ARlogo Annu. Rev. Astron. Astrophys. 1984. 22: 37-74
Copyright © 1984 by Annual Reviews. All rights reserved

Next Contents Previous

2.2 Masses and Kinematics

Global dynamics for large samples of Irrs have been derived from HI 21-cm line surveys (53, 109, 110, 135, 231, 241, 350, 351, 352). These observations provide integrated Doppler line widths, and usually the full width at 20% of peak intensity is interpreted as a good estimate of 2V'c sin i, where V'c is a characteristic circular gas velocity in a galaxy disk of inclination i. The distribution of HI line velocity widths shown in Figure 2 demonstrates that the majority of Magellanic Irrs have peak orbital velocities of Vc < 100 km s-1, with Vc ~ 50-70 km s-1 a representative value, in agreement with Brosche (41). Compared with spirals, for which Vc gtapprox 200 km s-1 is typical (41, 102, 287, 374), Magellanic Irrs are slow rotators and have low specific angular momenta (ltapprox 0.1 of the solar neighborhood value).

Figure 2

Figure 2. Distribution of HI line Doppler full widths at 20% of peak intensity for Im and Sm galaxies in the Fisher & Tully (110) survey that meet the conditions of (a) redshift velocity less than 1200 km s-1 and (b) estimated inclination of 50° or larger. Also shown are positions of several Irrs with known rotation curves based on references in the text (which are plotted at twice the maximum circular velocity) and the mean location of Scd galaxies [from Faber & Gallagher (102)]. Irr galaxies rotate more slowly than spirals, and virtually all systems with W20 180 km s-1 are transition spirals, while for W20 < 100 km s-1 pure irregular morphologies greatly predominate.

Dynamical masses within the optical radii Ropt of galaxies are normally calculated on the basis of spherical mass distributions M approx (Vc2 Ropt / G). This may lead to mass overestimates by a factor of two for Irr galaxies, which are disk dominated with low central mass concentrations. Since Magellanic Irrs have both smaller Vc and Ropt values than archetypal spirals, their masses are considerably smaller, with values of 108-1010 Msun typical of survey results. Dynamical masses for individual galaxies are subject to a variety of uncertainties [e.g. inclinations are difficult to estimate for barred, chaotic galaxies (351, 352), velocity fields may be complex (376), and warps in the gas disk may not be uncommon (326)]. Even so, the mass range (but perhaps not the mass distribution function) derived from surveys of Magellanic Irrs should be reliable. Mass-to-blue-light ratios found from the HI surveys mainly scatter between 2 and 10 for H0 = 100 km s-1 Mpc-1. As a fiducial point, the LMC has a mass of 5 ± 1 x 109 Msun (105), which yields a value of M / LB = 1.6 ± 0.2. There is then some justification for believing that the M / LB values based on HI survey results are systematically too high.

Internal kinematics have been measured in several nearer Irrs using accurate HI velocity maps derived from pencil-beam observations (181, 182, 183, 286; earlier references in 374; rotation curve data are summarized in 18) and from HI aperture synthesis studies (5, 6, 7, 21, 67, 136, 137, 227, 303, 326, 360a, 373, 377). Both of these methods suffer to some degree from modest angular resolution. Velocities from optical emission lines can provide good spatial resolution and, in some cases, good velocity resolution but are sensitive to motions induced in the ionized gas by young stars (54, 78, 80, 105, 134, 188, 189, 236).

From these observations we can identify an underlying unity to the properties of Magellanic Irrs. Cool gas is located in uniformly rotating disks. The dispersion velocity of HI in the disks is typically ~ 10 ± 2km s-1, a value that appears to be universal in HI disks of all galaxies (7, 182, 183, 375) and must therefore be an intrinsic property of interstellar matter rather than a property of specific galaxies. Rotation velocities within the gas disks, on the other hand, are quite sensitive to galaxy structure, with Vc declining from ~ 100 km s-1 in giant Irrs to virtually undetectable levels in extreme dwarfs and intergalactic HII regions. Among the very luminous clumpy Irrs and related blue compact galaxies, rotation velocities approach those of normal spirals (12, 53, 135), and thus these systems originate from a rather different state than the common Magellanic Irrs.

Mean rotation curves of Irrs also differ in form from those of spirals. Near-rigid-body rotation extends over most of the optical dimensions, with shallow velocity gradients of dV / dr ~ 5-20 km s-1 kpc-1 that reach peak velocities near the optical peripheries of the systems. In spirals the rigid-body rotation region is often quite limited in radius and has a steep gradient of dV / dr > 20 km s-1 kpc-1 (134, 287) [often larger than 50-100 km s-1 kpc-1 in more massive spirals (62, 133, 212)]. The peak velocities in spirals are attained in a nearly flat rotation curve that extends over most of the optical galaxy (37, 38, 102). Thus Irrs exist in a state with minimal differential rotation in star-forming regions, while star formation in spiral disks suffers strong shear due to differential rotation.

The forms of the rotation curves thus strongly affect the optical appearances of galaxies and insure that low-density disk systems will be morphologically distinct from spirals. Strom (334) demonstrated that the degree of gas compression during passage of interstellar matter through a density-wave spiral arm depends upon the maximum Vc value, and that for typical arm inclinations, only galaxies with Vc gtapprox 50-100 km s-1 will produce spiral arm shocks. Thus, slowly rotating galaxies should not be capable of arm shock-induced star formation and thus will not appear as spirals. Similarly, patterns produced by propagating star formation will not distort into spirals in the absence of differential rotation (116, 315, 316), nor will shear amplification of small inhomogeneities lead to the creation of spiral armlets (357). So whatever the reader's favorite spiral arm theory may be, it probably will not directly stimulate star formation in slowly rotating Irrs (cf. 255). Star formation processes in Irrs are free from internal dynamical forcing and therefore appear in a free-wheeling, chaotic natural state. From this perspective, the luminous clumpy Irrs with their spiral like rotation properties will be more difficult to produce, which may explain their rarity and tendency to be found in interacting systems where star formation can be externally stimulated (50, 177, 201).

The kinematics of Irrs are qualitatively consistent with their exponential brightness distributions. Mass models based on exponential disks (117) or low central concentration, Gaussian density distributions (360a) reproduce observed Irr rotation curves for M / LB ltapprox 5 in the central regions but fail in spirals (16, 17). A recent study of star cluster kinematics in the LMC similarly finds no evidence for a spheroidal component, even among old globular clusters (119). The Irrs thus have the expected character of pure disk galaxies containing (nearly) the observed amount of luminous mass, while spirals must be pinned on high-density central cores and also embedded in extensive stellar (and nonstellar?) halos. The lack of dense cores explains why Irrs rarely have optically identifiable nuclei and are not commonly sites for violent activity (20, 72; but see also 148a, 151). Evidently, a deep central gravitational potential well is a necessary ingredient for formation of massive, dense galactic nuclei and their associated fireworks.

When velocity fields in Irrs are observed with sufficient spatial and velocity resolution, numerous complexities appear. (a) HI is usually clumped into large clouds, which may not lie on the smooth rotation curves (as in the LMC; 241). (b) Irr galaxies are preferentially barred, and the bars primarily result from an enhanced density of older stars (75, 105, 188, 189), and usually do not lie on center [i.e. Irrs strangely prefer to be asymmetric rotators (78, 105)]. The presence of bars is to be expected if the Irrs are indeed dynamically cold systems that do not contain disk-stabilizing stellar or dark halos (93, 262, 318, 371). The combined impact of asymmetries and bars introduces significant perturbations into the velocity field (cf. 7, 236). Off-center bars remain theoretically poorly understood (63, 78, 106), which presents an impediment to construction of dynamical models for Irrs. (c) Injection of energy into the interstellar medium by massive stars can produce relatively large kinematic effects in slowly rotating Irrs. For example, giant HII regions expanding with velocities of ~ 20 km s-1 (124, 188, 347, 348) can significantly distort rotation properties as deduced from ionized gas. Blowouts in the cool interstellar medium due to stellar winds and supernovae lead to large-scale expanding gas shells (141, 243, 383) that are seen as ``holes'' in the HI distributions and can cause HI maps of Irrs to resemble Swiss cheese in both velocity and physical spaces (e.g. in the SMC; 19, 154, 331).

If one takes only optical light into account, then Irrs, like spirals (102, 255), must contain ``dark'' matter, but much of this is in the form of gas, including both easily detectable HI (and its associated helium) and elusive molecular material that nonetheless must be present at some level (197, 227, 390). Based on detections of CO emission in NGC 1569 and the LMC, the total mass of gas in inner regions of some Irrs may be as much as twice the HI value. The situation regarding dark mass, possibly nonnucleonic, that does not radiate detectable electromagnetic radiation is at present extremely unclear. Tinsley (354) and Lin & Faber (229) applied indirect arguments to conclude that dwarf Irrs probably have spirallike dark envelopes, but in both cases gas content, stellar population properties, and accuracy of dynamical masses introduce uncertainties. Feitzinger's (105) investigation of LMC dynamics, on the other hand, suggests there is little or no unaccounted mass, and a similar result was obtained by Gallagher et al. (126) for normal-mass Irrs (M ltapprox 1010 Msun) based on constant star formation rate models. Blue Irr galaxies of higher inferred dynamical mass, however, are found by Gallagher et al. to have excess mass for their optical luminosities; thus, two dynamical classes of Irr galaxies may exist, although the uncertainties are large.

The issue of spirallike dark envelopes in Irrs is relevant to fundamental points such as the necessity of dark matter for galaxy formation and whether the luminous baryon mass to total mass ratio could be nearly constant in all galaxies (cf. 101, 103, 229). Furthermore, the presence of dark envelopes in small systems would place useful constraints on relic elementary particle interpretations of dark matter [e.g. hot particles such as light neutrinos are unable to cluster on such small scales (68, 276, 360)]. We have seen that evidence for significant amounts of dark matter within Irrs is at best ambiguous. As in spirals, kinematic observations of Irrs at large galactocentric radii therefore must eventually play a crucial role, i,e. we should seek to identify dark envelopes from flat or rising rotation curves in regions exterior to most of the luminous mass. It does seem clear from the available data that single Irrs are unlikely to have rising rotation curves (but see 218), although the current observations of the gas kinematics of Irrs do not extend far enough or have sufficient spatial resolution to distinguish between flat and falling rotation curves. A new systematic high-resolution study of HI kinematics in inclined, noninteracting dwarf Irrs is needed to clarify the issue.

Next Contents Previous