4.2. Mid-Infrared Spectra
In the integrated spectra of galaxies, the mid-infrared marks the transition from emission dominated by stellar photospheres to re-radiation by interstellar dust. ISO has shown the details of this transition for the first time by providing continuous coverage, and filling in some crucial details as discussed in this and the next two sections. The relevant data were acquired with ISO-PHOT, PHT-S module from 2.5 to 5 and from 5.7 to 11.6 µm (Lemke et al. 1996); with ISO-CAM in the CVF (Circular Variable Filter) mode from 5 to 16.5 (C. Césarsky et al. 1996); and with SWS from 2.5 to 45 µm (de Graauw et al. 1996).
The transition from stellar to interstellar emission is well illustrated by the spectra of Virgo Cluster galaxies collected by Boselli et al. (1998). Its precise location and therefore the interpretation to attach to mid-infrared fluxes can be parametrized by a ratio such as IR/B (Section 3.1.2). Interstellar dust emission takes over by 5 µm when this ratio exceeds 0.5, and at shorter wavelengths for higher ratios. As might be expected, Elliptical galaxies are dominated by stellar emission, both photospheric and from circumstellar dust shells, and therefore provide the templates that one subtracts to isolate the interstellar emission component in Spiral galaxies (Boselli, Lequeux & Contursi 1997; Madden, Vigroux & Sauvage 1997). In addition, ISO sensitivity has allowed us to study the small amounts of ISM contained in Elliptical galaxies, and to look into differences with the ISM of Spirals (Knapp et al. 1996; Fich et al. 1999, Madden et al. 1999, Malhotra et al. 1999; note also the study of E+A galaxies in the Coma Cluster by Quillen et al. 1999).