We propose a systematic CAM imaging photometry in LW9 (14-16um), LW10 (8-15um)
and LW5 (6.5-7um) on a sample of 22 ultraluminous infrared galaxies (ULIRGs);
as well as PHT03 photometry at 25um, PHT37/39 photometry at 160 and 200um,
and PHT-S spectroscopy for 7, 10, and 4 of our targets, respectively.  These
observations will make a complete data set at 6.5, 12, 14, 25, 60, 100, 160
and 200um for our sample when IRAS data and PHT data released from other ISO
programs are combined.
The galaxy sample is a representative subset from a larger set of the ULIRGs
that are complete to a flux density of 2 Jy at 60um and to a redshift of 0.13.
Unlike any existing ISO program on ULIRGs, our sample is also selected to
distribute uniformly in terms of the (projected) nuclear separation, S, of the
two supposedly merging galaxy components, from unresolved cases with S being
less than one or two kpc to cases with S ~ 10 kpc.  Our main goal is to test
if there are correlations between the infrared emission, which accounts for
the bulk of the radiation from an ULIRG, and the advancing degree of the
underlying merging process as indicated by the nuclear separation.
If a statistically significant correlation is identified,  it would provide a
clean and direct support for the hypothesis that ULIRGs are formed exclusively
from galaxy mergers.
With these IR data, we will be able to quantitatively characterize the mid- to
far-IR dust continuum emission; to separate a possible AGN dust component from
the starburst component based on both spectral information over 6 to 200um and
spatial information from our CAM images at a resolution as high as ~2";
to determine the PAH emissions and their contribution to the mid-IR colors;
and ultimately to see if or how each of these dust components correlates with
the hypothesized merging sequence.  This study complements and extends the
existing ISO programs on ULIRGs, but is unique in its sample selection and
specific scientific goals.