3. THE INTERSTELLAR MEDIUM OF LOCAL GROUP IRREGULARS

3.1. The Magellanic Clouds

The Magellanic Clouds are the two most massive Irrs in the Local Group, and the only two Irrs in immediate proximity to a massive spiral galaxy. Their distances from the Milky Way are 50 kpc (LMC) and 60 kpc (SMC), respectively. They are the only two Local Group Irrs that are closely interacting with each other (and with the Milky Way). According to the earlier definition, the SMC qualifies as a dIrr.

The global distribution of neutral hydrogen within the Magellanic Clouds and other comparatively massive Irrs tends to show a regular, symmetric appearance, in contrast to their visual morphology. On smaller scales, the HI is flocculent and exhibits a complicated fractal pattern full of shells and clumps (e.g., Kim et al. 1998; Stanimirovic et al. 1999). The lack of correlation between the HI shells and the optically dominant HII shells suggests that HI shells live longer than the OB stars that caused them initially (Kim et al. 1999). The HI associated with HII regions is usually more extended than the ionized regions. The fractal structure of the neutral gas is self-similar on scales from tens to hundreds of pc (Elmegreen, Kim, & Staveley-Smith 2001) and appears to result from the turbulent energy input caused by winds of recently formed massive stars and supernova explosions.

With 0.5 × 10⁹ M (Kim et al. 1998) the LMC's gaseous component contributes about 9% to its total mass, while it is ~ 21% in the SMC (HI mass of 4.2 × 10⁸ M; Stanimirovic et al. 1999). In comparison to the Milky Way, the gas-to-dust ratio is roughly 4 times lower in the LMC (Koornneef 1982) and about 30 times lower in the SMC (Stanimirovic et al. 2000), implying a smaller grain surface area per hydrogen atom, fewer coolants, and thus a reduced H₂ formation efficiency (Dickey et al. 2000; Stanimirovic et al. 2000; Tumlinson et al. 2002). Indeed, the total diffuse H₂ mass is only 8 × 10⁶ M in the LMC and 2 × 10⁶ M in the SMC, which corresponds to 2% and 0.5% of their HI masses, respectively (Tumlinson et al. 2002). Also the reduced CO emission from both Clouds (3-5 times lower than expected for Galactic giant molecular clouds) is indicative of the high UV radiation field in low-metallicity environments and hence high CO photodissociation rates (Israel et al. 1986; Rubio, Lequeux, & Boulanger 1993). While high dust content is correlated with high H₂ concentrations, H₂ does not necessarily trace CO or dust (Tumlinson et al. 2002).

Photoionization through massive stars is the main contributor to the optical appearance of the interstellar medium (ISM) at ~ 10⁴ K in the Clouds and other gas-rich, star-forming galaxies. The LMC has a total H luminosity of 2.7 × 10⁴⁰ erg s^-1; 30% to 40% is contributed by diffuse, extended gas (Kennicutt et al. 1995). In the LMC nine HII supershells with diameters > 600 pc are known (Meaburn 1980). Their rims are marked by strings of HII regions and young clusters/OB associations. The standard picture for supershells suggests that these are expanding shells driven by propagating star formation (e.g., McCray & Kafatos 1987). However, an age gradient consistent with this scenario was not detected in the largest of these supershells, LMC4 (Dolphin & Hunter 1998). Nor are other LMC supershells expanding as a whole, but instead appear to consist of hot gas confined between HI sheets and show localized expansion. Supershells in several other galaxies neither show evidence for expansion (e.g., Points et al. 1999), nor the expected young massive stellar populations (Rhode et al. 1999). In contrast, the three HI supershells and 495 giant shells in the SMC appear to be expanding (Staveley-Smith et al. 1997; Stanimirovic et al. 1999).

The hot, highly ionized corona of the LMC with collisionally ionized gas (temperatures 10⁵ K) (Wakker et al. 1998) is spatially uncorrelated with star-forming regions. A hot halo is also observed around the SMC, but here clear correlations with star-forming regions are seen. This corona may be caused in part by gas falling back from a galactic (i.e., SMC) fountain (Hoopes et al. 2002). The OVI column density exceeds the corresponding Galactic value by 1.4 (Hoopes et al. 2002), consistent with the longer cooling times expected at lower metallicities (Edgar & Chevalier 1986).

The Magellanic Clouds, which only have a deprojected distance of 20 kpc from each other, interact with each other and with the Milky Way. Apart from an impact on the structure and star formation histories of these three galaxies (e.g., Hatzidimitriou, Cannon, & Hawkins 1993; Kunkel, Demers, & Irwin 2000; Weinberg 2000; van der Marel et al. 2002), this has given rise to extended gaseous features surrounding the Magellanic Clouds (Putman et al. 2003, and references therein). Part of these are likely caused by tidal interactions, but ram pressure appears to have played an important role as well (Putman et al. 1998; Mastropietro et al. 2004). Metallicity determinations for gas in the Magellanic Stream, which is trailing behind the Magellanic Clouds and subtends at least 10° × 100° on the sky, confirm that the gas is not primordial (Lu et al. 1998; Gibson et al. 2000). The H₂ detected in the leading arm of the Stream may originally have formed in the SMC (Sembach et al. 2001). No stars are known to be connected with the Magellanic Stream (Putman et al. 2003).

Another prominent HI feature is the "Magellanic Bridge" or InterCloud region (10⁸ M; Putman et al. 1998), which connects the LMC and SMC. Cold (20 to 40 K) HI gas has been detected in the Bridge (Kobulnicky & Dickey 1999), and recent star formation occurred there over the past 10 to 25 Myr (Demers & Battinelli 1998). Intermediate-age stars are also present in parts of the Bridge (carbon stars: Kunkel et al. 2000, and references therein). Higher ionized species with temperatures up to ~ 10⁵ K show an abundance pattern suggesting depletion into dust (Lehner et al. 2000). Interestingly, the metallicities of young stars in the Bridge were found to be [Fe/H] -1.1 dex (Rolleston et al. 1999), 0.4 dex below the mean abundance of the young SMC population, which is inconsistent with the proposed tidal origin 200 Myr ago (Murai & Fujimoto 1980; Gardiner & Noguchi 1996).