Whereas IRAS provided the first systematic survey of infrared (IR) emission from normal galaxies, it has been the photometric, imaging and spectroscopic capabilities of ISO (Kessler et al. 1996; Kessler et al. 2003) which have unravelled the basic physical processes giving rise to this IR emission. Thanks to the broad spectral grasp of ISO, the bulk of the emission from dust could be measured, providing the first quantitative assessment of the fraction of stellar light re-radiated by dust. The battery of filters has led to a definitive characterisation of the spectral energy distribution (SED) in the IR, revealing the contribution of the different stellar populations in powering the IR emission. The imaging capabilities have unveiled the complex morphology of galaxies in the IR, and their changing appearance with IR wavelength. They also allowed the exploration of hitherto undetected faint diffuse regions of galaxies. The contribution of different grain populations to the emission has been measured through their characteristic spectroscopic signatures. Knowledge of the emission from cooling lines of the interstellar medium (ISM) has been extended to low luminosity quiescent spiral and dwarf galaxies.
In this review we will concentrate on the mid-IR (MIR) to far-IR (FIR) properties of normal nearby galaxies. By normal we essentially mean that their SEDs are not powered by accretion. We will begin with spiral galaxies, since these have attracted most of the ISO observers' attention. From these objects we will move to the other class of gas-rich galaxies, the dwarfs. However, as the nearby extragalactic population is not only made of spirals and dwarfs, we will conclude by the exploring the very varied IR properties of early-type galaxies.