IV. Principal Characteristics of Compact and Post-Eruptive Galaxies

Shortly after some of the theoretical predictions concerning compact extragalactic aggregates had been confirmed observationally, the following statement was included in the report of the director of the Mount Wilson and Palomar Observatories (Yearbook No. 62 of the Carnegie Institution of Washington 1962/63),

A systematic search on 48-Schmidt telescope plates was undertaken for exceedingly compact galaxies that are actually difficult to distinguish from stars. Spectra obtained by Zwicky show that they run the whole range from ordinary G and K types to systems showing only emission lines and no continuum. It is conjectured that radio sources such as 3C48 and 3C273 lie at the end of this sequence, and an attempt will be made to test whether intermediate systems are weak radio sources.

Brief reports on the observations and the theory of compact and of post-eruptive galaxies have appeared in all of the subsequently Yearbooks Nos. 63 to 70 of the Carnegie Institution of Washington, while details can be found in the special articles and reviews listed in the two appended bibliographies. We therefore here give only a short tabulation of some of the outstanding characteristics of the objects included in this catalogue.

Structural Features

Restricting ourselves to features that can be seen in the ordinary photographic and visual ranges of wave lengths, we are only interested in those objects which exhibit nuclei, cores, discs and prominent knots of surface brightness greater than the 20th magnitude per square second of arc, as stated already.

In the limit we may have extremely compact galaxies of diameters smaller than a fraction of a second of arc that cannot even be distinguished from stars with the 200-inch Hale telescope and whose surface brightness actually far exceeds the 20th magnitude per square second of arc. In addition there exist quasistellar objects of this type that have associated with them faint jets, plumes, spiral arms, halos or filaments connecting them with other galaxies, the two patchy interconnected compacts at R.A. 8^h55.8^m and Decl. +6°31' (1950.0) being examples of this type. As the luminosity of the external formations increases relative to that of the compact central body, we enter the realm of galaxies endowed with compact parts that in the end merge with the extended and conventional families of ordinary galaxies which contain only insignificant parts of high surface brightness.

It should be emphasized that many of the red and very red compact galaxies included in this catalogue do not satisfy the condition 13) in the photographic range of wave lengths but only in the red or perhaps only in the infrared. Furthermore, it is quite possible that there exist galaxies that are not compact in the optical or any of the adjacent ranges of wave lengths. They may, however, be bright in the regions of the radio waves and of X- and -rays, indicating that temporary implosions and eruptive processes take place in them which produce specific non-thermal radiations at high intensity.

The distribution of the surface brightness in isolated compact galaxies may either be fairly flat throughout the whole system or it may rise more or less steeply towards the center, occasionally culminating in stellar-like peaks of enormous surface brightness and diameters equal to a small fraction of a second of arc, such as are found for the nuclei of some of the Seyfert galaxies. The variable compact I Zw 1 at R.A. 0^h51.0^m and Decl. +12°25' is a case in point. (We remark in parenthesis that all coordinates given in this catalogue refer to the epoch 1950.0.)

Many of the compact galaxies are found to be associated either with ordinary galaxies, with other compacts, or with combinations of both. One beautiful system of this kind is I Zw 96 at R.A. 14^h43.3^m and Decl. +51°35', showing a central E0 galaxy and several bridges and jets connecting It to a number of other objects, two of which have redshifts closely equal to that of the central galaxy. The symbolic velocity of recession, as determined by Sargent ⁽⁶⁷⁾ from both absorption and emission lines, is <V_s> = 27094 km/sec. ^*

Pairs and multiplets of compacts are unusually common. Often the various component objects are remarkably equal in structure, luminosity and color. In some extreme cases two or more compacts are imbedded in a more or less luminous matrix and may be in close contact, or almost coalescent. These systems offer particularly favorable conditions for the determination of reliable values of the masses involved, which can be obtained from the study of the differential redshifts of the various components. The object I Zw 4 at R.A. 1^h20.7^m and Decl. +34°19' is a good example of this type of coalesced system ⁽¹⁵⁾.

Although small groups of compacts are very common, only a very few clusters composed mainly of compact galaxies have been found ^{(55, 69)}. On the other hand, clusters of elliptical galaxies often contain many compacts. Some examples are shown in the Appendix of this catalogue. The Leo A and B clusters are examples of aggregates that contain many blue compact galaxies ⁽⁵²⁾.

Colors of compact and of post-eruptive galaxies

As was to be expected from our general theoretical considerations, compact galaxies show a much wider range of color and of spectra than the ordinary galaxies. Some of the basic reasons for this fact are as follows. As galaxies contract, either slowly or implosively, or, as the cores of ordinary galaxies grow in mass and compactness because of the accretion of stars and dispersed matter from the surrounding suburban formations, several important events take place. The radiation density inside of the compact objects increases and eventually causes the surface layers of the constituent stars to evaporate. These stars thus become blue and move towards earlier spectral types. The ever denser and faster moving interstellar gas clouds further help to whittle down the stars while their constituent ions, atoms, radicals and molecules get excited and give rise to a variety of sharp emissions. As the compactness of the galaxy grows, direct collisions ⁽⁷⁾ between stars become more frequent, causing their partial destruction and giving rise to ejection of various forms of matter. In the limit, collisions at velocities of several thousands of kilometers per second will result in the formation of neutron stars ^{(51, 53)} and eventually of the appearance of objects of the type HADES (H) ⁽³⁰⁾. This will lead to the production of neutron star studded compact galaxies and compact galaxies rich in H objects, that is models of cosmic bodies which we need to understand both the physics of the radioquiet ultracompact galaxies as well as of the radio waves emitting quasars. On this view the latter represent simply a relatively rare "pathological" species of the very much more numerous compact galaxies. They differ from radioquiet compact galaxies only because they have temporarily associated with them tenuous clouds of plasmas that emit the radio waves by synchrotron radiation or some other mechanism. One of the possible origins of such temporary plasmas lies in the ejection of gas clouds that result from implosions in compact galaxies. Quasars and quasistellar or ultracompact galaxies thus all belong to the general family of compact galaxies as integrally defined by the condition 13.

Some of the compact galaxies listed in our catalogue were found to be identical with already known radio sources, such as 3C371, 3C390.3, 4C35.6 and others. Although only a few dozen compact galaxies have been tested so far for their radio emission, it would appear that the great majority among them do not give rise to radiation intensities observable on the Earth of more than about one quarter of a flux unit (10^-26 W m^-2 Hz^-1).

Of great interest are those compact dwarf galaxies that observationally seem to be indistinguishable from giant H II regions in nearby galaxies. Among these the best investigated are ^{(16, 70)} II Zw 40 at R.A. 5^h53.1^m and Decl. +3°24' and I Zw 18 at R.A. 9^h30.5^m and Decl. +55°27'. These two systems were found to be strong emitters of 21 cm radiation ⁽¹⁷⁾. The fact that they are either very young systems, as measured in terms of the age of the Earth for instance, or that the luminosity function of the newly formed stars in them differs radically from that found in our galaxy confronts current cosmological theories with the greatest difficulties. An intensive search for similar objects is highly desirable.

Absolute luminosities and masses of compact galaxies

So far compact systems have been found with luminosities in the range from dwarfs to supergiant galaxies. Assuming that the redshifts observed in the spectra of all types of galaxies are essentially all cosmological in origin, the supergiant galaxies so far investigated lie in a range of indicative absolute visual magnitudes -20.0 > M_V > - 23.0. This may be compared with an average M_V = -21.7 for radio galaxies, a range -15.4 > M_V > -20.2 for Seyfert galaxies and a range M_V = -22.5 (for 3C47) to M_V = -25.7 (for 3C273) for quasistellar radio galaxies. These findings give additional support to our contention that quasistellar radio sources are but special limiting cases of compact galaxies.

Preliminary determinations of the masses of the most massive compact galaxies indicate that values as high as 10¹³ M exist, that is twenty times as high as those characteristic for the most massive supergiant common galaxies ^{(15, 16, 67)}.

The spectra of compact galaxies

The overwhelming majority of the spectra of ordinary galaxies, as listed for instance by M. L. Humason ⁽⁷¹⁾ and associates, are of the following types. Elliptical and S0 galaxies show almost exclusively spectra of the type G, while the spectra of normal and barred spirals cover the range from F0 to about G7, with only a handfull of the type A. A small fraction of the ordinary galaxies also show the Balmer, [O II] and [O III] lines in emission.

The recent investigations of compact galaxies have added a great number of spectral types to those previously known. And there seems to be no end in sight to the discoveries still to be made, especially if investigations are extended to both sides of the optical range of wave lengths.

Generally it may be stated that so far, in the optical range of wave lengths the following types of spectra of compact galaxies have been observed.

Emission Spectra:

a. Blue and red continuous and apparently completely featureless spectra.

b. Strong or weak continua with only 3727 of [O II] in emission.

c. Strong or weak continua with only the Balmer lines in emission.

d. Strong or weak continua with the Balmer lines, [O II] and [O III] in emission.

e. Strong or weak continua with many combinations of permitted lines of H, He, Fe II etc. and of forbidden lines of [O II], [O III], [O I], [Ne III], [Ne V], [S II] etc. The complexity of combinations will of course grow as the observations are extended into the ultraviolet, or lines are redshifted into the optical regions, as is the case for most quasars.

Unique are a few very bright compacts, with stellar cores showing emission lines, conspicuously many of Fe II, and similar to those of the well known but less luminous Seyfert galaxies. The variable blue compact I Zw 1 is one of the outstanding examples of this class ⁽⁷²⁾.

Absorption Spectra

f. Continua with the Balmer lines, D-lines of Na I, H and K of Ca II, 4226 of Ca I and the G-band in various combinations.

Mixed Spectra

g. Many combinations of the types a) to f). Occasionally H, H, H appear in emission while the higher members, from H to H are in absorption.

As to the widths of the spectral lines, they may all be sharp, all broad or, the forbidden lines are sharp and the permitted lines are broad.

Very broad Balmer lines in absorption have for instance been observed for VII Zw 475 at R. A. 12^h33.7^m and Decl. +81°53' indicating a velocity dispersion of the order of 9000 km/sec for the constituent stars of the system. It is also possible that the set of the H and K lines appears twice in the spectrum of this compact galaxy because of an Einstein gravitational redshift of 10,600 km/sec between the core (19,900 km/sec) and the halo (9300 km/sec). Analogous cases are I Zw 22 at R.A. 9^h55.9^m and Decl. +51°45' , with <V_s> = 25, 270 km/sec for the core and 14,230 km/sec for the halo. Another possible case of this character is I Zw 126 at R.A. 15^h45.9^m and Decl. +37°21' with <V_s> = 31, 300 km/sec for the core and 11,880 km/sec for the halo.

It will be very important to further check these conjectured interpretations of the mentioned spectra which t obtained just prior to my retirement. If the above stated conclusions can be confirmed, the contention, that the large redshifts of the quasars might be partly due to the Einstein effect will in great probability also be found to be true.

It should be added that the existence of smaller gravitational redshifts, of the order of 1000 to 2000 km/sec between the cores and the outskirts of compact galaxies almost certainly has been observed in the spectra of the compacts I Zw 188, I Zw 198, I Zw 206 and I Zw 208.

Finally I here wish to emphasize that all of the above mentioned compacts should by all means be observed with one of the powerful old fashioned instruments, such as the nebular spectrograph of the Palomar Hale telescope. Records obtained with the presently available image tubes are much too poor to show the detailed structural features necessary to check whether or not the differences in the symbolic velocities of recession of the order of 10,000 km/sec are due to the Einstein effect.

Variability of compact galaxies

The first variable compact galaxy, IV Zw 29, located at R.A. 0^h39.5^m and Decl. +40°3' was discovered by Zwicky ⁽⁶⁹⁾ in 1964. It is radio quiet and it also is the first galaxy ever to have been found variable in its output of light. The hundreds of films and plates obtained by Zwicky at Palomar in his search for supernovae show IV Zw 29 to have been variable in the range ⁽⁷³⁾ 16.0 < m_p < 18.2.

Another optically variable but radio quiet compact is I Zw 187 at R.A. 17^h37.1^m and Decl. +50°15'. Its spectrum was found to be featureless by Zwicky ⁽¹⁶⁾ and later on also by Oke and collaborators ⁽⁷⁴⁾.

The most interesting variable compact galaxy investigated so far is I Zw 1 at R.A. 0^h51.0^m and Decl. +12°25', the spectrum of which is similar to that of the Seyfert galaxies, with many emission lines of Fe II showing ⁽⁷²⁾. Its average photographic apparent magnitude, from records of the past forty years, is m_p = 14.3. With a symbolic velocity of recession V_s = 18, 150 km/sec this corresponds to an average indicative absolute photographic magnitude M_p = - 22.0. Outburst lasting from a few days to a few weeks, however, have occurred on several occasions that increased the brightness of I Zw 1 by about one magnitude. At the peak of such outbursts the indicative absolute photographic magnitude was M_p(max) - 23.0. Implosive and explosive events in I Zw 1 therefore produced increases in its luminosity of the order of 10¹¹ times that of the sun, or about thirty times that of a bright supernova at maximum. These findings possibly confirm the conjectures previously advanced by Zwicky ⁽⁷⁾ that in very compact galaxies, with masses of the order of 10¹² M and greater, one or more of the following events will take place.

First, in the most massive compact stellar systems about one thousand head-on collisions between stars moving with velocities of thousands of kilometers per second must be expected, ^** each one of them producing one or more neutron stars and outbursts of light equalling that of several billion suns. Statistically it will also happen that several dozen collisions may occur within a few days and thus give rise to outbursts of light equalling 100 billion times that of the sun, as observed.

Second, the kinetic energy of the gas clouds ejected in a supernova outburst is not usually converted into light. This, however, will happen if a supernova explodes in the interior of a compact galaxy, so that its luminosity is increased manyfold.

Third, collisions between massive and dense interstellar clouds may be responsible for the sudden increases in luminosity such as they have been observed for I Zw 1 and other variable compacts.

It will be of the greatest importance that the variable compacts mentioned here, as well as others, be observed as often as possible with small telescopes, actually a fruitful task for amateurs. Large and sudden outbursts of luminosity should be reported immediately, so that observers with large telescopes can obtain spectra near the peaks of these events and thus collect data that will eventually allow us to confirm or to discard the abovementioned three conjectures.

Encounters and collisions among galaxies

In addition to internal implosions and explosions that produce compact and post-eruptive galaxies, close encounters between galaxies and the resulting gravitational tidal actions are of great importance, as I have emphasized on many ocassions ^{(75, 76)}. Most interesting, however, are the cases in which a compact galaxy, or possibly an object HADES, has pierced another galaxy at great speed, leaving either a track of luminous gas clouds and emission line knots, or an almost straight channel swept clean of its original occupants of stars and gas or dust clouds. In the Appendix I present the "crossed galaxy" at R.A. 23^h15.7^m and Decl. +3°54' as a possible case of this kind.

Objects HADES in the centers of compactclusters of galaxies

It may be conjectured that very compact galaxies may ultimately collapse into configurations of the type of objects HADES H, thereby losing most of their effective mass and luminosity. If this for instance happens to many centrally located members of a cluster of compact galaxies, one consequence will be that the average symbolic velocity of recession of the brightest members of the afflicted cluster will be markedly lower than it would be expected from the average, though admittedly very rough redshift-apparent magnitude relation. One possible interesting case in point is the cluster Zw Cl 0257.8 + 3542 which contains the radio source 4C 35.6. The apparent photographic magnitude m_p = 17.6 for its brightest central member is much too faint for its observed symbolic velocity of recession of 14070 km/sec, which one actually would expect to be of the order of 52000 km/sec.

Also, in some clusters containing many objects H in the center, the brightest galaxies may be found on the outskirts of the cluster rather than in the central regions, thus producing a segregation reverse from that usually found.

Conclusion

Finally it is worth noticing that, the fewer parameters one uses to define a class of objects, the richer this class becomes. Thus compact galaxies have simply been defined as containing more than one star and having an apparent surface brightness greater than a certain critical value corresponding to the 20th magnitude per square second of arc. Because of this broad definition the class of compact galaxies includes all objects (or parts of objects) listed in this catalogue as well as radioquiet and radio emitting quasars, certain Markarian galaxies and many objects as yet to be discovered.

^* Sargent, in his extensive paper on compact galaxies ⁽⁶⁷⁾ unfortunately has incorrectly designated all objects from Zwicky's list No. 1. Starting from I Zw 3 all following I Zwn's in Sargent's paper must be given the original designation I Zw (n+1), thus for instance, I Zw 23 in Sargent's paper is really I Zw 24. Back.

^** per year Back.