Very Cold Gas and Dark Matter

We have recently proposed a new candidate for baryonic dark matter: very cold molecular gas, in near-isothermal equilibrium with the cosmic background radiation at 2.73 K. The cold gas, of quasi-primordial abundances, is condensed in a fractal structure, resembling the hierarchical structure of the detected interstellar medium. We present some perspectives of detecting this very cold gas, either directly or indirectly. The H${}_2$ molecule has an “ultrafine” structure, due to the interaction between the rotation-induced magnetic moment and the nuclear spins. But the lines fall in the km domain, and are very weak. The best opportunity might be the UV absorption of H${}_2$ in front of quasars. The unexpected cold dust component, revealed by the COBE/FIRAS submillimetric results, could also be due to this very cold H${}_2$ gas, through collision-induced radiation, or solid H${}_2$ grains or snowflakes. The $\gamma$-ray distribution, much more radially extended than the supernovae at the origin of cosmic rays acceleration, also points towards and extended gas distribution.