1.2. Short Historical Sketch

The anisotropy of galaxies' distribution in the sky and, in particular, a tendency to form close pairs, had already been noted early in the nineteenth century, which is well before the recognition of the physical nature of these objects as stellar systems. The very brightest pairs, primarily the interacting system M51, were noted by William Herschel. The first attempted systematic study of double galaxies was started by Lundmark (1927). To identify galaxy pairs, Lundmark proposed a variant of Aitkin's criteria for double stars. [add references] According to Lundmark, two galaxies comprise a pair if their angular separation does not exceed the angular diameter of the larger component. This criterion proved ineffective and was not used in later research.

A new stage in the study of double galaxies began with another Swedish astronomer, Holmberg. Using a large collection of sky photographs at Lund Observatory, Holmberg (1937) presented a catalogue of 827 small systems of galaxies including 533 pairs. In this extensive material, including apparent magnitudes, angular diameters, and apparent separations of the components, Holmberg presented the first statistical study of double galaxies. According to his criterion, two galaxies are a physical pair if their angular separation does not exceed twice the sum of their angular diameters.

The considerable influence of Holmberg's work on the development of extragalactic astronomy is still today leaving its mark, more than half a century later. Still, several very unsatisfactory aspects can be found in the Holmberg catalogue. Important amongst these are:

a) no pairs of galaxies in the southern sky could be included.

b) the negatives in the Lund collection were very heterogeneous, so that the Holmberg catalogue does not have adequate photometric accuracy for multiple galaxies.

c) as noted by Zonn (1963) and Vorontsov-Vel'yaminov and co-authors (1962-1968), about 40% of the pairs in the Holmberg sample are fictitious: due to the small scale of the negatives, many stars with photographic halos were included.

d) Holmberg's criterion does not distinguish isolated pairs from members of multiple systems (groups or clusters).

Further progress in the study of pairs of galaxies is associated with the name of Page. Page (1952) conducted the first all-sky measures of radial velocities of double galaxies from Holmberg's catalogue. In a series of papers, Page (1952, 1960, 1961, 1962, 1967, 1970) developed methods for measuring the masses of double galaxies and showed it is possible to determine the character of the orbital motion from the distribution of estimates of orbital masses.

Beginning with observational arguments and the same methods used by Page to infer anomalous masses for double systems, Ambartsumian (1956) introduced a radical proposal concerning non-equilibrium processes in multiple systems of galaxies. This hypothesis provoked much discussion at the Santa Barbara conference (Neyman and Scott, 1961).

An important milestone for all extragalactic astronomers occurred in 1956 with the culmination of many years' work in assembling the Palomar Observatory Sky Survey, comprising photographs in two colours covering the entire northern sky. Based on the Palomar Survey, Zwicky and collaborators produced a `Catalogue of Galaxies and Clusters of Galaxies' (CGCG) containing 27,841 galaxies with photographic magnitude brighter than 15.7 (Zwicky et al., 1961-1968). At almost the same time, Vorontsov-Vel'yaminov et al. (1962-1968) published the `Morphological Catalogue of Galaxies' (MCG), containing angular diameters, coordinates and structural descriptions for about 30,000 objects. Both catalogues noted numerous occasions of close double galaxies and the MCG specifically noted interactions. A significant number of cases of interacting galaxies and multiple systems recognized on the Palomar Survey were collected in the Atlas of Interacting Galaxies by Vorontsov-Vel'yaminov (1959, 1977). Large-scale images of many interacting pairs may be seen in Arp's 1966 `Atlas of Peculiar Galaxies'.

Many of the double galaxies in these catalogues do not form a satisfactorily selected sample because no consistent selection criteria were applied to identifying them. Pairs selected for evidence of interaction were byproducts of other programs. At this point there was a clear need for a new catalogue of double galaxies which would be selected solely from observed properties and which would be reliably restricted to isolated pairs of galaxies.

Such a project was started by the author and realised in the `Catalogue of Isolated Pairs of Galaxies' (Karachentsev, 1972). The selection of double galaxies and the compilation of the catalogue will be described in the next chapter.

Analogous attempts to apply strict, consistent selection to samples of double galaxies have been made by Turner (1976a, b) and Peterson (1979b). [add references]