Annu. Rev. Astron. Astrophys. 1997. 35: 637-675 Copyright © 1997 by Annual Reviews. All rights reserved

2.3. M 31 (NGC 224)

The stellar population of M 31 was reviewed by van den Bergh (1991b), and again we restrict discussion to significant subsequent developments.

The field nondisk population has been studied by several groups, following Mould & Kristian (1986; see also Crotts 1986). These authors established, from V and I data that reach several magnitudes down the giant branch, that the bulge/halo of M 31, at 7 kpc from its center, has mean metallicity like the Galactic globular 47 Tuc, [Fe/H] ~ -0.7, and a significant dispersion in metallicity, when assuming an old population, down to ~ -2 dex and up towards solar. Similar conclusions have been reached from HST data for the outer regions of M 31 (~ 10 kpc) by Holland et al (1996) and by Rich et al (1996) at ~ 30 kpc from the center, which limits the amplitude of any chemical abundance gradients, assuming always that one is dealing with an old stellar population.

These HST data also established firmly the scarcity of Blue Horizontal Branch (BHB) stars in the halo of M 31, which confirms the suggestion by Pritchet & van den Bergh (1987, 1988). A few BHB stars were found by Holland et al (1996), who suggest that the horizontal branch (HB) morphology is apparently too red for the derived broad metallicity distribution. If one assumes that the horizontal branch traces a population as old as the Galactic halo globular clusters, then the M 31 field population suffers a severe "second-parameter problem."

Assuming that the derived broad metallicity distribution is well-established, does this lack of a significant BHB population imply a young age for M 31? Age can affect HBmorphology in that younger populations are redder at a given metallicity, other things being equal (e.g. Lee 1993, who also demonstrates the effects of many other parameters), so that it is of interest to consider this possibility [while recalling that Richer et al (1996) argue quite convincingly, based on relative ages for those Galactic globular clusters with main sequence turn-off photometry, that age is not the dominant "second parameter" of HB morphology, at least in these systems]. Indeed, the presence of bright stars, identified as intermediate-age AGB stars, has been suggested from (prerefurbishment) WF/PC HST VI data at least within the inner 2 kpc of the bulge (Rich & Mighell 1995). Morris et al (1994) argued for a ubiquitous strong luminous AGB component, with a typical age of 5 Gyr, from their ground-based V and I data that reaches the bright giants in various fields of M 31, 16-35 kpc along the major axis of the disk and one probing the halo at 8 kpc down the minor axis (close to the field of Mould & Kristian 1986). Rich et al (1996), and also Holland et al (1996), find no evidence for an extended giant branch in their WF/PC2 HST data for fields in the outer halo, at 10-30 kpc from center, where again the RHB/clump is dominant, with essentially no trace of a BHB. Thus, the data describing possible metallicity/age effects remain unclear.

Large-scale surface photometry of the disk and of the bulge of M 31, in many broadband colors, was obtained and analyzed by Walterbos & Kennicutt (1988). They found that there was no color gradient in the bulge and that the inner disk and the bulge have essentially the same colors, i.e. those of "old, metal-rich stellar populations." This similarity of broadband colors has subsequently been found for a large sample of external disk galaxies, as discussed in Section 3, and clearly must be incorporated into models of the formation and evolution of bulges (see Section 5 below). Walterbos & Kennicutt also derived structural parameters for the disk and bulge that are consistent with the correlation between scale lengths found for the larger sample of more distant disk galaxies by Courteau et al (1996). In terms of total optical light, the bulge-to-disk ratio of M 31 is about 40%.

Pritchet & van den Bergh (1996) emphasize that a single R^1/4-law provides a good fit to their derived V-band surface photometry (from star counts), with no bulge/halo dichotomy. The R^1/4 component is significantly flattened, with axial ratio of 0.55, which is similar to the value for the metal-poor halo of the Milky Way (Larsen & Humphreys 1994, Wyse & Gilmore 1989).

In contrast to the metal-poor halo of the Milky Way, which is apparently flattened by anisotropic velocity dispersions, the bulge of M 31 has kinematics consistent with an isotropic oblate rotator, with mean rotational velocity of ~ 65 km/s and velocity dispersion of ~ 145 km/s (McElroy 1983), which are typical of external bulges (Kormendy & Illingworth 1982).

Thus, although Baade (1944a, b) identified the "bulge" of M 31 (which we may now define to be field nondisk stars at distances up to 35 kpc from the center of M 31) with Population II (similar to the Milky Way halo), the dominant tracers of the M 31 bulge do not share the characteristics of classical Galactic Population II, as they are neither of low mean metallicity nor have little net rotation (see Wyse & Gilmore 1988 for further development of this point, in the context of thick disks).

There are around 200 confirmed globular clusters associated with M 31 (e.g. Fusi-Pecci et al 1993). The distribution of their metallicities has a mean of around -1 dex, which is more metal-poor than the field stars, with a range of perhaps 1 dex on either side (e.g. Huchra et al 1991, Ajhar et al 1996). The inner metal-rich clusters form a rapidly rotating system, whereas the outer metal-poor clusters have more classical "hot" halo kinematics (e.g. Huchra 1993; see also Ashman & Bird 1993 for further discussion of subsystems within the globular clusters). The overall globular cluster system has a projected number density profile that may be fit by a de Vaucouleurs profile (although the central regions fall off less steeply) with an effective radius of ~ 4-5 kpc (Battistini et al 1993). This is more extended than the R^1/4 fit to the field stars. Thus, in terms of kinematics, metallicity, and structure, there may be evidence for a bulge/halo dichotomy in M 31 if the halo is traced by the globular clusters and the bulge by field stars. Note that, although there are exceptions, the spatial distributions of globular cluster systems and underlying galaxy light are similar to the first order (Harris 1991).

As seems to be the case for any system studied in sufficient detail, the morphology of the very central regions of M 31 is clearly complicated, with twisted isophotes (Stark 1977), gas kinematics that may trace a bar (e.g. Gerhard 1988), inner spiral arms (e.g. Sofue et al 1994), and two nuclei (Bacon et al 1994) that may indicate a tilted inner disk (Tremaine 1995). These phenomena have been modeled recently by Stark & Binney (1994) by a spherical mass distribution plus a weak prolate bar, with the bar containing one third of the mass within 4 kpc (the corotation radius). The association of the bulge with this bar, which one might be tempted to adopt by analogy with the Milky Way, is unclear.