Although it is known that our Galaxy is not particularly actively forming stars [the overall SFR is of the order of 3 M yr-1, compared to 50 M yr-1 in starburst galaxies like M82 (Berkhuijsen 1986; Telesco 1988; Heckman et al. 1990; see also Section 6.1)], some of the all-sky surveys mentioned above and other observations have led to the detection of several phases of the ISM outside its thin disk, in the halo. Therefore, again, Galactic studies are an ideal starting point into the subject.
4.1 H I in the Halo
Lockman (1984) detected H I emission from outside the thin disk of the Milky Way, the so-called ``Lockman layer''. An even thicker layer of H I - with a velocity dispersion of 60 km s-1 and a scale height of 3.2 kpc - was detected by Kalberla et al. (1997). H I surveys also revealed individual gaseous filaments which are located outside the thin disk, like for instance several chains of peculiarly moving high-velocity clouds (HVCs; e.g. van Woerden et al. 1985). Recently, Pietz et al. (1996) - using the Leiden-Dwingeloo Survey (Hartmann & Burton 1997) - detected a velocity bridge between the HVCs of the so-called ``chain C'' complex in the Galactic halo and intermediate-velocity (IV) and low-velocity (LV) clouds presumably falling into the disk and thereby interacting with the disk medium. These clouds provide evidence that the halo H I is kinematically linked to the WNM in the disk of the Milky Way.
4.2 Warm and Hot Ionized Gas in the Halo
Several different observing techniques led to the detection of ionized gas at various temperatures (about 104 K to a few times 106 K) in the halo of the Milky Way.
Diffuse warm ionized gas with temperatures in the 104 K range, similar to that found outside H II regions within the Galactic disk, was detected in H, [N II], [S II], and other line emission in the halo. The extraplanar DIG (eDIG), also called the ``Reynolds layer'' (e.g. Reynolds 1990a, b), is a layer of thermal electrons with a scale height of 0.67+0.17-0.14 kpc (Nordgren et al. 1992). Most of this ionized gas is found at z distances from the Galactic plane of at most a few kpc. Only in the Magellanic stream, which is interacting with the Galactic H I gas, has H emission been found as far away from the disk plane as 50 kpc (Weiner & Williams 1996). Hotter thermal gas, with temperatures of order 105 K, was detected in UV absorption lines against background continuum sources (Savage & de Boer 1979; Pettini & West 1982; Savage & Massa 1987).
The previously mentioned HVCs as well as other gas clouds cast shadows ag-ainst a diffuse background of soft (kT 0.3 keV) X-ray emission coming from beyond (Snowden et al. 1991; Herbstmeier et al. 1995). Distance measurements have shown that shadowing structures, like e.g. the Draco Nebula, are located high - several hundred pc - above the Galactic plane (Lilienthal et al. 1991), which places the shadowed X-ray emitting gas even farther away from the disk. The gas emitting in the 0.3 keV energy range is probably a hot thermal plasma with a temperature of about 3 x 106 K (e.g. Snowden et al. 1995, 1997). Breitschwerdt & Schmutzler (1994), however, argue that such emission might also come from delayed recombination of rapidly cooling gas at T 5 x 104 to 5 x 105 K.
In addition to these warm and hot ionized gas phases, there is now also evidence that the Galactic center region is driving a wind containing a 10 keV hot gaseous component (Koyama et al. 1986; Yamauchi et al. 1990).
4.3 The Galactic ``Cirrus''
IRAS detected both diffuse FIR emission in the Galactic halo (Hauser et al. 1984) and filamentary emission of dust associated with high-latitude H I clouds, the so-called ``Galactic cirrus'' (Low et al. 1984). Associated CO emission was also detected (e.g. Stark 1995), proving that a molecular cirrus exists in the Galactic halo. A ``thick disk'' of molecular gas above the central disk of the Milky Way has been reported as well (Dame & Thaddeus 1994). These detections imply that cold matter (T 50 K) exists in the Galactic halo.
4.4 Cosmic Rays and B-fields
Being located within the thin disk, it is difficult for us to determine whether gas for which we have no kinematic information is located in the halo or possibly in our local vicinity, still within the disk. To address this problem, Beuermann et al. (1985) developed a model showing that after removal of emission from nearby structures in the Galactic disk, seen in projection at high latitudes (like, e.g., the North Polar Spur), the distribution of radio continuum emission of the Milky Way in the 408 MHz all-sky survey is not compatible with the existence of only a thin disk. They found signatures of a radio halo, encompassing the disk, with a z thickness that depends on the local level of SF in the underlying disk (their Fig. 9). Complementary direct evidence for the existence of a Galactic CR (nucleon) halo comes e.g. from 10Be measurements (Simpson & García-Muñoz 1988).
Han & Qiao (1994) found a vertical component of the local Galactic B-field of 0.2 to 0.3 µG. Based on data of a large sample of galaxies, Lisenfeld et al. (1996b) have shown that the assumption of energy equipartition between CRs, B-fields, and the interstellar radiation field (ISRF) is not grossly wrong. Thus, the presence of a Galactic radio synchrotron halo implies the existence of B-fields at high z.
4.5 Signs of Ongoing Disk-Halo Interactions
Clearly, there is plenty of observational evidence for the existence of a gaseous Galactic halo, comprised of all components of the ISM found in the disk. Based on the shape of the radio halo, as modeled by Beuermann et al. (1985), we can quite safely assume that the Milky Way has an extended gaseous halo, reaching galactocentric radii of at least 10 kpc. To maintain such a halo, it must be fed with both matter and energy. Various energy sources can possibly feed halos; among them, outflows from SF regions in the Galactic disk (see Section 7.1.1). Assuming that this is correct, disk-halo interactions must be occurring in many locations. However, despite this evidence, there are hardly any signs for current ongoing disk-halo interactions in our Galaxy. So far, only two morphological structures have been clearly detected and interpreted as direct evidence for disk-halo interactions. The first is the so-called ``Stockert chimney'' (Müller et al. 1987; Kundt & Müller 1987), a radio continuum spur with a spectral index indicating thermal emission. However, this spur is not very long (only ~ 300 pc) and the underlying stellar associations do not produce much energy. There is better evidence for outflow from the Galactic disk in a second chimney, which was recently detected in the Perseus spiral arm (Normandeau et al. 1996). Here, not only can a cavity blown into the ambient (H I) gas be seen, but also a compression of molecular gas clouds within the cavity - traced by their head-tail structure. The third piece of evidence for ongoing disk-halo interactions is the link of HVCs with gas at lower velocities in the Galactic disk (Pietz et al. 1996; see above). These clouds are probably falling into the Galactic disk from the halo.
Thus, although our own Galaxy offers the possibility of studying the different phases and morphological components of the ISM at close range, there are also observational disadvantages, primarily due to the unfavorable viewing geometry and the associated lack of a grand overview. Our viewing range is limited, which leads to a scarcity of objects to study. For this reason observations of external galaxies can provide us with substantial additional information. Another motivation for investigations of external galaxies is to determine how many galaxies generate enough energy to support a gaseous halo and which conditions and processes lead to its creation. We can then also address the question of how our Galaxy compares to other spirals.