Annu. Rev. Astron. Astrophys. 1985. 23: 147-168 Copyright © 1985 by Annual Reviews. All rights reserved

5.3. The Problem of Dust and Gas

The above discussion about ellipticals with shells is not complete. The fate of the gas contained in the infalling disk needs to be considered too. A typical disk galaxy of mass 10¹⁰ M brings along 10^8.5-9 M of neutral gas, which has to go somewhere. Only Centaurus A (76) and NGC 2865 (cf. 77) have detectable amounts of H I. It seems easy to say that the gas will be heated up toward the temperatures of the X-ray-emitting gas known to exist in field ellipticals and S0s (68), e.g. in NGC 3923, or that it could form stars, but what then about the presence of dust and gas in ellipticals?

To get a feeling for the statistics first, we have constructed a sample of field ellipticals based on a catalog of E and S0 galaxies south of -33°, thought to be complete down to m_B = 13.8 (94, 95). We reclassified all the galaxies in that catalog with radial velocities less than 4000 km s^-1; ellipticals were distinguished from S0s by using both the ESO and SRC films, which resulted in a list of 78 ellipticals and 47 S0s complete down to m_B = - 20.5 (H = 50 km s^-1 Mpc^-1). We estimate that only up to 5 galaxies in this list could be misclassified.

We then compared this list with that of Malin & Carter (77) for ellipticals with shells and with the lists of ellipticals with dust (52, 120, 137, 155, and our own notes). Among the 78 ellipticals are 10 with shells, 9 with dust (of which 3 have shells as well), and 19 radio detections as listed in (96). Among the 47 S0s there is 1 with a shell, 4 with clear dust lanes always aligned with the major axis, and 5 radio detections. Of the 78 ellipticals 27 occur in clusters or very rich groups. Since neither shell galaxies nor dust-lane galaxies occur among these, we have a total of 51 ellipticals in a:representative field sample. Thus shells or dust is observed in about 20% of the field ellipticals, but there is surprisingly little occurrence of both phenomena together. Perhaps this is partly exaggerated; for example, if ellipticals are oblate and have shells of small azimuthal extent, the projection angles most favorable for detecting dust lanes may not be so favorable for detecting shells as well.

Slightly higher frequencies of occurrence of both phenomena have been reported recently using higher resolution data. Schweizer (115) finds on the basis of 4-m plates that 16 out of 36, field ellipticals have shells (ripples) and 9 out of 36 have dust. The occurrence of small dust lanes near the center in several more galaxies (97) may increase the frequency of dust in ellipticals to as high as 40%, comparable with the estimate of the detection rate of ionized gas with equivalent width at 3727 ([O II]) larger than 1 Å (21).

With respect to the radio detections, 9 occur among the 27 cluster or rich-group galaxies and 10 among the field ellipticals. Of the field elliptical detections, 2 occur in ellipticals with both shells and dust (For A and Cen A), 1 in a shell galaxy with no dust, and 4 in galaxies with dust and no shells. Thus the presence of dust increases the detection rate from 10% to 60%, while the presence of shells does not seem to affect the detection of radio emission. As far as ionized gas is concerned, 19 of the 78 galaxies have been included in Caldwell's (22) list. Eleven of these have been detected: 6 out of 6 dust-lane galaxies, and 5 out of 13 (40%) others. Thus the active galaxy phenomenon in ellipticals seems to be associated with dust and gas but not with shells or ripples. This supports indirectly the view that shells and ripples are stellar phenomena.

The results from the orbit calculations and from the analysis of preferred planes can also be applied to the ellipticals with dust. Unlike S0s, ellipticals do not have a well-defined main plane, and an ambiguity exists between near-oblate systems tumbling around the short axis and near-prolate systems tumbling around the long axis. This can only be settled if the kinematics of both stars and gas is known and if an assumption is made about the sense of tumbling with respect to the stellar streaming motion. Another difference with the polar ring situation is that the dust lanes occur mostly inside the optical image. In the case of Cen A, the main body is most likely prolate on account of the sharply defined shells, and then the warped dust lane can be in a stable configuration if the figure tumbles retrograde with respect to the gas streaming motions in the outer parts (142).

For several ellipticals with dust, the gas rotates around the major axis, while the stellar rotation, if present, is around the minor axis (e.g. 11, 21, 22, 30, 82, 119). In others, the two angular momentum vectors are more aligned (22, 119). In some cases the dust lane is really a thick annulus (22, 51) that is warped in its outer parts. Several dust lanes exist that are not aligned with any principal axis (51). The case of a gas disk rotating around the same axis as the stars, but with an antiparallel-spin vector (12, 23), strongly suggests an extragalactic origin for the gas. Neutral hydrogen in some ellipticals (see 61) apparently rotates in planes different from both the ionized gas and the stars, e.g. in NGC 4278 (30, 89, 108). This could indicate precession of a gas disk resulting from capture (48). Thus, if the gas is indeed of external origin, the swallowing of dwarfs again seems a popular candidate to explain its origin. The gas and dust should settle eventually into a preferred plane, but the fate of the dwarf's stellar component is now unaccounted for. So either the recipient ellipticals are different from those that have shells, or the dwarfs are of a different nature, or the collision parameters are different.