1
The Classification of Galaxies

Ronald BUTA

Astrophysics, University of Alabama, Tuscaloosa, USA

The classification of the forms of galaxies in a well-defined visual system is a critical step in the study of galaxies as physical objects. The Comprehensive de Vaucouleurs revised Hubble-Sandage (CVRHS) system is currently the most detailed approach that can be applied effectively to more than 95% of all galaxies. This chapter describes the different types of galaxies and the factors that may determine various morphological features.

1.1. Introduction

Galaxies are complex gravitational systems whose structure has been influenced not only by how they formed but also by the environment into which they were born. A century ago, getting a classifiable image of a single galaxy was a major effort involving long exposures of photograph plates. Today, there are classifiable images of literally millions of galaxies available through the Internet. Although it is not obvious how any galaxy arrived at its current morphological state, examination of the details of large numbers of galaxies have led to important physical insights into the roles played by both internal and external processes. It is for this reason that classical galaxy morphology and classification have survived into the modern era.

Galaxy morphology and classification began with simple descriptive terms for angular size, brightness and central concentration based on visual observations (e.g. Dreyer 1888). Although large 19th-Century reflectors did reveal through visual observation some genuine aspects of galaxy morphology, it was photography that led to the classification systems of the early-to-mid 20th Century, including those of Wolf (1908), Reynolds (1920), Hubble (1926, 1936), Lundmark (1926) and Morgan (1958). On the plates of the time, which were relatively more sensitive to blue light than to red light, details of spiral arms, disks and bulges could be used as classification criteria. In spite of great observational and theoretical progress in extragalactic studies during the past century, modern galaxy classification is still basically tied to the system proposed by Hubble (1936; Figure 1.1), only now the system is applied using digital images rather than photographic plates. The main reason the Hubble system has survived for nearly a century is that the aspects Hubble focused on (degree of central concentration, pitch angle and resolution of spiral arms, amount of distinct structure) correlated with measured properties of galaxies, such as luminosities, colors, stellar populations, HI content and global star formation history. This gave his view an astrophysically relevant edge that has driven much of extragalactic research since his time.

Figure 1.1. The Hubble (1936) "tuning fork" representation of galaxy morphology

This chapter describes the different classes of galaxies within the framework of the "Comprehensive de Vaucouleurs revised Hubble-Sandage" (CVRHS) classification system, a visual system that follows the precepts of de Vaucouleurs (1959), who proposed a personal revision of the Hubble-Sandage (HS) classification (Sandage 1961; Figure 1.2) that he believed provided a better description of galaxies without being too unwieldy. For many galaxies, the CVRHS classification is no more complicated than a de Vaucouleurs (1959) VRHS classification. However, the system has been designed to take into account more details that are of astrophysical interest today and which have become more noticeable and relevant in the era of digital astronomical imaging. These details include lenses, nuclear rings and bars, ansae bars, boxy/peanut bulges, boxy and disky elliptical galaxies, special outer rings and pseudorings, dust lanes and galactic disk warps.

Figure 1.2. The Hubble-Sandage (Sandage 1961) revised "tuning fork" representation of galaxy morphology, including stages of S0 galaxies

1.2. Classes of galaxies

Hubble recognized that there are basically two classes of galaxies: disk-shaped galaxies and non-disk-shaped galaxies. In a disk galaxy, the structure is dominated by a highly flattened stellar component. Within this disk, other structures may be seen, such as spiral arms, bars, rings, a central bulge and extensive distributions of interstellar gas and dust. The way these structures are seen depends on the inclination of the plane of the disk to our line of sight. For a given galaxy, we say the inclination i is 0° when the disk is seen face-on, and 90° when the disk is seen edge-on. Disk planes are randomly oriented to the line of sight (meaning they are uniformly distributed in sin i), which complicates the interpretation of highly inclined cases. Even a century ago, disk-shaped galaxies were known to be more common than non-disk galaxies to the point that the latter were considered of greater interest (Keeler 1899).

Although not all disk-shaped galaxies are spiral, the typical disk galaxy is a spiral, where luminous, outwardly winding arcs of stars and often star-forming regions form a major part of the morphology. Classic nearby spirals such as M51, M81, and M101 initially fueled the misperception that spirals are generally regular systems having only a bulge in addition to the disk and spiral arms. It was thought by Hubble that barred spirals are significantly less abundant than non-barred spirals, or what he called "normal spirals". It is now known that barred spirals are at least as abundant as non-barred spirals, and that some galaxies that appear to be non-barred in blue light can appear to be barred when imaged at infrared wavelengths (Eskridge et al. 2000).

The existence of non-disk-shaped (and therefore non-spiral) galaxies was at first somewhat controversial. Based on the limited plate material available in his day, Curtis (1918) believed that all galaxies were spiral, and any that did not appear to be spiral on his plates would be found to be spiral when observed with larger telescopes. Hubble, having access to better telescopes, disagreed with this conclusion and believed that genuine non-spiral galaxies existed. Hubble recognized such galaxies as elliptical galaxies, where the luminosity distribution is defined by a regular decline from a bright center to the faint outer regions. Ellipticals were thought to be characterized by no other features but their isophotal shapes, which ranged from round (E0) to a flattening approaching that of disk-shaped galaxies (E7). In fact, Hubble believed the sequence of E galaxy shapes blended smoothly into the domain of disk-shaped galaxies, which he split into two parallel sequences of normal and barred spirals characterized by the degree of central concentration, the degree of resolution of the spiral arms into what were likely to be star-forming regions and the degree of openness of the spiral arms. Hubble (1936) illustrated these features in his famous "tuning fork" of galaxy morphologies (Figure 1.1). Although this view is now obsolete, the tuning fork is still an effective way of binning galaxies into astrophysically meaningful classes.

In stellar astronomy, temporal terms are often used to describe certain kinds of stars. Stars of spectral classes O, B and A tend to be young and are known as "early-type" stars, while those of spectral classes K and M tend to be older and are called "late-type" stars. Those of spectral classes F and G are known as "intermediate-type" stars. Hubble found it convenient to use similar terms for galaxies, calling galaxies on the left part of the tuning fork early-type galaxies and those on the right part late-type galaxies. In general, E and S0 galaxies are said to be early-type galaxies, S0/a and Sa galaxies are called early-type spirals, Sc-Sm galaxies are called late-type spirals and Sab-Sbc galaxies are called intermediate-type spirals. Unlike for stars, no real temporal meaning was to be implied by these terms.

The value of the Hubble classification system, as well as its limitations, can only be appreciated by examining large numbers of galaxies and attempting to classify them within that system. The small number of astronomers who actually did this, such as Sandage, de Vaucouleurs, van den Bergh and Morgan, all saw the need for modifications or even alternative views, which led to major revisions of the system. This article uses images from the EFIGI project (Baillard et al. 2011; de Lapparent et al. 2011) to illustrate galaxies of different types within the framework of the CVRHS classification system (Buta et al. 2015). The classifications are from Buta (2019).

Figure 1.3. The de Vaucouleurs (1959) revised Hubble-Sandage (VRHS) system of galaxy classification, where the Hubble "tuning fork" has become a classification volume. Top: The VRHS stage sequence. Bottom left: The cross-section of families and varieties. Bottom right: Examples of families and varieties at stage Sbc (from Buta et al. 2007)

The basic outline of the HS system (Sandage 1961) is depicted in Figure 1.2 and that for the VRHS system is depicted in Figure 1.3. In the HS system, Sandage added the (r) and (s) subtypes, which are also used in the VRHS. The HS system still depicts galaxy morphology as a "tuning fork", but is more complicated than Hubble's original tuning fork (Figure 1.1). In the VRHS, the tuning...