Bruce G. Elmegreen and Debra Meloy Elmegreen

Spiral galaxies are flattened stellar systems that rotate in a nearly circular fashion. They generally contain stars with a wide range of ages. The gas and youngest stars are in a flattened disk, older stars occupy a slightly thicker disk, and the oldest stars tend to reside in a more three-dimensional distribution, which may include a central bulge, a system of globular clusters, and a low-luminosity halo. This division of stars into distinct populations was first recognized by Walter Baade in 1944.

Spirals form when compressional waves propagate through the disk, growing in length and amplitude because of self-gravitational forces. The waves appear as spirals because the angular rotation rate of the disk varies approximately inversely with distance from the center. In many galaxies, spiral waves become so strong that they drive shock fronts into the gas, trigger new stars to form, and impart large random motions to the older disk stars at critical resonance points. Then the spiral structure can dominate the appearance, internal dynamics, and evolution of a galaxy.

The optical appearance of spiral galaxies is extremely diverse. A classification system was developed in the 1940s by Edwin Hubble and expanded in the 1960s by Allan Sandage, and by Gerard and Antoinette de Vaucouleurs. Along the Hubble sequence from type Sa to Sb, Sc and Sd, the bulge gets smaller and the spiral arms become more open. Two parallel sequences, SBa to SBd and SABa to SABd, describe galaxies with central bars and ovals, respectively. Because Hubble originally envisioned an evolutionary sequence, the Sa and Sb galaxies are termed early type and the Sc and Sd galaxies are termed late type. Early type galaxies contain proportionately more old stars and less gas and star-forming activity than late type galaxies; presumably the early type galaxies formed stars more rapidly in the past.

The spiral structure itself varies from galaxy to galaxy, even within a Hubble type. A classification system based on the regularity of the spiral arms was introduced in the 1960s by Sidney van den Bergh and revised in the 1980s by Debra and Bruce Elmegreen. Approximately 10% of spiral galaxies contain only a single, bisymmetric spiral that extends in a grand design from the edge of the galactic bulge to the outer limit of the perceptible disk. Most galaxies have a multiple-arm spiral structure, or even a highly chaotic or flocculent spiral-like structure. Multiple arm galaxies represent 60% of the early and intermediate Hubble types with bars and intermediate types without bars. Flocculent spirals represent 60% of early Hubble types without bars.

An example of a grand design galaxy, Messier 81, is shown in Fig. 1, rectified by computer to a face-on orientation and enhanced in detail by subtraction of the average radial light profile. The spiral pattern could be a quasistationary wave mode, as discussed by Chia-Chiao Lin, Giuseppe Bertin, and collaborators. Such a mode comprises both inward and outward propagating spiral wave packets. Interference between these wave packets may produce the gentle oscillation of the arm amplitudes seen in the figure. Wave modes are thought to grow by the transfer of energy and angular momentum from the inner to the outer region at the radius where the spiral pattern corotates with the disk. The trailing spiral components propagate away from this radius, outward and inward, until they either lose their energy by resonant interactions with the peculiar (noncircular) motions of stars or they reflect or refract off regions where a high stellar velocity dispersion makes propagation impossible. The energy-losing resonances are called Lindblad resonances, after Bertil Lindblad; they occur where the period of a star's random excursion around its circular orbit equals the time spent between the main spiral arms. The reflection point occurs in the inner region, near the bulge. After reflection, the wave returns to the corotation radius as either a leading wave packet or an open trailing wave packet; there it stimulates more of the tightly wound trailing components to reinforce the original trailing waves. Theory predicts that an initially small spiral perturbation should double in amplitude approximately once in each orbit time. Eventually the growth of the spiral wave mode is limited by energy dissipation in shocked gas and by stars scattering at the Lindblad resonances.

Figure 1

Figure 1. Blue band image of M81, rectified to a face-on orientation and with the average radial light profile subtracted. The calibration bar on the left represents 100 pixels; the bar on the right has a length equal to the isophotal radius at 25 magnitudes per square arc second. (From Elmegreen, Elmegreen, and Seiden, 1989. Reprinted with permission from The Astrophysical Journal.)

Highly symmetric spirals like those in Messier 81 can also form when a galaxy is strongly perturbed by a close encounter with another galaxy. This mechanism was illustrated in an atlas of peculiar galaxies by Halton Arp (1966) and it was modeled theoretically by Alar and Juri Toomre (1972). Tidal forces during such an encounter can perturb the stars so much that their previously near-circular orbits become elongated, especially in the outer parts. This elongation may be initially along a common line of apses, but under the influence of galactic shear, the elongations at different radii rotate at different rates, causing the stars to bunch together in a symmetric spiral pattern. Self-gravity then amplifies this pattern, and it propagates into the inner disk, possibly reflecting off the bulge. The outer tidal arms may also stimulate the growth of an inner spiral wave mode, which then persists for many more orbits than the initial tidal wave.

An example of a multiple-arm galaxy is Messier 74, shown with computer enhancement in Fig. 2. The symmetric spiral in the inner region is surrounded by numerous, apparently independent, long-arm spirals in the outer region. Theory suggests that the inner symmetric spiral could be a single wave mode or a constructive superposition of several wave modes with different angular speeds, or it could be a rapidly evolving wave packet. The outer spirals could be the constructively interfering parts of several overlapping wave modes, or they could be random and independent wave packets triggered by local gravitational instabilities in the gas and stars.

Figure 2

Figure 2. Blue band image of M74, as in Fig. 1. [By permission from G. B. Elmegreen, in Galactic Models, New York Academy of Sciences, New York (1990).]

A different theory, developed by Philip E. Seiden and Humberto Gerola in 1978, suggests that some of the outer structure in multiple-arm galaxies could result entirely from star formation. Gaseous instabilities first trigger star formation, and then supernova explosions and shell-like compressions following star formation trigger more star formation, leading to chain reactions that are swept back into spirals arcs by shear.

Flocculent galaxies contain many more, and much shorter, spiral arms than multiple arm galaxies. Such short arms could form by the same mechanisms as the longer arms in multiple arm galaxies, with the larger number of arms causing the chaotic appearance, or they could differ because of a lack of global wave stimulation, or an inability to amplify global waves in flocculent galaxies. An example is Messier 63 (not shown).

Bars and oval distortions may drive spiral waves in some galaxies. Among galaxies with early Hubble types, those with bars are twice as likely to have symmetric spirals in their outer disks as those without bars. Among intermediate and late-type galaxies, however, those with bars have approximately the same proportions of grand design, multiple arm, and flocculent spirals as those without bars. This dependence of the bar-spiral correlation on Hubble type may result from a variation of bar length and strength with Hubble type, that is, the bars in early-type galaxies tend to be larger and stronger than the bars in late-type galaxies. Perhaps early-type bars end near the corotation resonance, where they can drive outward-moving trailing waves that fill the outer disk. The bars in late-type galaxies may end far inside this resonance, possibly at an inner Lindblad resonance.

An important quantity to determine is the angular rate of the spiral pattern, which equals the angular rate of the stars at the corotation resonance, in the modal theory. Stars move faster than this pattern rate inside corotation and slower outside corotation. Recent work suggests that the angular pattern speed can sometimes be determined from optical tracers of wave-orbit resonances. In the most symmetric spirals, the far outer extent of the spiral should be the outer Lindblad resonance, where outward moving wave packets convert their energy into random stellar motions. The inner extent may be an inner Lindblad resonance if there is a bar or oval there, because such ovals can reflect an inward-moving wave. If the inner galaxy is symmetric, or if there is a strong bulge, then the inner extent of the spiral need not be a resonance but merely a reflection point. The corotation resonance occurs at the radius where gaseous shocks and wave-triggered star formation end, and in some cases, where large patches of interarm star formation occur. The 4:1 resonance, where the time spent between the main spiral arms equals twice the period of the peculiar stellar motions (rather than one times this period, as for a Lindblad resonance), may contain small spiral-like spurs midway between the main arms. Once any one resonance location is identified in a galaxy, all of the others follow from the radial distribution of circular velocities.

Considerable progress has been made in understanding galactic spirals, but neither the theoretical framework nor the observations of galaxies are sufficiently detailed at the present time to determine the origin and evolution of spirals on a case-by-case basis. Part of the problem is that the basic state of a galaxy is uncertain, that is, the spatial and temporal distributions of the gas and stars, and of their velocity dispersions, are poorly known, as are the magnitudes and histories of environmental perturbations. Progress requires detailed observations at optical, infrared, and radio wavelengths of all types of spirals in a variety of environments. Fully nonlinear theories, including gas and stars, with or without external perturbations, are also needed.

Additional Reading
  1. Arp, H. (1966). Atlas of peculiar galaxies. Ap. J. Suppl. 14 1.
  2. Athanassoula, E. (1984). The spiral structure of galaxies. Phys. Rep. 114 321.
  3. Baade, W. (1944). The resolution of Messier 32, NGC 205, and the central region of the Andromeda nebula. Ap. J. 100 137.
  4. Bertin, G. (1980). On the density wave theory of normal spiral galaxies. Phys. Rep. 61 1.
  5. de Vaucouleurs, G., de Vaucouleurs, A., and Corwin, H. G., Jr. (1976). Second Reference Catalog of Bright Galaxies. University of Texas Press. Austin.
  6. Elmegreen, B.G. (1989). Grand design, multiple arm and flocculent spiral galaxies. In Galactic Models, J.R. Buchler, S.T. Gottesman, and J.H. Hunter, eds. Academy of Science New York.
  7. Elmegreen, B.G., Elmegreen, D.M., and Seiden, P.E. (1989). Spiral arm amplitude variations and pattern speeds in the grand design galaxies M51, M81, and M100. Ap. J. 343 602.
  8. Elmegreen, D.M. and Elmegreen, B.G. (1987). Arm classes for spiral galaxies. Ap. J. 314 3.
  9. Gerola, H. and Seiden, P.E. (1978). Stochastic star formation and the structure of galaxies. Ap. J. 223 129.
  10. Hubble, E.P. (1926). Extra-galactic nebulae. Ap. J. 26 321.
  11. Lin, C.C. and Shu, F.H. (1964). On the spiral structure of disk galaxies. Ap. J. 140 646.
  12. Lindblad, B. (1958). Stockholm Observatory Annals 20 No. 6.
  13. Sandage, A. (1961). The Hubble atlas of galaxies. Publication 618, Carnegie Institute of Washington.
  14. Toomre, A. (1977). Theories of spiral structure. Ann. Rev. Astron. Ap. 15 437.
  15. Toomre, A. and Toomre, J. (1972). Galactic bridges and tails. Ap. J. 178 623.
  16. van den Bergh, S. (1960). A preliminary luminosity classification of late type galaxies. Ap. J. 131 215.
  17. See also Galactic Structure, Large Scale; Galactic Structure, Spiral, Observations; Galaxies, Spiral, Nature of Spiral Arms.