We propose to carry out multiwavelength photometry at 4-200 mu of a
  complete all-sky sample of ultraluminous IRAS galaxies from the QDOT
  survey, and to perform spectroscopy with LWS and SWS on a few of the
 brighter galaxies in the sample.

  The discovery of ultraluminous infrared galaxies was one of the major
  results of the IRAS mission.  Subsequent study of these galaxies has
 been hampered by lack of access to most of the infrared waveband.
  Detailed radiative transfer models for starbursts and AGN
  contributions to the far infrared continuum are porrly constrained by
  the IRAS fluxes.  In most cases only 60 and 100 mu fluxes were
  detected by IRAS.  Although there is a concensus that interactions and
  mergers play a major role in generating ultraluminous infrared
  galaxies, there is major controversy about whether massive starbursts
  or shrouded quasars are the principal contributors to the emission.
  Detailed spectral energy distributions over a wide spectral range will
  be a powerful tool in resolving the controversy.  Study of a complete
  sample is essentail for statistical and quantitative conclusions to be
  reached.

  Spectroscopy of a smaple of the brighter galaxies at 4-200 mu will
  allow us to assess the contribution of line emission to the broad band
  fluxes, provide diagnostic evidence of the presence of starbursts, and
 give the contribution of broad features from very small grains or PAHs
 (and the magnitude of depletion of these features) in a variety of
  environments.  The 10 mu silicate feature and its shape is an
 especially important constraint on radiative transfer models.