During the past 20 years, it has been established by several different means that the ISM contains an important gaseous component with temperature of to K (Spitzer 1990). Such gas is called coronal, by analogy with the solar corona. Cooler interstellar gas is heated to temperatures above this range by supernova explosions (Cox & Smith 1974). At these temperatures the gas radiates soft X-rays and cools to about K, where it may remain until it is heated again by another blast wave. Depending on the mean density of the gas and the rate of energy input by supernovae, a small (10%) to large (90%) fraction of the ISM may be filled with gas at of about to K (McKee & Ostriker 1977). Under certain circumstances, a galactic fountain might exist, in which the heated gas escapes into the galactic halo, where it could cool and rain down on the disk (Shapiro & Field 1976).
The Sun is thought to be surrounded by a local bubble with a radius of 100 pc containing coronal gas that produces much, if not all, of the soft X-ray background radiation at energies below about 0.25 keV (McCammon & Sanders 1990). The fact that the background is brighter at the galactic poles, however, has led some investigators to suggest that at least some of the radiation is due to coronal gas in the galactic halo, the diffuse region that surrounds the main disk, at distances of several kiloparsecs from the galactic plane. Recent ROSAT (Roentgen Satellite) observations of shadows in the soft X-ray background that are attributable to foreground clouds tend to support this view (Marshall & Clark 1984; Burrows & Mendenhall 1991; Snowden et al. 1991). Observations with IUE have established that there is indeed ionized gas in the halo but have not established that the temperature of the halo gas is high enough to produce X-rays (Savage & Massa 1987).
Ultraviolet observations of interstellar OVI absorption lines with Copernicus first established the existence of coronal gas in the galactic disk, but could not determine whether such gas extends far into the halo, because the telescope was unable to observe the faint objects needed to probe the distant parts of the halo. Jenkins (1978) found that the mean density of OVI ions in the disk is and tentatively suggested a scale height of only 300 pc to describe the distribution of this gas away from the galactic plane. Such a small scale height would imply that the hot gas is confined to a region fairly close to the galactic disk.
The OVI ion appears with significant fractional abundance only at K for gas in collisional equilibrium, and there are no known sources of ionizing radiation that could explain widespread OVI gas by means of photoionization. On the other hand, the lower ionization lines attributed to gas far above the galactic plane, such as CIV and SiIV, can be explained by photoionization from a combination of halo stars and extragalactic UV background radiation. Thus a crucial test of the nature of the galactic halo can be made by searching for OVI absorption toward extragalactic objects at high galactic latitudes. A hot corona will have a large column density (total number of particles per unit area integrated along the line of sight) of OVI (Edgar & Chevalier 1986), whereas a cooler, photoionized halo, supported at a large distance from the galactic plane by something other than its thermal pressure, will have a much smaller column density of this ion (Fransson & Chevalier 1985).
On the Astro-1 mission, HUT was used to observe the far-ultraviolet spectrum of the quasar 3C273 (Davidsen et al. 1993). At galactic latitude , this line of sight passes almost vertically through the entire halo, providing an excellent probe for coronal gas. The quasar is a good background source for such an observation because it is expected to have an intrinsically smooth spectrum, with no features at the rest wavelength where OVI absorption in our Galaxy would occur. The HUT spectrum of 3C273 has strong absorption features at 1032 and 1038 Å, the wavelengths of the OVI doublet (see Figure ). However, there are several interstellar absorption lines that are expected to contribute to the feature at 1038 Å based on absorption lines seen with HST at longer wavelengths. There are also possible contributions to the OVI features from intergalactic hydrogen clouds whose Lyman- absorption has been seen with HST. The clouds discovered in the Virgo cluster have a redshift that causes their Lyman- lines to contribute to the 1032 Å feature, while Lyman- from a higher redshift cloud may also contribute to the 1038 Å feature. It is possible to place strong upper limits on all of these contaminating features, however, and the result is that OVI absorption is also required to produce the strong features seen with HUT.
Davidsen et al. (1993) find (OVI) is very likely . When compared with the mean density in the disk derived from Copernicus observations (Jenkins 1978), this suggests a scale height of at least 3 kpc, clearly placing the hot gas in the galactic halo if this line of sight is not atypical. A model for a photoionized halo predicted (OVI) (Fransson & Chevalier 1985), whereas a model based on a radiatively cooling corona or galactic fountain predicted (OVI) (Edgar & Chevalier 1986). Clearly, the former model is excluded by the HUT result, but the latter model is supported. However, the radiative cooling model of Edgar & Chevalier (1986) does not actually predict enough of the lower ionization species SiIV and CIV to match the observations of 3C273, indicating that the situation is more complicated. Recent work (Shapiro & Benjamin 1991) on the galactic fountain has shown that self-photoionization of the radiatively cooling gas alters the relative abundances of the lower ionization states appreciably. When this effect is incorporated in the calculations, it is found that the predicted column density ratios of OVI, NV, CIV and SiIV come into close agreement with the observations. Therefore, HUT has provided new evidence supporting the existence of a hot galactic corona, with K, probably originating in a galactic fountain. Of course, it is still necessary to measure the OVI column density along several other lines of sight to prove that the hot gas is widespread in the halo and to map out its distribution. These observations will be attempted with HUT on the Astro-2 mission.