The galactic fountain is formed by hot gas rising from the galactic disk. The lines of force of the interstellar magnetic field are dragged along in the flow. This lecture deals with the geometry and topology of the configuration of the field that is set up. The fountain flow acts as a dynamo, and the presence of the magnetic field plays an important part in the ionization balance of the gas as it returns towards the disk.

Neutral gas extends out of the disk to form a modest halo over most of the inner Galaxy except for the area at R ≲2.5 kpc. The HI halo probably consists of several populations including a relatively quiescent layer and a cloudy component that has peculiar velocities. This latter component often appears between 150 and 400 pc from the plane. Most of the neutral gas at |z| ≳500 pc has the same kinematics as the gas below it.

This contribution describes high-velocity clouds (HVCs), neutral hydrogen moving with velocities inexplicable by differential galactic rotation. They have been invoked as evidence for infall of gas to the Galaxy, as manifestations of a galactic fountain, as energy source for the formation of supershells, etc. It is becoming clear that a single model will not suffice to explain all HVCs. A better understanding is mainly hampered by the fact that the distance remains unknown. Many aspects to the study of HVCs will be discussed here.

The evidence for the existence of molecular clouds at large distances from the Galactic plane is reviewed. The molecular clouds at high Galactic latitudes are shown to be largely confined to the Galactic plane. There is evidence for one giant molecular cloud as much as four scale heights from the Galactic plane, but given the sample size from which the cloud is drawn, it is reasonable to suppose that it is part of the tail of the thin disk population. There is weak evidence that one star-forming molecular cloud may have originated in the Galactic halo. On the basis of kinematic evidence however, it is shown that there are three molecular clouds identified at high galactic latitude that, if not at high z, are likely to have resulted from interaction with gas in the halo. Understanding how these clouds have formed is likely to be an important key to understanding how the halo interacts with the disk gas.

Of the many and various means for observing the interstellar medium, and halo gas in particular, optical and ultraviolet absorption techniques provide both unique opportunities and unique limitations. As its title indicates, this article speaks specifically to the contributions from absorption methods to our knowledge of halo gas with temperatures below 105 K. A brief description of the trade-offs and benefits of the methods is therefore useful to set the stage for interpreting the observations of halo gas.

Warm (≈ 104 K), diffuse H+ is a significant component of the interstellar medium within the Galactic disk and lower halo. This gas accounts for about one quarter of the interstellar atomic hydrogen, consumes a large fraction of the interstellar power budget, and appears to be the dominant state of interstellar matter 1 kpc above the midplane. The origin of this ionized gas is not yet established; however, of the known sources of ionization only 0 stars and perhaps supernovae produce enough power to balance the “cooling” rate of the gas. If 0 stars are the source of the ionization, then the interstellar HI, including the extended “Lockman layer”, must have a morphology that allows about 14% of the Lyman continuum photons emitted by the stars to travel hundreds of parsecs within the Galactic disk and up into the lower halo.

It is shown that, in order to reproduce the observational properties of our Galaxy, models of galactic chemical evolution require a conspicuous amount of gas falling on the disk with very slow time decay. The constraints on the characteristics of this gas derived from the models are presented.

Surveys for atomic hydrogen at very high velocities (|V| ≥ 140 km s–1) in the galactic center and anticenter regions of the sky reveal a net inflow of gas toward the Milky Way. In the anticenter, the collisions of infalling clouds with galactic material trigger the most energetic structural disturbance in the Galaxy, the “anticenter supershell”.

Large amplitude waves have been found in the morphology and velocity patterns of several long filaments of HI at high latitude. HI in the filaments is controlled by magnetic fields and the velocity patterns and morphology bear the hallmarks of Alfvén waves. Enhanced emission features (EEFs), traditionally referred to as “clouds,” are seen wherever a segment of flux tube is viewed end-on. This suggests that HI emission structure teaches us about field geometry and not about cloud physics. Similar effects have been recognized in other regions mapped with high-resolution as well as in completely mapped high-velocity “clouds.”

The thick disk is the stellar disk-halo connection. At least near the solar circle, this component is on average as old as the system of disk globular clusters, or ~ 12 Gyr. This implies that it most probably formed early in the process of Galaxy formation, so that its properties – chemical abundances, stellar kinematics and spatial distribution – contain clues to the physics of these stages of Galaxy evolution. Its present-day importance for the interstellar disk-halo connection lies in the evolution of its constituent stars – gas loss through winds on the red giant and asymptotic giant branches, through planetary nebulae prior to white dwarf formation, and through supernovae. This gas loss results in mass injection, momentum injection and energy injection into the interstellar medium from a stellar population with a scale height of ~ 1kpc.

It seems very probable that young, luminous B stars occur in the galactic halo. An origin in the disc followed by ejection may account for many of these stars; certain proposed formation-ejection mechanisms involve the infall of halo material. The kinematics and supposed locations of some halo B stars seem incompatible with an origin in the disc, suggesting that these stars may have formed in the halo.

In the outer Galaxy (defined here as those parts of our system with galactocentric radii R>R0) the HI gas density (Wouterloot et al., 1990), the cosmic ray flux (Bloemen et al, 1984) and the metallicity (Shaver et al., 1983) are lower than in the inner parts. Also, the effect of a spiral density wave is much reduced in the outer parts of the Galaxy due to corotation. This changing environment might be expected to have its influence on the formation of molecular clouds and on star formation within them. In fact, some differences with respect to the inner Galaxy have been found: the ratio of HI to H2 surface density is increasing from about 5 near the Sun to about 100 at R≈20kpc (Wouterloot et al., 1990). Because of the “flaring” of the gaseous disk, the scale height of both the atomic and the molecular gas increases by about a factor of 3 between R0 and 2R0 (Wouterloot et al., 1990), so the mean volume density of both constituents decreases even more rapidly than their surface densities. The size of HII regions decreases significantly with increasing galactocentric distance (Fich and Blitz, 1984), probably due to the fact that outer Galaxy clouds are less massive (see section 3.3), and therefore form fewer O-type stars than their inner Galaxy counter parts. There are indications that the cloud kinetic temperature is lower by a few degrees (Mead and Kutner, 1988), although it is not clear to what extent this is caused by beam dilution.

Highly ionized gas in the galactic halo has been detected through UV absorption and emission lines. In absorption the species studied include Si IV, C IV and N V. The UV emission studies have recorded C IV and O III]. Absorption measurements toward galactic stars reveal that the |z| distribution of the gas is roughly exponential with a scale height of approximately 3 kpc and has column densities perpendicular to the galactic plane of N ~ 2×1013, 1×1014 and 3×1013 atoms cm−2, for Si IV, C IV and NV, respectively. Similar absorption line profiles for these species suggests a common process for their origin. The presence of N V absorption implies the existence of some gas with a temperature near T ~ 2×105 K. The highly ionized absorbing gas toward distant stars in direction b < −50° has simple and relatively narrow line profiles (FWHM ~ 45 to 70 km−1) and small average LSR velocities while the gas in the direction b > 50° reveals a complex pattern of motions with substantial inflow and outflow velocities. Galactic rotation has an appreciable effect on the absorption line profiles to very distant stars located in the low halo. C IV emission has been seen at greater than a 3σ level of significance in 4 of 8 directions. The emission brightens toward the galactic poles and has a polar intensity I(C IV) ~ 5000 photons cm−2s−1ster−1. If the emitting and absorbing gas coincide in space the measurements imply ne ~ 0.01 cm−3 and P/k ~ 2000 cm−3 K for gas with T ~ 105 K. This phase of the gas fills only a small volume of the space (f ~ 0.03) and accounts for only a small fraction of the total column density of gas perpendicular to the galactic plane [~3×1018 atoms cm −2 vs 3.5×1020 atoms cm −2 for H I and 1×1020 atoms cm −2 for H+]. However, the gas provides a large EUV/UV emission line flux (~1×10−5erg. cm−2 s−1) which corresponds to a H I ionizing flux of ~2×105 ionizations cm−2 s−1. Gas with T near 2×105 K cools very rapidly. Its origin may be associated with the cooling gas of a galactic fountain flow or with thermal condensations in cosmic ray driven fountains. In the nonequilbrium cooling of a Galactic fountain, a flow rate of 4 MO/ year to each side of the Galaxy is required to produce the amount of N V absorption found in the halo while a flow rate 5x larger is required to produce the observed level of C IV emission.

The interstellar medium and its hot gas component are very different from common conception.

We review the current data on the not-so-dark sky covering infrared, visible and ultraviolet wavelengths. Here, we are mainly concerned with the emission from the interstellar gas and dust above and below the galactic plane. Zodiacal light is not discussed in detail and emission from unresolved stars is briefly mentioned. Recent improvements in these studies have been made with the use of new satellite UV data, the use of high-performance CCD in the visible spectrum and extensive analyses from the Infrared Astronomical Satellite (IRAS). We show that cirrus clouds which subtend a large solid angle at high galactic latitudes are made of neutral gas and dust, are within a few hundred parsecs of the Sun, and are almost optically thin up to UV wavelengths. The brightness of these clouds, expressed as vIν = λIλ, is estimated to be within 10–8 and 10–7 W m–2 sr–1 at almost all wavelengths from λ = 0.1 to 300 μm and peaks at 150 μm, for a typical column density of 3 × 1020 H cm–2. They may yield the fundamental limitation to all extragalactic and halo studies.