1994A&A...285...79P

Based on dynamical constraints about the Hubble sequence evolution, observational data and a number of "conspiracies", we propose that the dark matter around spiral galaxies is in the form of cold gas, essentially in molecular form and rotationally supported. The existence of a much larger amount of cold gas in the outer discs of spirals is in accordance with dynamical studies taking into account the bar phenomenon and the secular growth of bulges, leading to the general conclusion that spiral galaxies have to evolve along the Hubble sequence from Sd to Sa over a couple of Gyr's. If so, the varying M/L ratio along the Hubble sequence suggests that dark matter is transformed into stars, i.e. dark matter should be in a sufficiently diluted form of gas. This hypothesis sheds light on several puzzles: 1) the conspiracy of the flat rotation curves, 2) the estimated short gas consumption time of spiral galaxies, leading to the "gas consumption problem", 3) the constant ratio of dark matter to HI mass in the outer spiral discs, 4) the larger amount of visible gas in interacting galaxies with respect to isolated ones, giving rise to gigantic starbursts, 5) the high frequency of absorption lines in front of quasars, together with the large sizes of the absorbing medium around galaxies. Several problems raised by a completely gaseous dark matter in the outer spiral discs are discussed. In particular the stability and self-consistency of cold gaseous discs is shown to be a less severe problem than commonly believed.