The radial behavior of B- and I-band surface brightness fluctuations i
n M32 has been measured. We identify a systematic variation in the sur
face brightness fluctuations of M32, which is seen as a local minimum
in the fluctuations at a distance of R similar to 20 '' from the cente
r. The systematic variation amounts to similar to 0.2 mag around the a
verage value (m) over bar(l) = 22.78 +/- 0.04, and similar to 0.3 mag
around (m) over bar(B) = 26.78 +/- 0.03. A possible population of asym
ptotic giant branch stars and associated post asymptotic giant branch
stars likely concentrated towards the center may cause the observed in
crease in the luminosity fluctuations for R < 20 ''. The increase in l
uminosity fluctuations outward may be attributed to a possible decreas
e in the mean metallicity of the underlying stellar population. A comp
arison of the present B-band luminosity fluctuations with stellar popu
lation models having different horizontal branch morphologies indicate
s that a red clump horizontal branch cannot account for the observed '
'blue'' fluctuations. During the reduction procedure the main source o
f contamination turned out to be a not insignificant contribution from
M31's disk population. Based on the present observations and on simpl
e simulations of the field we argue that the mere recovery of added st
ars does not necessarily guarantee that the derived stellar photometry
represents the true photometric properties of the stellar population.
Realistic simulations of the crowded field are essential in order to
select a criterion that will eliminate spurious sources due to the sur
face brightness fluctuations. Simulations show that our ground-based i
mages with similar to 1.'' 0 seeing do not have the spatial resolution
that will allow us to supplement the fluctuation measurement with ste
llar photometry of the most luminous stars. (C) 1996 American Astronom
ical Society.