We present a compilation of PRONAOS-based results concerning the temperature dependence of the dust submillimeter spectral index, including data from Galactic cirrus, star-forming regions, dust associated to a young stellar object, and a spiral galaxy. We observe large variations of the spectral index (from 0.8 to 2.4) in a wide range of temperatures (11 to 80 K). These spectral index variations follow a hyperbolic-shaped function of the temperature, high spectral indices (1.6-2.4) being observed in cold regions (11-20 K) while low indices (0.8-1.6) are observed in warm regions (35-80 K). Three distinct effects may play a role in this temperature dependence: one is that the grain sizes change in dense environments, another is that the chemical composition of the grains is not the same in different environments, a third one is that there is an intrinsic dependence of the dust spectral index on the temperature due to quantum processes. This last effect is backed up by laboratory measurements and could be the dominant one.

We map a 50' x 30' area in and around the M17 molecular complex with the French submillimeter balloon-borne telescope PRONAOS, in order to better understand the thermal emission of cosmic dust and the structure of the interstellar medium. The PRONAOS-SPM instrument has an angular resolution of about 3', corresponding to a size of 2 pc at the distance of this complex, and a high sensitivity up to 0.8 MJy/sr. The observations are made in four wide submillimeter bands corresponding to effective wavelengths of 200, 260, 360 and 580 um. Using an improved map-making method for PRONAOS data, we map the M17 complex and faint condensations near the dense warm core. We derive maps of both the dust temperature and the spectral index, which vary over a wide range, from about 10 K to 100 K for the temperature and from about 1 to 2.5 for the spectral index. We show that these parameters are anticorrelated, the cold areas (10-20 K) having a spectral index around 2, whereas the warm areas have a spectral index between 1 and 1.5. We discuss possible causes of this effect, and we propose an explanation involving intrinsic variations of the grain properties. Indeed, to match the observed spectra with two dust components having a spectral index equal to 2 leads to very large and unlikely amounts of cold dust. We also give estimates of the column densities and masses of the studied clumps. Three cold clumps (14-17 K) could be gravitationally unstable.

We present an extinction map of the Polaris molecular cirrus cloud derived from star counts and compare it with the Schlegel et al. (1998) extinction map derived from the far--infrared dust opacity. We find that, within the Polaris cloud, the Schlegel et al. Av values are a factor 2 to 3 higher than the star count values. We propose that this discrepancy results from a difference in $\tau_{FIR}/ A_V$ between the diffuse atomic medium and the Polaris cloud. We use the difference in spectral energy distribution, warm for the diffuse atomic medium, cold for the Polaris cloud, to separate their respective contribution to the line of sight integrated infrared emission and find that the $\tau_{FIR}/ A_V$ of cold dust in Polaris is on average 4 times higher than the Schlegel et al. value for dust in atomic cirrus. This change in dust property could be interpreted by a growth of fluffy particles within low opacity molecular cirrus clouds such as Polaris. Our work suggests that variations in dust emissivity must be taken into account to estimate Av from dust emission wherever cold infrared emission is present (i.e. molecular clouds).