Aspects of internal secular evolution have long been thriving "cottage industries" (an early review is in Kormendy 1982). Kormendy & Kennicutt (2004) provide a synthesis of these many lines of research, both observational and theoretical. Other reviews are in Sellwood & Wilkinson (1993), Kormendy (1993), Buta & Combes (1996) Kormendy & Cornell (2004), Kormendy & Fisher (2005), Athanassoula (2007), Peletier (2008), and Combes (2007, 2008). With limited space, this paper concentrates on new observations of pseudobulge properties.
Whatever the engine, internal evolution has similar consequences. Like all self-gravitating systems, galaxy disks tend to spread - the outsides expand and the insides contract (Tremaine 1989). This is as fundamental to disk evolution as core collapse is to globular clusters, as the production of hot Jupiters and colder Neptunes is to the evolution of planetary systems, and as evolution to red giants containing proto-white-dwarfs is to stellar evolution (Kormendy & Fisher 2005; Kormendy 2008). In galaxy disks, gas infall and star formation builds dense central components that get mistaken for bulges but that were not made by galaxy mergers. They come in several varieties depending on what drives the evolution. Pseudobulges made from disk gas are often but not always disky (Kormendy 1993; KK04; Fisher & Drory 2008a). Box-shaped bulges also are disk phenomena: they are parts of edge-on bars (Combes & Sanders 1981; Combes et al. 1990; Pfenniger & Friedli 1991; Raha et al. 1991; Kuijken & Merrifield 1995; Merrifield & Kuijken 1999; Bureau & Freeman 1999; Bureau, Freeman, & Athanassoula 1999; Athanassoula 2005, 2007). Nuclear bars are connected with disky pseudobulges (they rotate rapidly) and may be a subset of them. Other morphology that identifies pseudobulges includes nuclear star formation rings and spiral structure. It is convenient to have one name - "pseudobulges" - for all central, high-density products of disk secular evolution.
How to identify pseudobulges is discussed in KK04. Prototypical examples that are more disky than classical bulges were first recognized by their rapid rotation (Figure 2). Disks have large Vmax / and plot above the oblate line when seen at inclinations other than edge-on. Early identification of very disky (e. g., NGC 4736, NGC 3945) and moderately disky (e. g., NGC 2950, also a nuclear bar) pseudobulges have recently been augmented as shown in Figure 2.
Figure 2. Relative importance of rotation and random velocity as a function of observed ellipticity = (1 - axial ratio) for various kinds of stellar systems. Here Vmax / is the ratio of maximum rotation velocity to mean velocity dispersion interior to the half-light radius. The "oblate" line approximately describes oblate-spheroidal systems that have isotropic velocity dispersions and that are flattened only by rotation; it is a consequence of the tensor virial theorem (Binney & Tremaine 1987). This figure is updated from KK04.
Figure 2 is an approximate analysis based on long-slit, major-axis spectra. Integral-field spectroscopy from the SAURON team now provides beautifully detailed detections of rapidly rotating, disky pseudobulges. Often (Figure 3), it is exactly the high-surface-brightness center - where the projected brightness profile rises above the inward extrapolation of the outer disk profile - that shows rapid rotation and a corresponding inward decrease in velocity dispersion. Many of these kinematically decoupled components are also younger than the rest of the inner galaxy. Some counter-rotate (McDermid et al. 2006) and presumably are made from accreted material. But the phenomenon is common in barred and oval galaxies in which secular evolution is expected to be rapid. Besides NGC 4274 in Figure 3, excellent examples include NGC 3623 (SABa in the optical but clearly SB(r) in 2MASS JHK images: Jarrett et al. 2003), and NGC 5689 (SB0). These results are discussed in Ganda et al. (2006), Falcón-Barroso et al. (2006), Peletier et al. (2007a); see Peletier (2008) and Peletier et al. (2007b, c) for reviews. Quoting Peletier et al. (2007c): "SAURON observations show that 13 out of 24 Sa and Sab galaxies [and a similar fraction of late-type spirals] show a central local minimum in the velocity dispersion ... The sigma-drops are probably due to central disks that formed from gas falling into the central regions through a secular evolution process."
Figure 3. SAURON integral-field spectroscopy of the disky pseudobulge in the Sa galaxy NGC 4274 (adapted from Peletier et al. 2007c). The images show that NGC 4274 is a highly inclined barred galaxy; the bar is foreshortened, because it is oriented nearly along the minor axis. It fills an inner ring, as is normal in SB(r) galaxies (Kormendy 1979). The brightness profile (upper-right) is decomposed into a Sérsic (1968) function plus an exponential disk. The Sérsic function has n = 1.3, i. e., n < 2, as in other pseudobulges (Figure 5). The pseudobulge dominates the light at radii r 10". The kinematic maps (Falcón-Barroso et al. 2006) show that this light comes from a disky component that is more rapidly rotating (center), lower in velocity dispersion (right), and stronger in H line strength (left, from Peletier et al. 2007a) and hence younger than the rest of the inner galaxy.
Pseudobulges were also recognized photometrically in Kormendy (1993); progress since then has been rapid (KK04). In many galaxies, the "bulge" is essentially as flat as the disk and/or shows clearcut spiral structure. Both are signatures of high-density disks - classical bulges are dynamically hot and cannot have small-scale spiral structure. These features are spectacular in HST surveys of the centers of spiral galaxies (Carollo et al. 1997, 1998, 2001, 2002; Carollo 1999). Figure 4 shows examples. These are Sa - Sbc galaxies, so they should contain substantial bulges. Instead, their centers look like star-forming spiral galaxies. Contrast the definition of a classical bulge (Renzini 1999 following Sandage 1961): A bulge is nothing more nor less than an elliptical galaxy that happens to live in the middle of a disk.
Figure 4. Sa - Sbc galaxies with disky pseudobulges shown in 18" × 18" regions centered on the galaxy nucleus and extracted from Hubble Space Telescope (HST) WFPC2 F606W images kindly provided by Carollo et al. (1997, 1998). Displayed intensity is proportional to the logarithm of the surface brightness. Mean and minimum (pseudo)bulge-to-total luminosity ratios B / T observed by Simien & de Vaucouleurs (1986) are 0.35 and 0.13 for Sas, 0.22 and 0.10 for Sbs, and 0.18 and 0.05 for Sbcs. These galaxies contain nuclear star clusters but no E-like component with the above B / T values.
Classical and pseudo bulges can coexist (KK04; Kormendy et al. 2006; Erwin 2007), but the morphology in Figure 4 is not due to nuclear disks embedded in classical bulges that are hidden by the display parameters. Bulges have steep brightness profiles, so bulge light would dilute the contrast in the spiral structure very strongly at smaller radii. But the strength of the spiral structure depends little on radius: essentially all of the pseudobulge participates.
The imaging survey authors generally interpret disky bulges as consequences of secular evolution. Courteau, de Jong, & Broeils (1996) observe "spiral structure continuing into the central regions" and "invoke secular dynamical evolution and ... gas inflow via angular momentum transfer and viscous transport" as the explanation. Carollo et al. (2001) conclude that "exponential-type bulge formation is taking place in the local universe and that this process is consistent with being the outcome of secular evolution ... within the disks".
A "proof of concept" is largely in hand, so studies of secular evolution now concentrate on star formation (reviewed in Fisher & Drory 2008b) and on pseudobulge statistical properties. In Section 4, we compare the fundamental plane parameter correlations of pseudobulges, classical bulges, and ellipticals. First, we need to define what we mean by an elliptical. This leads to our second theme on environmental secular evolution and the formation of spheroidal galaxies (Section 3).