5.1. Molecular Lines
One of the great achievements of ISO was the detection of the long-sought molecular hydrogen transitions from rotationally excited states, and the realization of their promise as accurate tracers of the H2 mass (e.g. Draine & Bertoldi 1999), allowing for the first time a direct gauge of the dominant phase of the star-forming ISM. Most of the H2 data were collected by the SWS (de Graauw et al. 1996), and their interpretation and analysis in the Milky Way is discussed by Cox & Boulanger elsewhere in this volume. Several of these lines, especially S(0), S(1) and S(2) were detected in galaxies early in the ISO mission, permitting a detailed discussion of the gas phase conditions and heating mechanisms, as in the study by Valentijn et al. (1996) of the nuclear star burst of NGC 6946. Because of their high excitation levels however, these lines tend to trace H2 warmer than ~ 100 K, and are therefore easier to detect in the more intense star formation environments (Kunze et al. 1996; Kunze et al. 1999). Still, Valentijn et al. (1999) report the detection of H2 emission from the extended disk of the edge-on galaxy NGC 891. They detect S(0) and S(1) at eight positions, tracing the emission out to 12 kpc from the nucleus of the galaxy, and derive H2 temperature constraints and molecular mass estimates.
Hydroxyl (OH) was also detected in starburst galaxies, but in absorption rather than emission. A first report by Skinner et al. (1997) showed OH absorption at 35 µm in the spectrum of Arp 220, and confirmed for the first time pumping by infrared photons as the excitation mechanism behind OH mega-masers. Bradford et al. (1999) report OH in absorption at 35, 53 and 119 µm in NGC 253, and resolve the line at 119 µm with the LWS Fabry-Pérot mode. They estimate total column and excitation temperatures for the OH, and constrain the geometry of the molecular material and its relationship to the infrared-emitting dust.