Until recently, elliptical galaxies were thought mostly as old systems
having exhausted their interstellar medium (ISM). IRAS changed that
view: dust, as well as molecular gas, is now frequently detected.

A second idea that was revised recently is that dust should be colder
in ellipticals than in spirals because of the lack of star formation.
This is not the case as the very high density of stars in the central
parts of the ellipticals and the concentration of dust in these regions
lead to a higher equilibrium temperature (Sauvage & Thuan 1994).

But the origin of their 12 microns emission is still a matter of
controversy: Knapp et al. (1992) brought evidence for a circumstellar
origin, while Sauvage & Thuan (1994) show that an interstellar origin
would better explain the far-infrared properties.

The proposed observations aim at resolving that controversy. We propose
to make a short wavelength (7-13mic) spectroscopic survey of elliptical
galaxies. This survey will allow us to:

i - Assess the relative importance of circumstellar radiation as
compared to interstellar radiation, as the spectra will be markedly
different in the 2 cases.

ii - Compare the spectral properties of dust in ellipticals and
spirals, in view of the likely external origin of dust in ellipticals.

iii - Identify the heating sources. The MIR-emitting dust is very
sensitive to the shape of the heating spectrum. From its emission
spectrum, we shall constrain the origin of the dust heating, i.e.
residual star formation, hot evolved stars, or an active nucleus?