by Prof. Samuel PICKELNER, P. K. Sternberg State Institute of Astronomy, Lomonosov Moscow State University

SUBSYSTEMS OF GALAXIES

Spiral galaxies are the most fascinating kind. Their subsystems differ in the mode of stellar motions, distribution of stars in space and stellar characteristics proper. Lately five basic subsystems have been tentatively identified in our Galaxy; these are the globular, intermediate II, disk-like, older planar (flat, flattened) and younger planar ones. The larger part of our Galaxy is concentrated in the disk-like and intermediate II subsystems, measuring roughly 700 and 1,400 ps*, respectively. The globular subsystem, which makes up about a quarter of the Galaxy's mass, is about 4,000 ps across. And the planar (flat) subsystem accounts for about 2 percent of the Galaxy's mass, it is about 300 ps in diameter, and expands farther away from the center.

Subsystems differ above all in Hertzsprung-Russel (H-R) diagrams** plotted for the constituent stars after globular and galactic clusters. The absence of hot stars in globular clusters attests to their old age and indicates that the star formation process came to an end there more than 10 bn years ago. But the presence of hot stars in the planar subsystem shows that star formation is still on over there. The thinner a subsystem, the younger it is. Unlike the other ones, the planar subsystem is composed of gas by and large, and it also includes the younger stars less than 100 mn years old.

The diagrams of different subsystems reveal subtler distinctions, too, related to the chemical composition. The older the subsystem, the less it has heavy elements (compared with hydrogen and helium). For instance, the sun contains 100 to 200 times as much of these elements as the oldest stars.

All these differences are explained today within the general model of galaxy formation from primordial metagalactic gas. The protogalaxy was in the shape of an extended and slowly rotating cloud of gas. Although compressed by gravitation, the sluggish movement of the cloud had as good as no effect on the character of compression. The stars formed at that

* ps - parallax second, parsec - Tr.

** Express a connection between the luminescence and temperature (spectral class or chromaticity indicator) of stars.

Pages. 66

time retained their radial movement as well as much elongated orbits. These stars show a very low content of heavy elements. Some had a large mass, they evolved rapidly and, by the end of their life span, became supernovas. The Big Bang, the mechanism of which is yet to be elucidated, gave rise to relativistic particles. This was accompanied by nuclear reactions generating heavy elements which the fireball explosion (Big Bang) ejected into the environment and which gradually changed its chemical composition.

The metagalactic gas cloud began rotating ever faster with further compression and its radius decreased, while the angular momentum of the gas remained the same. Therefore at the second stage the compression came to be directed to the plane of the Galaxy, and little by little the cloud became disk-like. Intermediate subsystem stars came into being then. Thus, the stars of each subsystem were cloud-like when formed from the primary gas. Since stars keep moving without friction, their spatial distribution remains but steady, while the gas motions slow down step by step, and the gas is concentrated toward the plane of the Galaxy. Today the gas forms the planar subsystem.

The gas compression process occurred rather fast. For example, it takes a globular cloud under 1 bn years to turn into a disk. But the planar subsystem does not compress furthermore to convert to stars to the full. Why? Incidentally, in elliptical galaxies, where rotation is slow and the gas is compressed toward the center, nearly all of it has turned into stars. What keeps the gas from further compression to the Galaxy's plane? One of the possible causes thereof might be the gas motions. Even though the gas energy converts to heat fairly soon, there are energy sources to sustain such motions. For instance, the hot stars. They ionize and heat the ambient gas whose pressure soars 100 - 200 fold. The hot gas expands and gets the surrounding clouds moving. But all that is not enough for preserving the planar subsystem and prohibiting its condensation.

What keeps the gas in the Galaxy is its magnetic field. The resilience of magnetic lines of force (field lines) counteracts gravitation and is responsible for the comparatively large diameter of the planar subsystem.

Seemingly it should be in the form of a thin disk invaginated into the thicker ones of other subsystems. In actual fact, however, this is not a disk but spiral branches belonging to the neutral hydrogen of our Galaxy and discovered by 21 cm spectral line observations. Such branches can be easily detected in other spiral galaxies as well. Although a small part of their mass is concentrated in the branches, these are clearly visible on photographs thanks to young hot stars thousands of times brighter than ordinary stars. But gas-and-dust clouds of interstellar gas, which are concentrated in the arms, may be responsible for strong absorption-so much so that the branches look like dark strips against the background of the lighter part of the Galaxy. Spiral branches enclose the bulk of the gas mass, and its density in between is much lower.

Pages. 67

As seen from numerous observations, there is a magnetic field in the spiral branches. Among other things, it has been detected by the Zeeman effect* on 21 cm line. We know that the lines of atoms within a magnetic field are split into polarized components. In the simplest case the atomic line splits up into two components proportional to the extension of the field, or rather, to its projection on the line of sight. The interstellar field intensity is but low, the split-up is negligible, and this field baffles detection. It is a little easier to detect it in the spectrum of a bright radio source capable of piercing the gas cum field. The gas produces an absorption line and, since the source is much brighter than the luminescent gas, measurements are thus facilitated. At present we can measure magnetic fields only in gas clouds imaged on a bright source.

There are but few bright sources near the galactic plane. Using a 76 meter radio telescope at Jodrell Bank, British astronomers have measured a magnetic field in the direction of such sources. In most cases the results proved within the measurement error, i.e. were rather unreliable. Besides, only the longitudinal component of the field could be measured, and this component is small when the arm is about perpendicular to the line of sight. Yet in some directions the observations reliably indicated the presence of a magnetic field. For instance, in the Perseus arm toward Cassiopeia А Н_‌ ≈ - 6.7 • 10^-6 e, and in the Orion arm in the direction of Taurus A H_‌ ≈ + 6.4 • 10^-6 e. The mean error was equal to (3 ÷ 4) • 10^-6 e.

Prior to direct measurements the existence of the Galaxy's field had been proved by several independent methods. One was based on radiation. It is relativistic electrons in motion within magnetic fields that are responsible for most of the Galaxy's radiation. This kind of emission differs from the thermal radiation of ionized hydrogen clouds in its spectrum and some other characteristics, such as polarization. Therefore it can be easily identified. Nonthermal radiation was found to be particular intense in spiral branches, though it is also present outside, even outside the Galaxy's disk. Relativistic electrons, which are a component part of cosmic rays, were detected in direct measurements, too. Their radiation shows there is a 10^-5 e magnetic field in the Galaxy, especially in its arms.

The Zeeman effect and radiation furnish no data on the field's structure and the direction of field lines. Such data are obtained by observing the stellar light polarization. It was found about 15 years ago that the light of distant stars is polarized, and the degree of polarization depends on a star's reddening and its position in the sky. This means that polarization is connected with interstellar absorption. The polarization plane, i.e. the direction of prevalent fluctuations of the lightwave electric vector, usually concurs with the direction of the galactic equator, though with apparent deviations here and there. Plotting the polarization plane slope as a function of a star's galactic longitude, we shall see the dots to be mostly grouped close to the abscissa, which means the polarization plane of the stars represented by these dots is about parallel to the Galaxy's plane. But in some directions, e.g. near the galactic longitudes l ≈ 80° and l ≈ 20°, the dots are almost evenly dispersed on the ordinate, i.e. the polarization plane orientation is chaotic here. The polarization magnitude in these sections of the starry sky is lower than elsewhere within the Milky Way. The cause of interstellar polarization is due to the absorbing dust particles having an elongated form and being oriented by a magnetic field. They rotate like propellers around a short axis directed along the magnetic field. Sidewise they look like elongated sticks oriented across the magnetic field. Such specks of dust exhibit a stronger light absorption capacity with the light oscillations oriented along the long axis. That is why this component is weaker in the radiation of distant stars, their polarization plane being parallel to the field lines. Looking lengthwise, we shall find dust particles to be randomly oriented on the image plane, and there should be no polarization. Since field lines (lines of force) are not quite straight, even in this particular case they do not concur with the line of sight everywhere. Therefore a small polarization is observed here, and its plane, determined by random fluctuations of the field, is oriented differently in different stars. Consequently, we may say that in the directions l ≈ 70 - 80° l ≈ 15 - 25° the line of sight is oriented lengthwise. In the former case the line of sight passes along the Orion arm. In the latter case (direction) no arm is detected; however, as demonstrated by Dr. I. Pronik, the strap in between arms contains many clouds of gas and dust. Thus, the field is directed along spiral branches or similar structures. This agrees with the theory that the arm can be steady only if field lines pass lengthwise. But should they cross it, some of its layers, field lines including, may come off and it will break up shortly.

The dust polarizing stellar light is concentrated in the clouds by and large. Therefore polarization indicates the presence of a magnetic field there. But magnetic fields occur outside the clouds as well, as seen in the shape of nebulae, dark ones in particular. Most of them are elongated. Dr. Schein (astrophysical observatory in the Crimea) attributes this phenomenon to the expansion of nebulae within the magnetic field. Such expansion is not inhibited along the field lines, but is impeded crosswise.

Very important data on the field's structure in the vicinity of the sun were recently obtained by a new method based on the rotation of the radiation polarization plane in magnetized plasma. As found still earlier, the radiation of many extragalactic radio sources is polarized linearly. Such emission can be broken down into two circularly polarized components. In each of them the electric vector gyrates in a circle, but in oppo-

* Zeeman effect - split-up of the spectral line of atoms within a magnetic field. - Ed.

Pages. 68

site directions. The resultant vector fluctuates along the straight line whose direction depends on the phasic differences of circular polarizations. The rate of propagation for right- and left-polarized waves along the plasma magnetic field is somewhat different, hence their phasic difference changes but slowly. As a consequence, the linear polarization vector turns about with the wave movement. The magnitude of this turn is proportional to electron concentration, the longitudinal component of field intensity, the pathway and the square of the wavelength. These values determine the rotational measure which can be found by measuring the position of the polarization plane on two different waves. The longitudinal component of field intensity includes the polarization sign too, and thus one can locate the direction of the magnetic field intensity vector from the rotation of polarization.

A group of American researchers has made a statistical study of sources with changed polarization. Their number was found to be 37. As the rotational measure was calculated for them, it proved to be about proportional to cosec of the galactic latitude. Which means that rotation occurs in a flat layer near the Galaxy's plane. The spiral arm in the vicinity of the sun can be a layer like this, too. In the Northern Hemisphere the field is directed to l ≈ 250°, while in the Southern - to 1 ≈ 70°. These longitudes coincide with the arm's direction. Consequently, field lines are indeed oriented along the arm, though their direction is opposite in the southern and northern halves of the arm.

CONDITIONS NEXT TO THE POLARITY INTERFACE

If a galaxy's either side carries an opposite sign, there ought to be a divide boundary (interface) near its plane, or a surface with rather specific conditions where field intensity is equal to zero. As we have already seen, the field pressure counteracts the gravitational pull and prevents the arm from contracting. But since the field intensity value is zero on the neutral surface, the arm' pressure compresses this region. Equilibrium is possible only if the gas pressure next to the neutral surface is equal to the field pressure in the arm. At normal gas temperature the gas density should be about 300 particles per 1 cm³ , whereas the mean density of arms is ca. 1 atom per 1 cm³ . Consequently, the gas near the neutral surface should be tightly compressed, and its layer, very thin.

Under normal conditions the gas and the magnetic field are interconnected, the gas compression is accompanied by the compression of a bundle of lines offeree within the gas. The point is that the lines of force (field lines) travel rather slowly through the conducting medium of a partially ionized interstellar gas. But should the layer (in which the field undergoes noticeable changes) be thin enough, we cannot neglect the diffusion of field lines through the gas. In our case the field lines separated by a thin layer of dense gas diffuse through this layer and, merging, come to be neutralized, annihilated,

Pages. 69

Central part in Hercules constellation.

for their directions are opposite. Meanwhile the layer of the compressed gas grows, augmented by a new gas devoid of magnetic field and compressed by external pressure. In 10 bn years the thickness of the compressed layer should reach ~2 ps. Bare of the field, this gas layer can give rise to interstellar clouds.

We cannot tell why most of the interstellar gas is concentrated in the clouds. For one, they could have come into being by dint of gravitational instability. We know that a homogeneous medium, if it is long enough, breaks up into several parts, and each contracts under the impact of gravitation. The size of these constituent parts depends on gas density and temperature. But regular clouds are too small for the gravitational pull to keep them from expansion.

Other processes implicated in gas formation do not apply in this case either. However, the gas compressed toward the neutral plane builds up a solid layer. If stars are born in this layer, hot stars among them, they ionize part of the gas which, expanding, will eject the ambient mass of dense gas from the layer. That's how interstellar clouds could take body and form. True, the ejected gas has little, if any, of the magnetic field, while ordinary clouds do have it. When a cloud gets into an arm, the field will first recede and then, pressing into the cloud, split it into thin parallel fibers separated by lines of force. Thereupon these lines enter into the fibers, which are not thick enough. The field will kind of fix the fibers that will persist for a long time. As a matter of fact, interstellar gas clouds are indeed of fibrous structure. This is clearly seen on a photograph of reflection nebulae in the constellation Pleiades, where the clouds are illuminated by bright stars. Thus far no explanation has been offered for the fibrous structure, and one cannot escape the impression that the lines of force were indeed pressed into the gas originally devoid of magnetic field. This phenomenon is described as flute instability.

FORMATION OF SPIRAL BRANCHES

Composed of gas, spiral branches are simultaneously tubes of field lines not necessarily of identical directivity. The field is an essential property of an arm, and arm formation cannot be divorced from it. That is why at the beginning of this article, when considering subsystem formation, we spoke about the origin of the planar subsystem, not arms, where magnetic field cannot be left out.

The magnetic field of galaxies must have been in the same medium where galaxies were formed. This field was weak (ca. 10^-11 e) in the metagalactic gas. At first it was compressed into large clouds which gave birth to galactic clusters and individual galaxies inside. Even today field lines seem to connect galaxies with the inter-galactic medium.

Field lines became compressed with the compression of a galaxy, with field intensity on the rise. Initially magnetic forces had no impact on the nature of condensation (compactification), they were too weak compared with gravitation and gas pressure. But when compressed into a disk or cylinder, magnetic forces grow faster in intensity than pressure and gravitation, and thus they should be taken into account at the final stages of compression.

Now let us imagine a spiral arm as a regular tube of field lines crossing a galaxy's core. A tube like that can be formed from a field dispersed within a disk only if lines of forces are compressed in one of the sections at least. The gas should be compressed together with the field, too. Today we do not see such kind of tight compactification, and the density along spiral branches changes not as much. Yet the compressed gas can turn into

Pages. 70

stars. For this reason we should look out for an area of enhanced stellar density. This is the center, the core. The density of older stars forming the globular and intermediate subsystems tends to increase a good deal toward the center in which it is very high, hundreds of times higher than in the sun' environs.

Now, taking into account the magnetic field and central condensation, let us complete our picture of galaxy formation we have sketched in the first part of the present article. So, the gas contracted into a disk, and so did the magnetic field. But a large portion of the gas came to be concentrated toward the center, where the density of field lines increased accordingly. Looking at the disk from above, we shall see the following pattern of distribution for lines of force. The magnetic tube is kind of constricted in the center, and then radiates within the galactic plane and sideways. Lines of force are resilient. Since they are compressed tight by the gas in the center and cannot come apart, they will straighten out, i.e. compress the ambient gas toward the tube's axis. Although the magnetic field is already compressed into a disk, and magnetic forces have grown in intensity, they are still weaker compared with gravitation and pressure. These strong forces, however, are in equilibrium, their motions are inhibited, the pressure decreases, and condensation keeps on. Under like conditions even a small but steady force is capable of imparting direction to condensation. As a result two new condensations are formed on either side of that in the center within the selfsame tube of field lines. These lines, compressed as they are in these condensations, diverge on coming out. At this stage magnetic lines compress the gas farther from the center, and trigger the formation of new condensations within the same tube of field forces. This process recurs until the wave of condensation crosses the galaxy.

The tube's compression will go on until the magnetic forces (growing, as we have said, faster than gravitational ones) have caught up with the forces of gravitation. Now the arm's equilibrium depends but little on the pressure, and therefore the extinction of the motions no longer results in further condensation. The excess gas converts to stars, with the residual gas sustained by field lines also turning into stars, albeit slowly. That is why the arms are still there, even though they are not much older than intermediate subsystems.

The spiral form of branches is due to the differentials in galactic rotation. Such differential rotation has a higher velocity toward the center, the inner parts of the arm travel forward, and thus a helical structure takes body and form. But here we come up against another predicament. The differential rotation is so strong that it should have twisted the arms to their present shape in less than 1 bn years, while the arms have been around much longer. This enigma has not been unlocked yet.

The above scenario connects the formation of spiral branches with the tube's compression in the central condensation. No condensation, no spiral branches either! In point of fact, there is a class of galaxies described as irregular ones. Usually they have no central condensation, or else it is faint. Despite the presence of a gas and magnetic field, they are actually devoid of spiral structure, with the gas and field distributed in the disk. The gas is still there because of the same pressure as in the arms; the field hampers condensation, though it is not condensed into a tube as in spiral galaxies. True, we can cite an opposite example: elliptical galaxies have a very strong central condensation, but no spiral branches. But there is little gas over there: owing to the slow rotation it congregated toward the center to turn into stars because of high density.

How did arms with oppositely directed lines of force come to be? The nature of a galaxy's compression at the incipient stage is a function of its rotation, the direction of its axis in particular. These parameters do not depend on the field, since magnetic forces are weak initially. However, the metagalactic gas field used to have some initial direction. Say, if the axis of rotation and the field are perpendicular, then the field is compressed into a disk; in this case all field lines carry an identical sign. But if the field is directed at an agle with the rotational axis, ultimately, upon the gas being condensed into the disk, some lines of force will go, as before, through the center, while others will fold up. An arm is formed in the same mode-the field is condensed into a tube, but in this tube some field lines are of different directionality, and the number of differently directed lines is not equal. Note that observation data also indicate asymmetry: in the Southern Hemisphere of our Galaxy the rotation measure is 1.5 - 2-fold higher than in the Northern.

Consequently, here the tube of lines of force which passes across the center lies below the galactical plane.

Thus, the spiral form of galactical arms results from a galaxy's differential rotation. The higher the concentration of the mass toward the center, the faster the differential rotation is, and the sharper the arm's twist. And yet about half of the spiral galaxies are in the shape of a spiral with a strap. Instead of arms they have a straight pivot from the ends of which spiral branches ramify. The pivot's straight form points to the absence of differential rotation in the center.

Zemlya i Vselennaya (Earth and Universe), No. 4, 1965

Permanent link to this publication: