A new search for the gravitational lens effects of a significant cosmological density of supermassive compact objects (SCOs) on gamma-ray bursts has yielded a null result. We inspected the timing data of 774 BATSE-triggered GRBs for evidence of millilensing: repeated peaks similar in light-curve shape and spectra. Our null detection leads us to conclude that, in all candidate universes simulated, $\Omega_{SCO} < 0.1$ is favored for $10^5 < M_{SCO}/M_{\odot} < 10^9$, while in some universes and mass ranges the density limits are as much as 10 times lower. Therefore, a cosmologically significant population of SCOs near globular cluster mass neither came out of the primordial universe, nor condensed at recombination.

The fast temporal structures and cosmological distances of gamma-ray bursts (GRBs) afford a natural laboratory for testing theories of frequency-dependent propagation of high-energy photons, as predicted for quantum gravity (QG). We calibrate the sensitivity of the proposed Gamma-ray Large Area Space Telescope (GLAST) by performing simulations which include: the response of GLAST to a GRB fluence distribution; a distribution of spectral power-law indices similar to the EGRET sample; and consideration of gamma-gamma attenuation, significant above ~ 10 GeV for redshifts z > 3 - 5. We find that GLAST should detect > 200 GRBs per year, with sensitivity to a few tens of GeV for a few bursts. GLAST could detect the energy- and distance-dependent dispersion (10 ms / GeV / Gpc) predicted by QG with 1 - 2 years of observations. Attribution to QG would require correlation of GRB redshifts with the temporal and energetic signatures.

We suggest a connection between the pulse paradigm at gamma-ray energies and the recently demonstrated luminosity distribution in gamma-ray bursts: the spectral evolution timescale of pulse structures is anticorrelated with peak luminosity, and with quantities which might be expected to reflect the bulk relativistic Lorentz factor, such as spectral hardness ratio. We establish this relationship in two important burst samples using the cross-correlation lags between low (25-50 keV) and high (100-300 keV and > 300 keV) energy bands. For a set of seven bursts (six with redshifts) observed by CGRO/BATSE and BeppoSAX which also have optical or radio counterparts, the gamma/X peak flux ratios and peak luminosities are anti-correlated with spectral lag. For the 174 brightest BATSE bursts with durations longer than 2 s and significant emission above 300 keV, a similar anti-correlation is evident between gamma-ray hardness ratio or peak flux, and spectral lag. For the six bursts with redshifts, the connection between peak luminosity and spectral lag is well-fitted by a power-law. GRB 980425 (if associated with SN 1998bw) would appear to extend this trend qualitatively, but with a lag of ~ 4-5 s and luminosity of ~ 1.3x10^47 ergs s^-1, it falls below the power-law relationship by a factor of ~ 460. As noted previously by Band, most lags are concentrated on the short end of the lag distribution, near 100 ms, suggesting that the GRB luminosity distribution is peaked on its high end, e.g. N(L) proportional to L^beta, with positive beta.

If dark matter in the form of compact objects comprises a large fraction of the mass of the universe, then gravitational lensing effects on gamma-ray bursts are expected. We utilize BATSE and Ulysses data to search for lenses of different mass ranges, which cause lensing in the milli, pico, and femto regimes. Null results are used to set weak limits on the cosmological abundance of compact objects in mass ranges from 10$^{-16}$ to 10$^{-9}$ $M_{\odot} $. A stronger limit is found for a much discussed $\Omega = 0.15$ universe dominated by black holes of masses $\sim 10^{6.5} M_{\odot}$, which is ruled out at the $\sim$ 90% confidence level.

GRBs vary more rapidly than any other known cosmological phenomena. The lower limits of this variability have not yet been explored. Improvements in detectors would reveal or limit the actual rate of short GRBs. Were microsecond "spike" GRBs to exist and be detectable, they would time-resolve stellar mass objects throughout the universe by their gravitational microlensing effect. Analyzing the time structure of sufficient numbers of GRB spikes would reveal or limit $\Omega_{star}$, $\Omega_{MACHO}$, and/or $\Omega_{baryon}$.

Recent claims of angular correlations between gamma-ray bursts (GRBs) and clusters of galaxies are evaluated in light of existing but previously uncorrelated GRB positional data. Additional GRB data sets we use include sub-samples of soft BATSE 3B bursts, bursts located by the Interplanetary Network (IPN), and GRBs localized by COMPTEL. We confirm a previously reported excess by Rood and Struble (1996) of the 185 rich, nearby clusters of galaxies (Abell, Corwin, and Olowin 1989, ACO) in the 1-$\sigma$ error circles of 74 BATSE 3B positions, but find a typical correlation strength of only $~$2.5-$\sigma$ for typical sub-samples. However, none of the 185 ACO clusters occur in the 1-$\sigma$ error boxes of 40 IPN GRBs or 18 COMPTEL GRBs. When all ACO clusters are correlated with BATSE 3B GRBs however, we find an increasingly strong correlation for GRBs with decreasingly small error boxes, reaching above the 3.5-$\sigma$ level. We also find a slight excess of {\it soft} BATSE GRBs near the positions of 185 rich, nearby ACO clusters, but the significance of the correlation averages only $~$2.5-$\sigma$ for sub-samples delineated by softness. We caution that the statistical significance of all these correlations is marginal, and so conclude that the excess is at best only suggestive of a physical association. Statistical fluke is still a strong possibility. BATSE could confirm or refute such correlations in a 10-year lifetime.