Just as modeling of stellar atmospheres is important, we are also beginning to pursue models of interstellar clouds. We need to gain a better understanding of chemical processes and the spectra that would identify them as interstellar clouds collapse, are shocked, form new molecular species, or are lit up by the formation of a new star. We can test such models of chemical networks against spectra ISO is now providing.
A most useful tool for studying a variety of gaseous domains is the CLOUDY
program of
Gary Ferland. His collaborator, Peter van Hoof, displayed how this program
facilitates the
study of photodissociation regions (PDRs)and starburst galaxies. Both Frank
Bertoldi and
Pierre Cox showed us the information that now can be derived for PDRs. In
particular, a large
number of rovibrational emission lines of molecular hydrogen H 
are clearly
discerned. If
we select only the rotational transitions of the ground vibrational state, we
can plot the
excitation energy of a given rotational state J against observed column
densities N(J)
divided by statistical weight g(J)= n(2J+1). Here n = 3 for ortho and 1 for
para states,
respectively the odd and even angular momentum states including J = 0. The
slopes of these
plots for the PDRs in S140 and NGC 7023 reveal surprisingly high temperatures,
respectively,
500 and 580K.
In the planetary nebular NGC 7027 Xiaowei Liu showed us spectral features of
CH 
that
he, José Cernicharo and coworkers had identified. Once again, long-available
laboratory
spectra permitted this identification. Unidentified lines in this and other
planetary nebulae,
however, persist and may arise from unknown molecular species.
A poster by Paolo Saraceno and coworkers showed water vapor emission to be
unexpectedly
low in shocked interstellar regions. They found water vapor cooling to be
typically an order of
magnitude or more lower than cooling through CO transitions. Saraceno feels that
water vapor
may be under abundant in these regions. If so, this raises another problem. In
the sun oxygen
is roughly twice as abundant as carbon. We speak of a cosmic abundance of oxygen
which is
roughly twice that of carbon. In interstellar clouds the abundance of oxygen
appears to be
anomalously low. By now we have the ability to observe oxygen in its atomic form
and as
OH, as H O ice and vapor, as CO, and as CO . Carbon is present in atomic
form and
as CO, CO , CH 
and larger organic molecules, and perhaps as graphite in
grains. This
makes it difficult to see an abundance of oxygen that might still be twice as
high as that of
carbon, unless oxygen was present in molecular form, O . But this may already
be ruled
out by the apparent absence of solid O in ices. Laboratory spectra of ices
consisting of
H O, CO, and CO show changes in shape when O is admixed. Pascale
Ehrenfreund told us that such features are not found in the spectra of
interstellar ices,
suggesting that O is not a significant component. Searches with ISO for
gaseous O
are also being carried out. It will be interesting to see what they will teach
us. Molecular
oxygen is difficult to observe because of its symmetric dipole structure, and we
may not be
able to put significant upper limits on its abundance. The interstellar
abundance of oxygen is
one of the log-standing problems of interstellar chemistry that ISO may still
have an
opportunity to clarify.