\rm pc}})^{- 2}$ in natural units. If $r_* \simeq 1 {\rm Mpc}$ then $| \Lambda | < 10^{- 114}$ and, if $r_* \simeq 10^{28} {\rm cm}$ then $| \Lambda | < 10^{- 122}$ in natural units.

We present the static, spherically symmetric solutions of the low energy $d = 4$ string. They form a simple generalisation of the Schwarzschild solution and describe the gravitational field of a spherical star in perturbative string theory. If the Robertson parameter $\gamma$ is different from one, these solutions describe singular objects with naked singularities at their horizons, and whose Hawking temperature is infinite. These objects might represent ``massive remnants'', the end states of black holes. Our analysis also implies that Brans-Dicke theory has a naked singularity. Invoking cosmic censorship conjecture will rule out Brans-Dicke theory.