Below [Fe/H] = −3.0, there is an enormous range in [C/Fe]. We discuss the properties of C-rich ([C/Fe] > +0.7) and C-normal ([C/Fe] ≤ +0.7) stars in this regime, and suggest that there existed two different gas cooling channels in the very early Universe.

I present an evaluation of the kinematic properties of halo red giants thought to have formed in globular clusters based on the strength of their UV/blue CN and CH absorption features. The sample has been selected from the catalog of Martell et al. (2011). The orbital parameters of CN-strong halo stars are compared to those of the inner and outer halo populations, and to the orbital parameters of globular clusters with well-studied Galactic orbits. It has been found that both the clusters and the CN-strong field stars exhibit kinematic and orbital properties similar to the inner halo population, indicating that globular clusters could be a significant source of inner halo field stars, and suggesting that both globular clusters and CN-strong stars could belong primarily to the inner halo population of the Milky Way.

Globular Clusters and metal poor stars represent two aspects of the population II stars in our Galaxy. The presentation will focus on the metal poor end of population II of galactic field stars.

I discuss a search for red giant stars in the Galactic halo with light-element abundances similar to second-generation globular cluster stars, and discuss the implications of such a population for globular cluster formation models and the balance between in situ star formation and accretion for the assembly of the Galactic halo.

During the past decade observations across the electromagnetic spectrum have led to broad progress in the understanding of galaxy clusters and their far more abundant smaller siblings, groups. From the X-rays, where Chandra and XMM have illuminated old phenomena such as cooling cores and discovered new ones such as shocks, cold fronts, bubbles and cavities, through rich collections of optical data (including vast and growing arrays of redshifts), to the imaging of AGN outbursts of various ages through radio observations, our access to cluster and group measurements has leaped forward, while parallel advances in theory and modeling have kept pace.

This Special Session offered a survey of progress to this point, an assessment of outstanding problems, and a multiwavelength overview of the uses of the next generation of observatories. Holding the symposium in conjuction with the XXVIIIth General Assembly provided the significant advantage of involving not only a specialist audience, but also interacting with a broad cross-section of the world astronomical community.

The meeting “Galaxy Evolution through Secular Processes,” hosted as Special Session 3 at the XVIIIth General Assembly of the IAU in Beijing, brought together a broad community of astronomers actively working on topics related to the evolution of galaxies. This evolution is currently thought to have two phases: (1) a formative chaotic phase, where baryonic matter collected into dark halo “seeds” with the occurrence of frequent mergers; and (2) a slower secular evolutionary phase where internal perturbations such as bars and spirals interacted (and likely still interact) with a galaxy's stars and gas clouds to redistribute material and slowly change the morphology, such as growing bulges or pseudobulges from disk material.

This Special Session is devoted to the secular evolution of disk galaxies. Here ‘secular’ means ‘slow’; i.e., evolution on time scales that are generally much longer than the galaxy crossing or rotation time. Internal and environmentally driven evolution both are covered.

I am indebted to Albert Bosma for reminding me at the 2011 Canary Islands Winter School on Secular Evolution that our subject first appeared in print in a comment made by Ivan King (1977) in his introductory talk at the Yale University meeting on The Evolution of Galaxies and Stellar Populations: ‘John Kormendy would like us to consider the possibility that a galaxy can interact with itself.. . . I'm not at all convinced, but John can show you some interesting pictures.’ Two of the earliest papers that followed were Kormendy (1979a, b); the first discusses the interaction of galaxy components with each other, and the second studies these phenomena in the context of a morphological survey of barred galaxies. The earliest modeling paper that we still use regularly is Combes & Sanders (1981), which introduces the now well known idea that box-shaped bulges in edge-on galaxies are side-on, vertically thickened bars.

It is gratifying to see how this subject has grown since that time. Hundreds of papers have been written, and the topic features prominently at many meetings (e.g., Block et al. 2004; Falcoń-Barroso & Knapen 2012, and this Special Session). My talk here introduces both internal and environmental secular evolution; a brief abstract follows. My Canary Islands Winter School review covers both subjects in more detail (Kormendy 2012). Kormendy & Kennicutt (2004) is a comprehensive review of internal secular evolution, and Kormendy & Bender (2012) covers environmental evolution. Both of these subject make significant progress at this meeting.

Secular evolution happens because self-gravitating systems evolve toward the most tightly bound configuration that is reachable by the evolution processes that are available to them. They do this by spreading – the inner parts shrink while the outer parts expand. Significant changes happen only if some process efficiently transports energy or angular momentum outward. The consequences are very general: evolution by spreading happens in stars, star clusters, protostellar and protoplanetary disks, black hole accretion disks and galaxy disks. This meeting is about disk galaxies, so the evolution most often involves the redistribution of angular momentum.

We now have a good heuristic understanding of how nonaxisymmetric structures rearrange disk gas into outer rings, inner rings and stuff dumped onto the center. Numerical simulations reproduce observed morphologies very well. Gas that is transported to small radii reaches high densities that are seen in CO observations. Star formation rates measured (e.g.) in the mid-infrared show that many barred and oval galaxies grow, on timescales of a few Gyr, dense central ‘pseudobulges’ that are frequently mistaken for classical (elliptical-galaxy-like) bulges but that were grown slowly out of the disk (not made rapidly by major mergers). Our resulting picture of secular evolution accounts for the richness observed in morphological classification schemes such as those of de Vaucouleurs (1959) and Sandage (1961). State-of-the art morphology discussions include the de Vaucouleurs Atlas of Galaxies (Buta et al. 2007) and Buta (2012, 2013).

Pseudobulges as disk-grown alternatives to merger-built classical bulges are important because they impact many aspects of our understanding of galaxy evolution. For example, they are observed to contain supermassive black holes (BHs), but they do not show the well known, tight correlations between BH mass and host properties (Kormendy et al. 2011). We can distinguish between classical and pseudo bulges because the latter retain a ‘memory’ of their disky origin. That is, they have one or more characteristics of disks: (1) flatter shapes than those of classical bulges, (2) correspondingly large ratios of ordered to random velocities, (3) small velocity dispersions σ with respect to the Faber-Jackson correlation between σ and bulge luminosity, (4) spiral structure or nuclear bars in the ‘bulge’ part of the light profile, (5) nearly exponential brightness profiles and (6) starbursts. None of the above classification criteria are 100% reliable. Published disagreements on (pseudo)bulge classifications usually result from the use of diffferent criteria. It is very important to use as many classification criteria as possible. When two or more criteria are used, the probability of misclassification becomes very small.

I also review environmental secular evolution – the transformation of gas-rich, star-forming spiral and irregular galaxies into gas-poor, ‘red and dead’ S0 and spheroidal (‘Sph’) galaxies. I show that Sph galaxies such as NGC 205 and Draco are not the low-luminosity end of the structural sequence (the ‘fundamental plane’) of elliptical galaxies. Instead, Sph galaxies have structural parameters like those of low-luminosity S+Im galaxies. Spheroidals are continuous in their structural parameters with the disks of S0 galaxies. They are bulgeless S0s. S+Im → S0+Sph transformation involves a variety of internal (supernova-driven baryon ejection) and environmental processes (e.g., ram-pressure gas stripping, harassment, and starvation). Improved evidence for galaxy transformation is presented in several papers at this meeting.

Gravity-bound isolated systems, from stars, planetary systems, star clusters to galaxies, share common properties where evolution is the rule. Typically if they start forming at a well defined epoch they tend to change significantly over a timescale comparable to their present age. So evolution is never truly stopped, it just proceeds slower and slower: after a rapid, violent phase a slower, secular phase follows. In galactic astronomy for many decades the paradigm was rather that after a short violent time galaxies would settle in a stable steady state just consuming gas into stars. Actually today it appears that the progressive appearance of galaxy systematic morphologies and the slowing pace of mergers indicate that common intrinsic dynamical factors continue to shape galaxies towards similar properties irrespective of their largely different formation histories and initial conditions. Newtonian physics supplemented by a weakly dissipative component provides an amazing amount of explanations for the galaxy properties, like exponential stellar disks, spirals, bars, and peanut-shaped bulges. The purpose of this talk is to review these mechanisms of dynamical secular evolution.

During the lifetime of a galaxy, secular radial mass redistribution is expected to gradually build up a bulge and transform the Hubble type from late to early. The dominant dynamical process responsible for this transformation is a collective instability mediated by density-wave collisionless shocks (Zhang 1996, 1998, 1999). The ability of this new mechanism to secularly redistribute the STELLAR mass provides a general pathway for the formation and evolution of the majority of Hubble types, ranging from late type disk galaxiess to disky ellipticals. ATLAS3D results (Cappellari et al. 2013) showed that spirals and S0s and disky ellipticals form a continuous trend of evolution which also coincides with the aging of the stellar population of galactic disks. The importance of stellar accretion is also revealed in the results of the COSMOS team which showed that the evolution of the black-hole-mass/bulge-mass correlation since z = 1 was mainly due to the mass redistribution on pre-existing STELLAR disks which were already in place by z = 1 (Cisternas et al. 2011). The weaker correlation between the masses of late-type bulges and AGNs observed at any given epoch in our view is a result of the quicker initial onset of accretion events in AGN disks compared to that in galactic disks, since the dynamical timescale is shorter for smaller AGN accretion disks.

The same secular dynamical process can produce and maintain the well-known scaling relations and universal rotation curves of observed galaxies during their Hubble-type transformation (Zhang 2008), as well as reproduce many other observed structural and kinematic properties of galaxies such as the size-line-width relation of the interstellar medium and the age-velocity dispersion relation of solar neighborhood stars in our own Galaxy. A by-product of this analysis is a powerful new method for locating the multiple corotation resonances in galaxies (Zhang & Buta 2007; Buta & Zhang 2009).

The current work also highlights the connection between collective effects in galactic dynamics and nonequilibrium phase transition processes in other branches of physics such as fluid turbulence and spontaneous breaking of gauge symmetry in high-energy physics. The continuous build-up of the Hubble sequence of galaxies through secular mass accretion also hints at the baryonic nature of galactic dark matter and poses challenges to the existing LCDM paradigm, since the well-known adiabatic compression process during baryonic mass inflow produced by secular evolution would lead to a concentration of the cold dark matter to the central region of early-type galaxies, which is not observed.

Resonances play an important role in the evolution of the disks of spiral galaxies, and in particular in the chemical abundance evolution. The dominant effect is that of corotation; this effect can be even used as a tool to estimate the age of the present spiral arm pattern, which are usually found to be long-lived, contrary to a recent common belief. We investigated a sample of galaxies for which the corotation radius is known and for which there are available in the literature measurements of abundance gradients for Oxygen. A very good correlation is found between corotation radii and the radii at which there is a break in the slope of the gradients. The gradients are usually decreasing in the inner regions and become flat or rising at larger radii. In several galaxies, including the Milky Way, one observes not only a change in the slope of the abundance gradient, but also an abrupt step in metallicity, at corotation. This step is due to the fact that corotation separates the disk of a galaxy in two regions (inside corotation and outside corotation) which are isolated one from the other, so that the two sides evolve in an independent way. The barrier between the two regions is produced by the flow of gas in opposite directions in the two sides and by the ring-shaped void of gas observed at corotation. Besides this, an independent effect of corotation is a minimum of star formation associated with the minimum velocity at which the spiral arms (seen as potential wells) are fed with interstellar gas. Still another effect is the scattering of stars by the resonance, which causes their migration to different galactic radii. Other resonances, like 4:1, have properties almost opposite to corotation; they stimulate star-formation, and tend to gather the stars in the resonant orbit, instead of scattering them out, as shown by numerical simulations. Due to this property, one can see arms which have the shape of resonant stellar orbits, which depart from logarithmic spirals.

Simulations of isolated galaxy disks that are stable against bar formation readily manifest multiple, transient spiral patterns. It therefore seems likely that some spirals in real galaxies are similarly self-excited, although others are clearly driven by tidal interactions or by bars. The rapidly changing appearance of simulated spirals does not, however imply that the patterns last only a fraction of an orbit. Power spectrum analysis reveals a few underlying, longer-lived spiral waves that turn at different rates, which when super-posed give the appearance of swing-amplified transients. These longer-lived waves are genuine unstable spiral modes; each grows vigorously, saturates and decays over a total of several orbit periods. As each mode decays, the wave action created as it grew drains away to the Lindblad resonances, where it scatters stars. The resulting changes to the disk create the conditions for a new instability, giving rise to a recurring cycle of unstable modes.

Disc galaxies have a number of structures, such as bars, spirals, rings, discy bulges, m = 1 asymmetries, thick discs, warps etc. I will summarise what is known about their origin and in particular whether it is due to an external or an internal process. The former include interactions, major or minor mergers etc, while the latter include instabilities, or driving by another component of the same galaxy, as e.g. the bar or the halo. In cases where more than one process is eligible, I will analyse whether it is possible to distinguish between different origins, and what it would take to do so. This discussion will show that, at least in some cases, it is difficult to distinguish between an internal and an external origin.

Based both on observations and simulations, recent works propose that the speed of the spiral pattern in disk galaxies may decrease with increasing radius; the implications are that patterns are actually short-lived, and that the azimuthal color/age gradients across spiral arms predicted by density wave theory could not be produced. We, however, have consistently found such gradients, and measured spiral pattern speeds by comparing the observations with stellar population synthesis models (González & Graham 1996; Martínez-García et al. 2009a, b; Martínez-García & González-Lópezlira 2011). Here, we summarize our previous results in non-barred and weakly barred spirals, together with six new, as yet unpublished, objects. On the other hand, we have indeed found a trend whereby pattern speeds at smaller radii are larger than expected from a model that assumes purely circular orbits (cf. Figure 1), likely due to the effect of spiral shocks on the orbits of newborn stars. The results suggest that spirals may behave as steady long-lived patterns.

We use visual classifications of the brightest 250,000 galaxies in the Sloan Digital Sky Survey Main Galaxy Sample provided by citizen scientists via the Galaxy Zoo project (www.galaxyzoo.org, Lintott et al. 2008) to identify a sample of local disc galaxies with reliable bar identifications.

These data, combined with information on the atomic gas content from the ALFALFA survey (Haynes et al. 2011) show that disc galaxies with higher gas content have lower bar fractions.

We use a gas deficiency parameter to show that disc galaxies with more/less gas than expected for their stellar mass are less/more likely to host bars. Furthermore, we see that at a fixed gas content there is no residual correlation between bar fraction and stellar mass. We argue that this suggests previously observed correlations between galaxy colour/stellar mass and (strong) bar fraction (e.g. from the sample in Masters et al. 2011, and also see Nair & Abraham 2010) could be driven by the interaction between bars and the gas content of the disc, since more massive, optically redder disc galaxies are observed to have lower gas contents.

Furthermore we see evidence that at a fixed gas content the global colours of barred galaxies are redder than those of unbarred galaxies. We suggest that this could be due to the exchange of angular momentum beyond co-rotation which might stop a replenishment of gas from external sources, and act as a source of feedback to temporarily halt or reduce the star formation in the outer parts of barred discs.

These results (published as Masters et al. 2012) combined with those of Skibba et al. (2012), who use the same sample to show a clear (but subtle and complicated) environmental dependence of the bar fraction in disc galaxies, suggest that bars are intimately linked to the evolution of disc galaxies.

Bars serve a crucial signpost in galaxy evolution because they form quickly once a disk is sufficiently massive and dynamically cold. Although the bar fraction in the local Universe is well-established since the mid-60s, a variety of studies have concluded varying bar fractions due to different definitions of bars, use of low quality data or different sample selection. The Spitzer Survey of Stellar Structure in Galaxies (S4G) offers us the ideal data set for resolving this outstanding issue once and for all. S4G consists of over 2000 nearby galaxies chosen based on optical brightness, distance, galactic latitude and size in a 40 Mpc volume. With a 4 minute integration time per pixel over >1.5 × D25 diameter for each galaxy, the data provide the deepest, homogenous, mid-infrared (3.6 and 4.5 microns) data on the nearby Universe. The data are so deep that we are tracing stellar surface densities << 1 solar mass per square parsec. With these data we can confidently constrain the bar fraction and thus shed important light on the evolutionary state of galaxies as a function of mass, environment and other galaxy host properties.

Rings are common in galaxies. Several kinds of rings are known: collisional, polar, and resonance rings, of which the latter is by far most common. Resonance rings are prime tracers of the underlying dynamical structure of disk galaxies, in particular of orbital resonances and of manifolds. Rings are also indicators of angular momentum transport, and this is a key factor in secular evolution (see the various reviews in Falcón-Barroso & Knapen 2012).

Resonance rings come in three flavours, primarily defined by their size, namely nuclear, inner, and outer rings. From studies like those of Buta (1995), Knapen (2005) and Comerón et al. (2010, 2013) we know that the radii of nuclear rings range from a few tens of parsec to some 3.5 kpc, while inner rings and outer rings have typical radii of 1.2 and 2.5–3 times the length of the bar. Many host galaxies of rings are barred, but so are most galaxies in general. Some 20% of all rings occur in non-barred galaxies, which implies that rings do not, or hardly, occur preferentially in barred galaxies (Knapen 2005, Comerón et al. 2010, 2013). In most non-barred ringed galaxies an oval, a past interaction, or even a prominent spiral pattern lies at the dynamical origin of the ring, but this needs additional scrutiny.

From an inventory of all known nuclear rings, Comerón et al. (2010) reach the following conclusions. Star-forming nuclear rings occur in 20 ± 2% of disk galaxies with −3 < T < 7; 18/96 occur in disk galaxies without a bar (19%); they are found in S0 to Sd galaxies, peaking in types Sab Sb; when nuclear rings occur in barred galaxies, the ring radius is limited to one quarter of the bar radius; and stronger bars host smaller rings (cf. Knapen 2005).

We are now using the Spitzer Survey of Spiral Structure in Galaxies (S4G; Sheth et al. 2010) to expand our survey to inner and outer rings (Comerón et al. 2013). We aim to study the relations between ring and host properties – as we did before for nuclear rings. We will use the S4G sample size and image depth to reach further insight into the secular evolution of galaxies by measuring structural properties of rings, as well as those of components like bars and disks. We will then be able to tackle outstanding questions such as the origin of rings in non-barred galaxies, and how exactly ring properties are determined by the bar.

Bars are the most important driver of secular evolution. A significant fraction of barred galaxies also harbor small secondary bars. Secondary bars are visible even in near-infrared images, so they are not just dusty and blue, but stellar features (Erwin & Sparke 2002). Since they are quite common, secondary bars are probably long-lived stellar features. The random relative orientation of the two bars indicates that they are dynamically decoupled with different pattern speeds (Buta & Crocker 1993). Corsini et al. (2003) presented conclusive direct kinematic evidence for a decoupled secondary bar in NGC 2950. Dynamically decoupled secondary bars have long been hypothesized to be a mechanism to drive gas past the ILR of primary bars to feed active galactic nuclei (Shlosman et al. 1989). However, the dynamics of secondary bars are still not well understood, and it is still unclear what role secondary bars play in the AGN fueling process.

Numerical simulations offer the best approach to understanding double-barred systems. Decoupled secondary bar in the earlier gaseous simulations only last a short time (< 1 Gyr, e.g. Friedli & Martinet 1993). Orbital studies of double-barred systems discovered a family of loop orbits that may be building blocks of long-lived nuclear stellar bars (Maciejewski & Sparke 1997, 2000). To complement orbital studies, which are not fully self-consistent, N-body simulations are preferred to further our understanding of double-barred systems. Debattista & Shen (2007) and Shen & Debattista (2009) managed to form long-lived double-barred systems with purely collisionless simulations, where a pre-existing rotating pseudo-bulge is introduced initially. The shape and size of secondary bars in the models are comparable to observed ones. They found that the rotation of the two bars is not rigid. The amplitude and pattern speed of the secondary bars oscillate as they rotate through their primary counterparts. Although the secondary bar rotates faster than the primary bar in this model, the stellar velocity field in the central region only shows a weakly twisted kinematic minor axis.

Recently more simulations of double-barred galaxies with simpler initial conditions are explored (Du, Shen & Debattista 2014). We expect that the new models can be used to cross-check with the kinematic properties of double-barred galaxies from IFU observations such as SAURON and Atlas3D.

We present a study of the kinematics of a sample of isolated spiral galaxies in the Spitzer Survey of Stellar Structure in Galaxies (S4G). We use Hα Fabry-Perot data from the GHαFaS instrument at the William Herschel Telescope (WHT) in La Palma, complemented with images at 3.6 microns, in the R band and in the Hα filter. The resulting data cubes and velocity field maps allow a complete study of the kinematics of a galaxy, including in-depth investigations of the rotation curve, velocity moment maps, velocity residual maps, gradient maps and position-velocity (PV) diagrams. We find clear evidence of the secular evolution processes going on in these galaxies, such as asymmetries in the velocity field in the bar zone, and non-circular motions, probably in response to the potential of the structural components of the galaxies, or to past or present interactions.

Bar driven secular evolution plays a key role in changing the morphology and kinematics of disk galaxies, leading to the formation of rapidly rotating boxy/peanut bulges. If these disk galaxies also hosted a preexisting classical bulge, how would the secular evolution influence the classical bulge, and also the observational properties.

We first study the co-evolution of a bar and a preexisting non-rotating low-mass classical bulge such as might be present in galaxies like the Milky Way. It is shown with N-body simulations that during the secular evolution, such a bulge can gain significant angular momentum emitted by the bar through resonant and stochastic orbits. Thereby it transforms into a cylindrically rotating, anisotropic and triaxial object, embedded in the fast rotating boxy bulge that forms via disk instability (Saha et al. 2012). The composite boxy/peanut bulge also rotates cylindrically.

We then show that the growth of the bar depends only slightly on the rotation properties of the preexisting classical bulge. For the initially rotating small classical bulge, cylindrical rotation in the resulting composite boxy/peanut bulge extends to lower heights (Saha & Gerhard 2013). More massive classical bulges also gain angular momentum emitted by the bar, inducing surprisingly large rotational support within about 4 Gyrs (Saha et al. in prep).

Since Edwin Hubble introduced his famous tuning fork diagram more than 70 years ago, spiral galaxies and early-type galaxies (ETGs) have been regarded as two distinct families. The spirals are characterized by the presence of disks of stars and gas in rapid rotation, while the early-types are gas poor and described as spheroidal systems, with less rotation and often non-axisymmetric shapes. The separation is physically relevant as it implies a distinct path of formation for the two classes of objects. I will give an overview of recent findings, from independent teams, that motivated a radical revision to Hubble's classic view of ETGs. These results imply a much closer link between spiral galaxies and ETGs than generally assumed.