The following empirical or physical parameters describing individual and statistical properties of galaxies are reviewed: morphological type including new or revised types; intrinsic luminosity distribution functions of two main components (spheroid and disk); true ellipticities of different types; characteristic scale parameters (effective diameter and luminosity density) of spheroidal and flat systems; masses and densities of spheroidal, disk and mixed systems from velocity dispersion and rotation velocities; rotation periods, maximum rotation velocities, angular velocities and momenta of different types; neutral hydrogen masses and densities; spectral energy distributions and colour indices; number of independent parameters from principal component (factor) analysis; luminosity functions and selection effects; clustering and space distribution; maximum density-radius and velocity dispersion-radius relations in systems of galaxies.

This report is a review of several proposed pictures for how matter came to be concentrated in lumps like galaxies and clusters of galaxies. The proposals cover a broad range of unrelated possibilities. I attempt here to subject these proposals to critical discussion, to discover the major problems for each case and the more promising lines for further study.

The distribution of galaxies on the sky is not random, but has structure on a scale of ~ 30 Mpc. Similar structure in the cosmic background radiation is not present. Another expected source of structure in the background radiation is the finite size of the light horizon in the early Universe. The lack of observable structure in the background radiation implies extreme initial isotropy. Quantum effects may be responsible for this isotropy, and for the initial perturbations from which galaxies and clusters of galaxies formed.

Analysis of Brown's (1964, 1968) data on the orientation of galaxies inside superclusters confirms his conclusion that the directions of the major axes of the galaxies are significantly correlated. For a number of superclusters, the minor axes are found to be oriented along the direction of the flattening of the supercluster as a whole. By analogy with the Local Supercluster where the same phenomenon is associated presumably with the rotation established by de Vaucouleurs (1959), one may conclude that, in general, superclusters rotate. It is shown that rotation on scales as large as these cannot be explained by tidal or by any local effects, and must have a primeval nature analogous to the 3 K microwave background. Rotation on such scales contradicts the standard idea of galaxy formation from density fluctuations but favours the concept of primeval whirls which gave rise to pregalactic density inhomogeneities.

An up-dated review is given of the evidence for the presence of intergalactic matter and radiation in the Universe. It is concluded that the only important constituents which may make a sizable contribution to the total mass-energy are intergalactic gas and condensed objects with a very high mass-to-light ratio. If the QSOs are not at cosmological distances, cold atomic hydrogen may still be the most important constituent and may contribute much more mass than do the galaxies. The X-ray observations still do not unambiguously show that very hot gas is present, though it is very likely on general grounds that some hot gas is present in clusters of galaxies.

The question of whether or not large amounts of matter, enough to close the Universe, are present, remains unsettled. From the theoretical standpoint the answer depends almost completely on the approach taken to the problem of galaxy formation and to the cosmological model which is favoured.

Some recent observations of head-tail galaxies are presented. The properties of these objects are reviewed and are shown to be consistent with the radio trail hypothesis.

High sensitivity measurements in the vicinity of nearby galaxies have shown the existence of appreciable quantities of neutral hydrogen at large distances (2 to 5 Holmberg radii) from the centre. The galaxies studied were the M81/M82/NGC 3077 system, M31, M33, IC 342 and M51. In all cases the gas is at velocities consistent with being bound to the parent galaxy. Of these the M81 system contains the most circum-galactic neutral hydrogen. In several galaxies there is as much neutral hydrogen outside the Holmberg dimensions as inside.

Dynamical aspects of the early history of our Galaxy are discussed, with emphasis on the two-component nature of disk galaxies. Topics include the Eggen, Lynden-Bell and Sandage picture of the Galaxy's collapse, dynamical problems of the disk, and the bulge component.

New H I observations indicate that the mass-to-light ratios in the disks of M31 and M33 are independent of radius.

A general review is given of chemical abundance determinations; particular emphasis is given to abundances of galactic and extragalactic metal-poor objects since presumably they represent the abundances of the primeval material from which our Galaxy was formed. The following results are stressed: (a) most of the helium present in the galaxies of the local group as well as in other galaxies was produced before these objects were formed, (b) the heavy elements were produced mainly as the result of stellar evolution, (c) there is a chemical abundance gradient in our Galaxy and, by analogy with other galaxies, it is expected to be steeper near the nucleus, (d) the carbon and oxygen content of our Galaxy increased at a rate different from the metals, reaching their present abundance earlier than the other heavy elements, and (e) the increase of the iron abundance in the disk of our Galaxy with time has been small while that of carbon is negligible; furthermore, as a group the super-metal-rich stars correspond to the old disk population. Several models of galactic chemical evolution are reviewed.

It is pointed out that the metallicity of galactic and extragalactic globular clusters correlates with the density of the regions in which they were formed. Within individual galaxies the regions of highest density have the highest metallicity.

Some of the differences in family traits between the globulars in the Andromeda Nebula, the Galaxy and the Magellanic Clouds might be understood if the M31 cluster system is ~ 1 × 109 yr older and the SMC cluster system ~ 1 × 109 yr younger than galactic globular clusters.

Scanner observations of five galaxies in the region near 9910 Å showed that the strength of the dwarf-sensitive Wing-Ford band is below the level of detection. This result is at variance with the predictions of strongly dwarf-enriched population models and in accord with the evidence for giant-dominated radiation at 2.3μ reported by Baldwin et al. (1973) on the basis of the strong CO band found in three galaxies. In the absence of any evidence for a power-law mass function, the consequences of a rounded function are examined.

The numerical study of a gravitating system of colliding particles has many potential applications, for instance the formation of flat galaxies, the formation of the solar system and the evolution of Saturn's rings. Preliminary results are presented for the galactic case. The system tends towards a final equilibrium state and it seems that such a collision mechanism can flatten a protogalaxy.

Axisymmetric systems, having differing amounts of initial angular momentum and containing 2000 stars (mass points) each, are allowed to collapse under their own gravitational attraction. Collapse and violent relaxation are found to lead to the formation of equilibrium structures after only a few free-fall times. The systems contain a sufficient number of mass points so as to be effectively collisionless and the equilibrium structures are consistent with accurate satisfaction of the stationary collisionless Boltzmann equation. Rotating equilibrium models produced in this way resemble elliptical galaxies, with elliptical isophotes. With cosmologically reasonable amounts of initial angular momentum, E0 to E4.5 galaxies can be produced. The equilibrium models show rotation curves in general agreement with that expected from violent relaxation theory. In particular the models show a central region with solid-body rotation. Further out, the systems become differentially rotating. When infall effects are included one can produce models with extended envelopes in good agreement with observed elliptical galaxies. It is suggested that the crucial factor in determining whether an elliptical or a spiral galaxy is formed is whether or not star formation is complete by the time the proto-galaxy reaches its point of maximum collapse.

Detailed dynamical model calculations based on a conventional collapse picture of galaxy formation, and conventional assumptions concerning star formation and stellar evolution, are found to be able to reproduce satisfactorily the basic structural and photometric properties of elliptical galaxies. The quasar phenomenon may be identifiable with the formation of the nucleus of a giant elliptical galaxy.