In addition to this, it has been pointed out by Holmberg that the deter mination of the radial velocity possibly involves some systematic errors associated with the brightness of the galaxies.
In general, however, it seems most likely that at least part of the multiple galaxies are actually receding from each other.
Ambartsumyan considers that systems of the Trapezium type are young and unstable. The same view is held by the Burbidges /3/ with respect to the tight galactic chains discovered by B.A. Vorontsov-Vel'yaminov. But these groups also include some elliptical galaxies, which are considered very old formations. The author has already pointed out elsewhere /4/ that the presence of very long connecting filaments and tails indicates, however, that such galaxies have existed for a long time and are fairly stable. The elliptical galaxies included in chains and trapeziums may be subjected to the following considerations: 1. Such elliptical galaxies are young, although it is not yet known in what way they might differ from old elliptical galaxies.
2. These are not elliptical galaxies, but only appear to be so (this view has not yet been confirmed).
3. Notwithstanding mechanical considerations, the systems in which they are included may be old and stable. Might it not also be possible that in an old, initially stable system a mutual repulsion could subsequently develop, as a result of galactic evolution? Many people are puzzled by the question of the origins of the literally fantastic energies which make the fragments of a divided galactic center scatter with velocities of hundreds of kilometers per second. The break-up of galactic centers may perhaps be attributed to the ordinary properties of superdense prestellar material, but then it becomes difficult to explain what causes the mutual recession of already formed galaxies which interpenetrate and have independent rotations.
If we base ourselves on the law of gravitation, the available data fail to allow for the enormous masses in the centers of some galaxies still consisting of superdense matter. The existence of such active centers in the younger galaxies should be detected by observation. Ambartsumyan considers that the "surge" from the center of the NGC4486 galaxy is a sign of such activity. That this is indeed a "surge"—as it is commonly called—
has not yet been proved. Galactic centers should be investigated more thoroughly, in order to find whether they show traces of activity.
Some new data seem to confirm, though perhaps not completely, Ambartsumyan's view that galactic centers have a special significance. Thus, for instance, the outflow of gas observed to take place from the nuclei of our Galaxy and other galaxies is hard to explain by ordinary processes.
The unusually fast rotation of the centers which markedly differs from that of the surrounding regions, and the dark filaments issuing almost from the exact center of the M31 galaxy, are also indicative of the peculiar nature of galactic centers. This last might seem to prove Ambartsumyan's idea that spiral arms are formed by the ejection of relatively small superdense masses from the center. It is difficult to imagine though how spiral arms could be formed in this way. First of all, a fragment of superdense matter should move in a straight line after being ejected. Further, in order to form a spiral arm it should continuously release along its path matter which would turn into stars and gas. It is true that galaxies of the Sc type are observed, in which the spiral arms consist partly of widely spaced conden sations, but such cases are generally rare among spiral galaxies. It is also difficult to explain the fact that spiral galaxies mostly have two arms, that the arms often branch out not from the center but from the ends of a bar, and that there are galaxies with a large number of branches issuing from the center in a single equatorial plane.
To all appearances, the spiral arms develop inside an already existing disk or from it, rather than disks being the result of the dispersion of spiral arms. The arms are usually completely disconnected from the center. They often begin at the periphery of the disk or within it. There are sometimes internal and external spirals, which are not connected with each other. They branch out from a ring, sometimes in pairs from a single point, or are completely isolated from the internal regions.