A new hydrogen line survey of M31 is described. A rotation curve of the galaxy is derived from these data. The northern half of the rotation curve can be traced to 27 kpc, the southern to 30 kpc. They are in general agreement; both sides show an extensive region of essentially constant rotational velocity.

Given the wealth of data, the rotation curve, and the necessity for out-of-plane hydrogen demonstrated in the preceding paper, it seems desirable to attempt to establish a systematic procedure for determining the three-dimensional distribution of hydrogen. Under the assumption of cylindrical rotation this is, in principle, possible for most of the galaxy.

Aperture synthesis observations of the neutral hydrogen in the extreme NE and SW regions of M31 have been made using the Cambridge Half-Mile Telescope, with an angular resolution of 3′.6 × 5′.5 and a resolution in radial velocity of 16 km/s (Newton and Emerson, 1977). These observations show that a warp in the HI distribution exists in opposite directions at each end of the galaxy.

Distortions in the distribution of the HI in the outer parts of M33 are thought to be due to a warp in the HI plane. Measurements by Reakes and Newton with the Half Mile Telescope at Cambridge provide new evidence on the extent and kinematics of this gas. The integrated HI map with an angular resolution of 7 × 14 arcmin in shown in Fig 1a. The wings extend roughly symmetrically in outline to 70 arcmin (14 kpc) from the nucleus in the plane of the sky both to the NW and the SE. The radial velocity field in Fig lb shows large deviations from normal rotation in a plane in the outer parts. A model in which the gas rotates in circular orbits whose inclination and position angle of the major axis vary with radius, fits the HI distribution and kinematics quite well. Much larger extensions in R and z are required than in the model of Rogstad et al. (1976) and with a slower trend of i and P.A. with radius.

Extended HI-distributions around late-type galaxies have been found at a surface density of roughly 1020 atoms cm−2 (e.g. Roberts 1972, Davies 1973). For the northern part of M33 a low density Hi-component (∼ 1019 cm−2) was observed to extend to about 1.5 Holmberg radii (Huchtmeier 1973). As part of a program to study the HI-distribution and kinematics of HI-shells (i.e. the HI outside the Holmberg diameter, dH) and to search for high-velocity cloud phenomena in a dozen nearby late-type galaxies the 100-m radio telescope in Effelsberg has been used to map the neutral hydrogen in and around M33 with high sensitivity to the limit where sidelobe contributions become important. Unpublished studies of the antenna pattern and that of Reich et al. (1976) place this limit to about 1018 cm−2 in the case of this galaxy. This limit can be reached in one hour of observing time. In 1973 an area of approximately 2°.5 by 2°.5 has been observed, which has been undersampled with the chosen grid separations of 9′ in δ and 12′.9 in α. Integration times inside dH were considerably shorter than outside.

Before reviewing the radio properties of active spiral galaxies it is appropriate, as well as instructive, to think about a working definition of an active galaxy. This may seem a trivial point when we consider classes of objects like Seyfert galaxies or radio galaxies but this is much less so when we inspect the nearby systems. In a sample of nearby galaxies one is likely to encounter much more milder forms of activity than that known from studies of more distant galaxies. This is, of course, due to the fact that the latter have been selectively taken from a much larger reservoir of objects. (Although this may seem a disadvantage from an observational point of view, there are some clear advantages as well. Firstly, we can obtain a much better linear resolution once we have been able to isolate the active nearby galaxies and, secondly, we eventually may hope to learn more about the physical situation that existed prior to the onset of the active period and how the surrounding medium reacts to the activity.)

When Heeschen (1968) found that the nuclear sources in the two elliptical galaxies NGC 4278 and NGC 1052 contained flat spectrum unresolved sources he conjectured that these might be weaker versions of the compact optically thick and strongly variable sources found in the quasar nuclei. This conjecture appears to be correct since these are both now known to have diameters <0.1 pc (Cohen et al. 1971) and to be variable on time scales of order one year. This variability has recently been confirmed by observations with the VLA (Heeschen, private communication).

The term “nearby galaxies” is not very precise. If we restrict ourselves to galaxies within the local group, we are really only talking about our Galaxy, M31 and M33. Since the Galactic Center has been reviewed extensively by Oort in Annual Reviews (1977), and I feel that there is nothing exceptional to say about the nuclei of M31 and M33 as far as phenomena other than their stellar content and central dynamics are concerned, to discuss interesting properties we must consider more distant objects. If we go out to the distance of the Virgo cluster, we already include objects such as NGC 5128, M82 and M87. Each of these galaxies shows or was claimed to show evidence of different kinds of violent nuclear activity. Indeed, it is obvious that within the volume occupied by the supercluster (whether or not it is really a physical entity) there must be many galaxies in which nuclear activity can be detected.

The non-stellar activity in the nuclei of nearby galaxies poses problems of its own. But it gains added interest insofar as it may provide clues to the nature of quasars and the unusually energetic nuclei of some more remote galaxies. The evident qualitative resemblance between these spectacular phenomena and some of the nearby galactic nuclei discussed at this symposium suggests that we may be witnessing a scaled-down or slow-motion version of the same physical mechanism; and the likelihood that dead quasars vastly outnumber living ones suggests that defunct remnants - perhaps displaying some low-level residual activity - may lurk in the centres of most large galaxies.

Available evidence on the mass-to-light ratios in binary galaxies, small clusters and in rich clusters is reviewed. The interpretation of the binary data remains ambiguous. The relative velocity of M81 and M82 implies that m/LV > 12. The Local Group data also appear to favour the existence of some hidden mass associated with M31 and the Galaxy. In the Virgo cluster a significant fraction of the virial mass may be concentrated in the halo of M87. In Coma the missing mass was probably stripped from the haloes of the two dominant E galaxies and is now distributed throughout the cluster.

It is suggested that metal-poor stars in the galactic disc have a higher CNO abundance than do halo stars of similar metallicity. This may explain why globular cluster stars lie on taller red giant branches than do moderately high-velocity stars near the Sun. The dominant old red giant population in the disc of M31 appears to be similar to that in the Galaxy. Most of the red giants in M33, NGC 6822 and in the LMC lie on taller red giant branches than do those in the Galaxy and M31. This suggests that old evolved stars in these relatively low-luminosity galaxies have a lower heavy element abundance than do their counterparts in the galactic disc. The red giants in M32 appear to be intermediate between those in M31 and M33. This supports the idea that M32 was once much more luminous than it is at present.

A review of the observed facts and the observed theories of distortions of the outer parts of galaxies.

[Manuscript not received]

Computer simulations of galaxy encounters (Toomre 1974 and references therein) have strongly supported the idea that the luminous bridges and tails in many multiple galaxies are a result of gravitational interaction. The need for an observational test of the kinematics of these models has prompted observations of NGC 4038/39 (the “antennae”) in the 21 cm line of neutral hydrogen (HI) with the Westerbork Synthesis Radio Telescope (WSRT) at an angular resolution of 100″×150″ (α×δ). A detailed description of the observations and data reduction is given elsewhere (van der Hulst, 1977). Neutral hydrogen is definitely associated with the tails as was already indicated by the filled aperture observations of Huchtmeier and Bohnenstengel (1975) at the lower resolution of 8′.7. The distribution of HI column density is shown in Figure 1 superposed on a print of a plate obtained by F. Schweizer with the 4 m telescope at Cerro Tololo. Owing to the low declination of the object (−19°) the WSRT map is affected by several sidelobes; the main effect is an attenuation of extended east-west structures, for which reason the HI along the southern tail in Fig. 1 shows a gap. Computer experiments and a comparison with the filled aperture observations show that the WSRT data are entirely consistent with an HI distribution which is uniform along this tail and increases with about a factor 2 at the extreme end, where the tail is very blue and small HII regions are present (Schweizer, elsewhere in this volume). The tails contain 2.9 × 109 M⊙ of HI (assuming a distance of 20 Mpc), 85% of which is in the southern and more prominent tail. About 1.1 × 109 M⊙ of HI is associated with the galaxies which are not spatially resolved at the present angular resolution.

Today I would like to show you some animals from the zoo of Alar Toomre. These animals are all characterized by long tails and belong to two subspecies: The first consists of animals which always huddle in pairs, each individual having one long tail. The second, rarer subspecies comprises those strange animals which live alone but have two tails!

It is increasingly certain that the principal concentrations of neutral hydrogen gas are in spiral and irregular galaxies. The intergalactic medium (IGM) appears to be either a near-void with <ρ> ≲ 1.5 × 10−11 cm−3 from the 3C9 measurements of Gunn and Peterson (1965) or at best a hot, highly ionized gas with very little neutral hydrogen present (Field and Perrenod, 1977). In clusters of galaxies, the intracluster medium (ICM) also appears to be mostly hot and ionized, as suggested by X-ray observations (Jones et al. 1977) and by observations of head-tail radio galaxies. Even if most of the IGM and ICM were ionized, however, one might imagine that there exist high-density concentrations where the cooling rate is sufficient to allow recombination, particularly in the vicinity of galaxies. One might expect such condensations to be revealed either by 21-cm emission, or by absorption against a bright background continuum source.

While most of the sky has been surveyed in the 21 cm line at velocities within a few hundred km/s with respect to the LSR, the coverage is highly inhomogeneous in sensitivity, angular and spectral resolution and velocity extent. As interesting information has recently become available on the structure of the clouds, however, inferences on the dynamical and thermal state of the gas have become possible. Here, we shall concentrate on the direction in which this structural information influences the dichotomy stated in the title of this paper.