Interstellar Matter at Large Distances from the Galactic Plane.

The interstellar gas at large distances z from the galactic plane is studied by the absorption lines it produces on the spectrum of distant stars off the Milky Way. From the statistics of multiple lines in various ranges of z, it is shown that some gas clouds probably exist at z = 1 kpc. The number of clouds observed in 0 5 <z < 1 kpc has been found to be larger than would be expected from the known distribution of their velocity components in the galactic plane. The apparent asymmetry in the distribution of high-velocity clouds is explained as the result of decreased chances of collisions in the z-direction and also in terms of an intrinsic anisotropy in the mechanism accelerating the clouds. The typical time required for the clouds to reach their actual probable height from z = 0 is evaluated to be 40 X 106 years From the line intensities and by assuming cosmic abundance of the elements, a relation between the linear dimensions and the densities of the clouds is established. Irrespective of whether the clouds are H I or H ii regions, it is found that their continued existence for 40 X 106 years requires the operation of a process preventing them from expanding. The physical conditions prevailing in a galactic halo or corona exerting pressure on the clouds are next analyzed. It is shown how the observations rule out a halo with an electron temperature T, around 10 K. A corona with T = 106 K, as postulated by Spitzer, on the other hand, is found admissible, provided that the high-velocity clouds at high z are H ii regions. The large energy input by conduction from the corona may be balanced by radiative losses only at about T - 10 K. Next the ionization e uilibrium in the clouds is briefly discussed, and it is suggested that the anomalous abundance ratio Na Ca observed in interstellar space is the result of using an unrealistic mean stellar radiation field in the photoionization computations. In this context, the results of a calculation of the ionization equilibrium of aluminum is presented It is shown that the Al I line at X 3964 should have a strength about one-twentieth that of Ca I X 4226. In a final section the possible mechanisms by means of which interstellar clouds may be accelerated are discussed. It is shown how the operation of the Oort-Spitzer process requires a ratio between the total amounts of ionized and neutral interstellar matter much larger than is observed. The relevance of magnetic fields in accelerating small masses of ionized field4ree material is thereby emphasized.