Astronomers noticed long ago that there exist many close groupings among the extragalactic nebulae. First, the great nebula in Andromeda was seen to possess two companions, NGC 205 and NGC 221. Among the brightest galaxies there are some conspicuously beautiful pairs such as the Whirlpool nebula, Messier 51, with the two companions NGC 5194 and 5195. Soon after the 60-inch reflector became available at the Mount Wilson observatory Dr. F. G. PEASE from 1911 until 1919 photographed a great number of individual and of double nebulae describing them in two classical papers (3) which preceded the final identification, shortly afterwards, of many of these objects as extragalactic stellar systems by H. D. CURTIS (4), K. LUNDMARK (5), E. P. HUBBLE (2) and others.
Curiously enough the study of double and multiple galaxies was severely neglected for a long time. Even today the data available are far too scant to allow us to draw any far reaching conclusions as to the intrinsic nature of groups of stellar systems. There are cosmologists who hold that these groups were formed out of large amorphous masses of primordial gases while others, including the author, think it more likely that most multiple galaxies represent the result of inelastic collisions and subsequent mutual capture of galaxies which were formed a long time before the encounters. The luminous intergalactic formations of stars recently found by the author (1) to connect many widely separated galaxies have greatly added to the mystery of the formation of a certain class of complexes of stellar systems. On the other hand this discovery, if explored properly, promises to furnish us with a number of entirely new clues about the fundamental problem of the evolution of galaxies and of groups of galaxies.
K. LUNDMARK who contributed so much toward the consolidation of our views about the realm of the extragalactic nebulae or, the metagalactic system, as he proposed to call it, also was probably the first to point out the unique value of the study of multiple galaxies for the derivation of the physical characteristics of distant stellar systems (6). His contributions to the subject were clearly reviewed by E. HOLMBERG (7) who says "KNUT LUNDMARK was the first to recognize the great importance of the double galaxies and to make use of them in obtaining the absolute characteristics of the anagalactic objects (extragalactic nebulae). In several papers (6) he calls attention to these problems. Thus LUNDMARK has investigated about 8000 NGC-objects and has found about 200 double and multiple systems. Furthermore he has made a search for these objects on the Franklin-Adams Plates, on Crossley plates and on plates taken with the great instruments at Mount Wilson."
"Concerning the absolute characteristics of the anagalactic objects LUNDMARK has pointed out that from the double galaxies we can, among other things, obtain the differences in absolute magnitudes and in absolute dimensions. From these we can derive the dispersions in the corresponding quantities, and through processes of integration, or rather interpolation, numerical expressions for several connections can be derived. Thus the double galaxies are of very great importance for the derivation of absolute quantities and interrelations within the metagalactic system."
HOLMBERG (7) himself in an extensive study published in 1937 continued the investigations started by LUNDMARK. He determined the apparent properties of 827 double and multiple systems of galaxies by means of the rich collection of Bruce plates of the Heidelberg Observatory. Some of his results and conclusions are as follows.
1. There is an unbroken line of transition: double galaxies - multiple galaxies - metagalactic clusters - metagalactic superclusters or clouds.
2. HOLMBERG considers as a double galaxy any system in which the
apparent distance
12
between its components is equal to, or smaller
than twice the sum of the largest apparent diameters
a1 and a2. Thus
| (1) |
In a multiple system the condition (1) must be valid for any one of the components taken together with one of the others. HOLMBERG states that the double galaxies selected according to the definition (1) prove in general to be physical systems, and that the probable number of optical systems should amount to only a few percent.
3. The components of multiple systems may be elliptical, spiral or
irregular galaxies. Using HUBBLE's distance scale
(2) HOLMBERG derives
for the members of the systems he investigated an average absolute
photographic magnitude M = - 14.35 and a dispersion in magnitudes
M = 1.02, not
very different from what HUBBLE found for individual
galaxies (2).
HOLMBERG also derives absolute diameters A of the member
galaxies and a relation between A and M.
4. HOLMBERG is of the opinion that the formation of double galaxies is probably due to capture. Starting from this premise he finds that the observed number of double galaxies requires an effective age of the metagalactic system of the order of 4 × 1012 years.
5. From the radial velocities determined for the components of
double systems HOLMBERG obtains for the selected galaxies an average
mass of about 1011
M
.
More work was recently done by TH. PAGE (8) who obtained small dispersion spectra (330Å / mm) of about forty nebulae which are the members of twenty double systems. PAGE comes to the conclusion that "although the observations do not yet permit an accurate determination of the frequency distribution of masses among these nebulae, there is good evidence in these and other data that many individual masses are over twice the mean, and many others are one tenth the mean or less. In fact these limited data indicate a sharp division between heavy weight and light weight galaxies of 1.5 × 1011 and 0.5 × 1010 solar masses respectively."
In section C we shall discuss recent advances which make it doubtful whether the statistical study made by PAGE on the basis of entirely insufficient material has any degree of reliability. Some of the main objections to this study obviously are a) PAGE did not apply any effective criteria whatever to distinguish between nebulae which are close neighbors in space and optical doubles. The pooling of both of these types of apparent and of real pairs may very well have given rise to the conclusion that there is a sharp division between light weight and heavy weigh galaxies, a result which is most unlikely when analysed theoretically and when compared with other well established data. Indeed, many thousands of double and of multiple galaxies have now been studied as they appear on 48-inch Schmidt plates. From the distribution of types and from the differences in apparent magnitudes of the member nebulae within these groups one must conclude that there is no drastic gap in the number of galaxies either as a function of luminosity or of mass. b) PAGE also did not have at his disposal any criterion to decide if neighboring galaxies actually form dynamic units or if they are escaping from one another. c) In order to determine the masses from the differences in radial velocities one must obviously know the absolute separations of the components in the double nebulae. This separation can be obtained only from reliable data on the absolute distances of the nebulae which are involved. Distance determinations for extragalactic objects have recently undergone severe revisions. For nebulae whose apparent velocities of recession are less than 5000 km./sec. the final results for the distances are still in doubt by factors of the order from two to five.
The recent advances made in the observation of multiple galaxies which will be described in the following will make possible a more rigorous use of the differences in radial velocities for the determination of the masses of the components of double galaxies. For the time being, however, the only reliable dynamic methods for the derivation of the masses of cluster nebulae are based on the procedures originally proposed and applied by the author (9). These procedures make use of the Virial theorem and the thermodynamic statistical analysis developed by EMDEN for the study of polytropic gravitational gas spheres (10).