G.R. Knapp reviewed observations of the structure of circumstellar dust shells around AGB stars. The photospheres of several red giant stars have recently been found to be non-circularly symmetric, and the asymmetries are aligned with large-scale asymmetries in the circumstellar envelopes. These asymmetries, and those in the subsequent evolutionary stage, planetary nebulae, may be the result of asymmetries in the shapes of the stars themselves.
Knapp and J.E. Gunn, in collaboration with P.F. Bowers (U.S. Naval
Observatory) and J. Vasquez Poritz
(Cornell undergraduate) analyzed sensitive short-wavelength VLA
observations of six globular
clusters. Several new candidate low-mass x-ray binary stars were
identified by detection of radio
point source emission at the positions of cluster x-ray sources. No new
planetary nebula candidates
were found. The observations also set limits of typically 
on diffuse
ionized gas in the centers of the clusters.
Knapp and A. Jorissen (Brussels), with K. Young (Caltech) detected a very fast (
kms
) molecular wind from the carbon star V Hya using the CSO.
Comparison with other objects
showing this phenomenon shows that such winds may signal the onset of
planetary nebula formation.
W.N. Colley (graduate student) and J.R. Gott III studied the naked-eye
observability of
microlensing events for both known stars and possible massive compact halo
objects (MACHOs). They
found that if both the dark matter disk and halo are composed of MACHOs in
the Jupiter-mass range,
microlensing events of naked-eye stars, undergoing at least 1 magnitude of
magnification, occur at the
rate of 1 per 2400 yr., and have durations of from hours to days; they thus
surmise that the change of
at least one event occurring in the era of recorded history (the last 5000
yrs.) would be 
%. For magnification by known stars, events are expected
at the rate of 1 per 40,000
yr., so we should not be surprised not to have witnessed an event in the
last 5000 yr. However, in
the last 200,000 yr., while humans have inhabited the Earth, we expect of
order five events, and thus
a 99% chance that at least one observable event has occurred.
J.R. Gott has continued to investigate the implications of the Copernican Principle for our future prospects. If our location is not special among intelligent observers, the total longevity of our species is predicted to be similar in order of magnitude to that observed for other mammal and hominid species. This study points out how Copernican estimates are identical to those obtained with an appropriate Vague Bayesian Prior. This is no accident -- if the Vague Bayesian Prior is chosen correctly then all intelligent observers should be able to use it and then the Copernican results should be obtained.