We have inverted high-phase-angle Voyager images of Neptune to determi
ne the atmospheric extinction coefficient as a function of altitude an
d the scattering phase function at a reference altitude. Comparisons b
etween theoretical models and observations help separate the contribut
ions from molecular Rayleigh and aerosol scattering and help determine
the variation of the aerosol size, concentration, and scattering prop
erties with altitude. Further comparisons between models and data allo
w us to place constraints on the location and composition of the hazes
, the concentration and downward Bur of certain condensible hydrocarbo
n gases, the eddy diffusion coefficient in the lower stratosphere, and
the thermal profile in parts of Neptune's stratosphere. We find that
a distinct stratospheric haze layer exists near 12(-1)(+1) mbar in Nep
tune's lower stratosphere, most probably due to condensed ethane, The
derived stratospheric haze production rate of 1.0(-0.3)(+0.2) x 10(-15
) g cm(-2) sec(-1) is substantially lower than photochemical model pre
dictions. Evidence for hazes at higher altitudes also exists. Unlike t
he situation on Uranus, large particles (0.08-0.11 mu m) may be presen
t at high altitudes on Neptune (e.g., near 0.5 mbar), well above the r
egion in which we expect the major hydrocarbon species to condense. Ne
ar 28 mbar, the mean particle size is about 0.13(-0.02)(+0.02) mu m wi
th a concentration of 5(-3)(+3) particles cm(-3). The cumulative haze
extinction optical depth above 15 mbar in the clear filter is similar
to 3 x 10(-3), and much of this extinction is due to scattering rather
than absorption; thus, if our limb-scan sites are typical, the hazes
cannot account for the stratospheric temperature inversion on Neptune
and may not contribute significantly to atmospheric heating. We compar
e the imaging results with the results from other,observations, includ
ing those of the Voyager Photopolarimeter Subsystem, and discuss diffe
rences between Neptune and Uranus. (C) 1995 Academic Press, Inc.