In light of the discovery of the first-ever double pulsar system, PSR J0737-3039, we re-examine an earlier proposal to directly detect gravity waves from neutron stars, which was predicated on a hypothetical system almost identical to the later discovered double pulsar. We re-derive the effect in more detail, and confirm the initial estimate--sometimes doubted in the literature--that it includes a 1/b dependence, where b is the impact parameter of a pulsar with respect to its foreground, gravity-wave emitting, neutron star companion. A coherent modulation in pulsar time-of-arrival measurements of 10 nano-sec/sec is possible. A one-year intermittent experiment on an instrument comparable to the SKA could thus detect the exceedingly faint gravity waves from individual neutron stars.

Gravitation could modulate the interstellar scintillation of pulsars in a way that is analogous to refractive interstellar scintillation (RISS). While RISS occurs when a large ionized cloud crosses the pulsar line-of-sight, gravitational interstellar scintillation (GISS) occurs when a compact gravitational deflector lies very near to that line-of-sight. However, GISS differs from RISS in at least two important respects: It has a very distinctive and highly predictible time signature, and it is non-dispersive. We find two very different astronomical contexts where GISS could cause observable diffraction-pattern distortions: Highly inclined binary pulsars, and the kind of compact interstellar clouds suspected of causing extreme scattering events.

We find that general relativity can be naturally free of cosmological singularities. Several nonsingular models are currently available that either assume ad hoc matter contents, or are nonsingular only over a sector of solution space of zero measure, or depart drastically from general relativity at high energies. After much uncertainty over whether cosmological inflation could help solve the initial-singularity problem, the prevailing belief today is that general relativistic cosmology, with inflation or without, is endemically singular. This belief was reinforced by recent singularity theorems that take account specifically of inflation. Here, a viable inflationary cosmology is worked out that is naturally free of singularities despite the fact that 1) it uses only classical general relativity, 2) it assumes only the most generic inflationary matter contents, and 3) it is a theory of the chaotic-inflation type. That type of inflation is the most widely accepted today, as it demands the least fine-tuning of initial conditions. It is also shown how, by dropping the usual simplification of minimal coupling between matter and geometry, the null energy condition can be violated and the relevant singularity theorems circumvented.

Gravitational waves, although generally associated with extremely microscopic effects, can displace by hundreds of kilometers the pulsar interstellar scintillation patterns that bathe the Earth. The combination of the pulsar and the interstellar medium acts as a kiloparsec-long, nature-provided gravity wave amplifier. We show how an effective scheme for the detection of periodic gravity waves can be constructed based on this effect. This approach to detection does not require the development of new, ad hoc technology, but the optimization of existing observational techniques in a few different fields of astronomy. Part of the scheme is a new, purely numerical detection technique that can also be used in the data processing of other projects of periodic gravity wave detection.

It is generally accepted that a first ever direct detection of gravity waves would herald a new era in astronomy and in fundamental physics. Ever since the early sixties, increasingly larger human and material resources are being invested in the detection effort. Unfortunately, the gravity wave effects one has had to exploit so far are extraordinarily small and are usually very many orders of magnitude smaller than the noise involved. The detectors that are presently at the most advanced stage of development hope to register extremely rare, instantaneous longitudinal shifts that are expected to be orders of magnitude smaller than one Fermi. However, it was recently shown that gravity waves can manifest themselves through much larger effects than previously envisaged. One of these new effects is the periodic, apparent shift in a star's angular position due to a foreground gravity wave source. The comparative largeness of this effect stems from its being proportional not to the inverse of the gravity wave source's distance to the Earth, but to the inverse of its distance to the star's line of sight. In certain optimal but not unrealistic cases, the amplitude of this effect can reach the critical bar of one micro-arcsecond, thus raising the prospect that the long awaited first direct detection of gravity waves could be achieved by a high precision astrometry space mission such as GAIA.

An observer, situated several thousand light-years away from a radio pulsar, finds himself embedded in the diffraction pattern resulting from the propagation of the radio waves through the irregular interstellar medium. The observer's movement relative to the pattern causes an apparent scintillation of the pulsar. A binary star, situated close to the pulsar's line-of-sight, is generating relatively strong gravity waves. The rays originating from the pulsar experience a tiny periodic deflection due to the gravity waves produced by the binary star. This deflection displaces the diffraction pattern laterally in a manner that is familiar from refractive interstellar scintillation, except that this gravity wave effect is not dispersive. The displacement has the same period as the gravity waves. Its amplitude equals the product of the tiny deflection angle and the large distance from the binary star to the observer. This periodic displacement can reach a few hundred kilometers, which can be comparable to the size of the features in the diffractive pattern. Thus, there seems to be a possibility that the exceedingly faint gravity waves can manifest themselves macroscopically. Observationally, the end effect could be a substantial, deterministic alteration of the scintillation time structure.

Recently, the possibility has emerged of an early detection of astrophysical gravity waves. In certain astronomical configurations, and through a new light-deflection effect, gravity waves can cause apparent shifts in stellar angular positions as large as $10^{-7}arcsec$. In these same configurations, the magnitude of the gravity-wave-induced time-delay effect can exceed $10^{-14}$. Both these figures lie just at present-day theoretical limits of detectability. For instance, cases are described where the very faint neutron-star gravity waves could soon become detectable. The detection meant here involves direct observations of the very wave-forms.

When a source of gravity waves is conveniently placed between the Earth and some source of light, preferably a pulsating source, the magnitude of time delays induced by the gravity waves could, in optimal situations, be not too far out of the reach of already existing technology. Besides the odd case of near-to-perfect alignment one might be lucky enough to encounter in the Galaxy, there exists several astronomical sites where good alignment occurs naturally. A good example is when the light source and the gravity-wave source are, respectively, a high-frequency pulsar and a neutron star, locked together in a tight binary. We are lead to believe that neutron-star gravity waves might be directly observable in timing data of systems such as PSR B1913+16.

It is suggested that gravity waves could, in several cases, be detected by means of already (or shortly to be) available technology, independently of current efforts of detection. The present is a follow-up on a recently suggested detection strategy based on gravity-wave-induced deviations of null geodesics. The new development is that a way was found to probe the waves close to the source, where they are several orders of magnitude larger than on the Earth. The effect translates into apparent shifts in stellar angular positions that could be as high as $10^{-7}$ arcsec, which is just about the present theoretical limit of detectability. (Calculation improved; results unchanged.)