In this chapter, we have examined the observational data for basic reasons why members of double galaxy systems are distinct from single galaxies. Some of these differences arise from the similar conditions of formation of the pairs, and others from the synchronized evolution induced by their proximity. We recapitulate the main properties.
1. In comparison with single galaxies, members of double systems exhibit a much higher fraction of elliptical galaxies, and the relative number of elliptical systems rises with decreasing separation between pair components (a morphological segregation effect). Among double galaxies a significant excess is observed of cases with the same structural type of the components, by comparison with random expectations. The correlation function for morphological type has a significant value only for common and neighbouring Hubble types. The amplitude of this correlation function increases for tighter pairs as well as for systems showing signs of interaction.
2. The mean value of the orbital mass-to-luminosity ratio for double galaxies changes monotonically along the Hubble sequence from 10f for ellipticals to about 4f for the latest types. The same tendency exists for individual mass-to-luminosity ratios calculated from rotation curves. Both the masses and mass-to-luminosity ratios of pair members exhibit a marked correlation, which testifies to common conditions of formation of the components of double systems.
3. More than half of the catalogue pairs exhibit evidence of interaction. The relative number of interacting double systems decreases monotonically with increasing separation between galaxies, reaching zero for X = 110 kpc, a maximum scale of interaction. Any galaxy of a certain Hubble type has the same chance of exhibiting signs of interaction. However, the distribution of types of interaction with Hubble type is markedly different. Thin tails and bridges as well as disturbances in the overall form of galaxies are characteristic of objects dominated by type I stellar populations with predominantly circular orbits. Amorphous, symmetric atmospheres are usually observed around elliptical galaxies in which the stellar orbits are predominantly radial and the internal velocity dispersion is high. This connection between the type of interaction and the characteristic stellar motions within the components of pairs agrees in detail with that expected from model simulations (Clutton-Brock, 1972).
4. The intensity and number of emission lines in the spectra of double galaxies have been used to define four spectral classes. The spectral characteristics of pair members exhibit a tight correlation with their structural types. This partly explains the observed excess of pairs having the same spectral class for both components. In addition to this correlation with morphological types, the shared spectral characteristics for members of double systems seem to reflect tidal interaction. Among the interacting spirals, the number of objects exhibiting strong emission is two to three times greater than for doubles without any signs of interaction. This enhancement of emission is probably due to the active redistribution and collision of gas clouds caused by tidal disturbances, which increases the flow of gas to the nucleus, thereby fueling its activities. Another mechanism, the physical transfer of gas from spiral components of pairs to ellipticals, does not have significant observational support.
5. Incontrovertible evidence for active star formation in pairs of galaxies is furnished by the excess number of blue Markarian objects among them. Evidence of strong emission in the nuclei and bursts of star formation is often found in double systems with small radial velocity differences and small separation, i.e., just those in which tidal effects are most effective. The excess luminosity in Markarian galaxies, a factor of 1.5 by comparison with other pair components, may furnish an estimate of the mean amplitude of the increase in star formation.
6. Photometric measurements of double galaxies provide evidence of strong correlation between the colour indices of components (the Holmberg effect). The closeness of pair members in colour cannot be due solely to the excess number of systems with common Hubble types. In pairs of elliptical components in which the reserves of gas for increased star formation are small, the colour correlation probably occurs as a result of their similar chemical compositions at the epoch of galaxy formation. The observed blue colours of galaxies in SS pairs and the correlation in colour indices for spiral members in pairs, may easily be explained as a result of simultaneous star formation, periodically induced by tidal disturbances.
7. A measure of the compactness of galaxies may be found in their mean surface brightness, which, for double galaxies, is approximately the same for all structural types but falls upon going from absorption-line objects to galaxies with strong emission lines. The decreased compactness and rich emission spectrum are often encountered for galaxies of low luminosity and small linear dimension. It is possible that exhaustion of the gas supply in dwarf galaxies proceeds at a slower rate than among normal galaxies, so that such objects can exhibit active star formation even at the present epoch. The mean surface brightness of pair members exhibits a dependence on separation. However, this may be explained as a result of selection effects caused by the isolation criterion, and so does not allow any conclusions on the role of evolutionary factors. Elliptical components in pairs have a significantly lower surface brightness than do isolated elliptical galaxies, which may be indicative of changes in their structure due to tidal effects.
8. In recent years much observational evidence has accumulated to indicate that the processes of star formation in double systems are considerably more active than in field galaxies or members of rich clusters. Recapitulating the numerous arguments in the literature, we see an increased occurrence of supernovae in interacting pairs, infrared and radio detection excesses among double galaxies, and high percentages of double systems among Seyfert galaxies and quasars. One might produce a picture in which the basic cause of activity in paired galaxies is tidal, involving collisions of gas clouds and induced disk star formation, as well as the redistribution of gas into the circumnuclear regions.
In contrast to this we cannot resist noting that many attempts at modelling the tidal influences in pairs have been somewhat over-simplified. N-body models moving in the gravitational fields of two massive centers tend to resemble more the rings of Saturn than real interacting galaxies since they neglect collective processes (Friedman, 1978). One frequently encounters tidal models limited to a simple monotonic approach of one galaxy to another, and therefore not exhibiting effects which could arise from motion on more general orbits.
Very little attention has been given to the role of various resonant effects in the process of synchronous evolution. Disturbances produced in a galaxy by its nearest neighbours may distinguish its evolution from that characteristic of a normal field galaxy. There is therefore an obvious need to model multi-component pairs, incorporating gaseous subsystems and paying proper attention to star formation. Such experimental calculations will obviously be extremely expensive in computer time.