The Supernova Rate in Starburst Galaxies
                        Michael William Richmond

                              ABSTRACT

I conducted an optical search for supernovae in a sample of 142 
starburst galaxies over the period December, 1988 to June, 1991.
The sample was drawn from two sets of galaxies: a subset of the
IRAS Bright Galaxy Sample (Soifer et al. 1987), selected on the
basis of their high far-infrared luminosities, and galaxies 
exhibiting strong emission lines of hydrogen in their nuclei
(Balzano 1983).  All the galaxies were nearby (z <= 0.03) and
relatively bright (m(B) <= 16).  I used the one-meter Nickel
Telescope at Lick Observatory to obtain unfiltered or R-band
CCD images with a limiting magnitude of m(R) = 18 of as many
galaxies as possible on a single night at two-week intervals.
Over the course of the survey, visual comparison of new and 
reference images revealed a total of five supernovae in the
sample, each of which occurred outside its host galaxy's nucleus.
Using a set of template curves for each type of supernova, in both
V and R bandpasses, which I constructed from the literature, I
determine supernova rates in the extra-nuclear regions to be
0.7 h^2 SNU for type Ia, 0.7 h^2 SNU for type Ib/c, and ~0.6 h^2 SNU
for type II, with large uncertainties but upper limits of 2.2 h^2,
2.5 h^2, and 1.7 h^2 SNU, respectively.  These rates are similar
to those measured in "normal" galaxies.  Because supernovae in the
nuclei of galaxies are difficult to detect by comparing two images
by eye, I performed aperture photometry on the nucleus of each image 
and searched the resulting light curves for increases in brightness
caused by supernovae.  I found no evidence for a supernova-induced
brightening in any nucleus, and, with a few reasonable assumptions,
can place upper limits of 9 h^2, 12 h^2, and 8 h^2 SNU on the rates
of type Ia, Ib/c, and II supernovae inside the nuclei.  I derive
relationships between stellar population, supernova rate and flux
in Balmer emission lines for a numnber of population models; 
for 34 galaxies in the sample with measured nuclear H-alpha and 
H-beta fluxes, I apply the relationships to place constraints on their 
stellar populations.  In an appendix, I describe my contributions 
to the Berkeley Automatic Imaging Telescope.