zotero/storage/NLU6C682/.zotero-ft-cache

Preface

My purpose in publishing this book is to communicate information which would not otherwise be accessible. About 10 years ago my first book on this subject appeared: Quasars, Redshifts and Controversies. That first book had really been written between 1984 and 1985, but it took a seemingly interminable two years to publish because uncountable numbers of publishers turned it down. One university press, that of my alma mater, was enthusiastic about it until they gave it to a member of the Astronomy faculty to read. Another, Cambridge University Press, declined to publish it, but once it was published bought a large number of copies at very low cost to sell through their distribution. (At least the distribution was a useful step).
Finally, Donald Goldsmith came to the rescue of what I view as academic freedom of communication and published it under the aegis of his small company, Interstellar Media. I felt enormously grateful to him for enabling the observational material to be presented, regardless of what he or any one else felt about the ultimate outcome of the debate. Of course, I was hoping that once all the evidence was correlated and described in a way not allowed by referees, scientists would turn their instruments and analysis to investigating the many crucial objects which contradicted current theory.
Instead, the book became a list of topics and objects to be avoided at all cost. Most professional astronomers had no intention of reading about things that were contrary to what they knew to be correct. Their interest usually reached only as far as using the library copy to see if their name was in the index. But before that disappointment really registered with me, something rather wonderful happened. I started getting letters from scientists in small colleges, in different disciplines, from amateurs, students and lay people. The amateurs in particular amazed and delighted me, because it quickly became clear that they really looked at pictures, knew various objects and reasoned for themselves while maintaining a healthy skepticism toward official interpretations. As an example, Canadian physics students brought me from Europe to address their annual convention. I was stunned when they ushered me into a room where a table was piled high with copies of my book to autograph. I realized that these were books they had bought on their own initiative and with their own money. In the end, the book was translated into Italian and Spanish, and I still hear from people all over the world who are interested in how it is all going to turn out. So regardless of the difficulties and frustrations, and no matter what else happens, I feel that book was the most important and rewarding work I have ever undertaken.

Halton Arp, Seeing Red: Redshifts, Cosmology and Academic Science (Apeiron, Montreal, 1998)

i

ii

Preface

More than 10 years have passed and, in spite of determined opposition, I believe the observational evidence has become overwhelming, and the Big Bang has in reality been toppled. There is now a need to communicate the new observations, the connections between objects and the new insights into the workings of the universe—all the primary obligations of academic science, which has generally tried to suppress or ignore such dissident information. In spite of—or because of—the success of the first book, it is even more necessary now to secure independent and effective publication of these kinds of science books. The present volume is a bigger book with prospects for wider circulation. In consideration of these aspects, with Don Goldsmith’s advice and assistance, I feel fortunate that the present publisher, Roy Keys, is presenting this new work, Seeing Red: Redshifts, Cosmology and Academic Science.
One useful aspect of the present book is that it illustrates what can develop from one simple assumption, such as the nature of extragalactic redshifts. Both sides in the dispute have complex, rather fully worked out views which they believe to be empirically supported and logically required. Yet one side must be completely and catastrophically wrong. It makes one wonder, perhaps with profit, whether there are other uncertain assumptions on which much of our lives are built, but of which we are innocently overconfident.
The present book is sure to outrage many academic scientists. Many of my professional friends will be greatly pained. Why then do I write it? First, everyone has to tell the truth as they see it, especially about important things. The fact that the majority of professionals are intolerant of even opinions which are discordant makes change a necessity. Those friends of mine who also struggle to get the mainstream of astronomy back on track mostly feel that presenting evidence and championing new theories is sufficient to cause change, and that it is improper to criticize an enterprise to which they belong and value highly. I disagree, in that I think if we do not understand why science is failing to self-correct, it will not be possible to fix it.
Briefly, I suppose my view is that science never matured through the “age of enlightenment.” When society at long last learned that major decisions were too important to be left in the hands of kings and generals, a more democratic process was evolved. But science always insisted that only those who possessed arcane knowledge were capable of deciding what was true and what was not true in the world of natural phenomena.
Now we have a situation where new facts are judged by whether they fit old theories. If they do not, they are condemned with the judgment:

“There is no way of explaining these observations, so they cannot be true.”

That encourages the dissident to come up with an explanation of how it could be true. It disagrees with convention. Then the jaws of the trap spring shut and the theory is labeled:

“....prima facie evidence that the proponent is a crackpot and the evidence is false.”

Preface

iii

This, then, is the crisis for the reasonable members of the profession. With so many alternative, contradictory theories, many of them fitting the evidence very badly, abandoning the accepted theory is a frightening step into chaos. At this point, I believe we must look for salvation from the non-specialists, amateurs and interdisciplinary thinkers—those who form judgments on the general thrust of the evidence, those who are skeptical about any explanation, particularly official ones, and above all are tolerant of other people’s theories. (When the complete answer is not known, in a sense everyone is a crackpot—Gasp!).
The only hope I see is for the more ethical professionals and the more attentive, open-minded non professionals to combine their efforts to form a more democratic science with better judgment, and slowly transform the subject into an enlightened, more useful activity of society. This is the deeper reason I wrote this book and, although it will cause distress, I believe a painfully honest debate is the only exercise capable of galvanizing meaningful change.
If there is any credit due for all this, I should mention that when I left the United States in 1984, I came to the Max-Planck Institut für Astrophysik, first on an Alexander Humboldt Senior Scientist award; I then stayed on as a guest scientist. I must acknowledge that if it had not been for the use of the facilities of the Institut, the hospitality, support and friendship of the researchers, I would not have been able to carry out the present work. It was my amazingly good fortune that many of the key, active objects I had observed with the big telescopes on the Pacific Coast were just being observed with the frontier-breaking X-ray telescope at the Max-Planck Institut für Extraterrestrische Physik (MPE). It picked out the most energetic objects with ease, and the telescope was still small enough so that it had a sufficiently large field to include the crucial objects which were related to the central progenitor galaxies.
All of the staff and faculty were enormously kind and helpful. To single out just a few: Rudi Kippenhahn, who initially nominated me for the Humboldt award and arranged for me to stay on afterwards; Hans-Christoph Thomas in the neighboring office, who was always ready to assist me in complex computer problems; and Wolfgang Pietsch at MPE, who taught me what rudiments of X-ray image processing I was able to learn, and showed me his many observational breakthroughs. We all have our precious beliefs, and the greatest courage is to respect a differing belief. Here I found people who believed the way one did science was the overriding ethic, and, with poetic justice, I think it leads to the greatest advances.
The following book is arranged with the first two chapters establishing that high redshift quasars emerge from the active nuclei of nearby galaxies. The next two chapters show that smaller companions of nearby galaxies also have intrinsic (non-velocity) redshifts, which persist down to the stars and gas that make up the galaxy. Chapter 5 discusses how the Local Supercluster is composed of similar groups and types of objects, and shows how their intrinsic redshifts decrease from the quasars down to the oldest galaxies. Chapter 6 introduces the startling evidence that faint groups of highredshift, non-point-source objects on the sky are generally not distant clusters of

iv

Preface

normal galaxies, but instead are more like smaller, intrinsically redshifted components of broken up quasars.
Chapter 7 discusses how gravitational lenses cannot explain the association on the sky between quasars and lower redshift galaxies. It presents arguments that the quasars are not lensed background objects but younger material actually emerging from the central object. Chapter 8 presents the evidence for quantization, a phenomenon that could not occur if redshifts were caused by velocities. Chapter 9 discusses the theory. It points out how the Friedmann/Einstein expanding universe (the so-called “Big Bang”) is based on a mistaken assumption—and why it cannot explain the observations. A more general solution of the basic equations is presented and it is discussed how it predicts the observed creation of quasars and their evolution into normal galaxies.
Finally, Chapter 10 recounts a number of examples where Academic Science has been unable to modify its theories and commitments to accommodate new observational facts. Directions of possible change are briefly discussed.
But the text, I feel, is not as important as the pictures. If non-specialists find parts of the text too technical, it is recommended just to scan through these sections. Actually, the pictures tell the story. One can look at some of the key pictures and simply understand by analogy with everyday experience the important aspects of how objects are related to each other, and how they must develop with time. In fact, the whole book could be reduced to a few pictures in which a person’s ability to recognize patterns and sequences would convey most of the meaningful information. If individuals have confidence in what they “see,” they can live serenely with the knowledge that they do not yet have ultimate understanding.

Introduction

WHY ARE REDSHIFTS THE KEY
TO EXTRAGALACTIC
ASTRONOMY?

Redshifts: If we look at the light from an object after it has been spread out
from short to long wavelengths, we will see peaks and valleys due to emission and absorption from its atomic elements. One thing we can then measure is how much these features are displaced from their wavelengths in a laboratory standard.

It turns out when we observe galaxies and quasars, such features are generally shifted to longer wavelengths, in some cases by amounts up to 4 or 5 times the local laboratory values. This redward displacement of lines in the spectrum is considered to increase with distance and to be the most significant information we have about the faint smudges that are supposed to represent the most distant objects we can see in the universe. But if the cause of these redshifts is misunderstood, then distances can be wrong by factors of 10 to 100, and luminosities and masses will be wrong by factors up to 10,000. We would have a totally erroneous picture of extragalactic space, and be faced with one of the most embarrassing boondoggles in our intellectual history.
Because objects in motion in the laboratory, or orbiting double stars, or rotating galaxies all show Doppler redshifts to longer wavelengths when they are receding, it has been assumed throughout astronomy that redshifts always and only mean recession velocity. No direct verification of this assumption is possible, and through the years many contradictions have arisen and been ignored. The evidence presented here is, I hope, convincing because it offers many different proofs of intrinsic (non-velocity)

Halton Arp, Seeing Red: Redshifts, Cosmology and Academic Science (Apeiron, Montreal, 1998)

1

2

Introduction

redshifts in every category of celestial object—from stars through quasars, galaxies and clusters of galaxies. Moreover, this one key observable will ultimately lead us to consider a universe governed by the non-local effects of inertial mass and quantum mechanics, rather than the local dynamics of general relativity.

Cosmology: Because it concerns our ultimate origins and our future desti-
nies, most people are interested in the nature of the universe in which we live. We call this picture of our environment in its broadest possible sense cosmology.

There is now a fashionable set of beliefs regarding the workings of the universe, greatly publicized as the Big Bang, which I believe is wildly incorrect. But in order to enable people to make their own judgments about this question, we need to examine a large number of observations. Observations in science are the primary and final authority. In the present book I endeavour to discuss these observations in as much detail as necessary to understand them. If the basic data were not so fiercely resisted by conventional cosmologists, the details would not need to be extensively discussed. But as it is, each block in the edifice has to be defended against endless objections. Moreover, the link between many different results is what ultimately gives the whole new picture credibility. The separate observations have to be related to each other, and this takes some patience and effort, although it is exciting to see the pieces fit together in the end. In order to make this process more stimulating, I recount some of the personal and human reactions that accompany these events. This, I hope, will aid the reader in understanding not only the facts, but why they have been received as they have. After all, science is a human undertaking, and people will only read the detailed scientific evidence if someone speaks freely about what it means in the context of real human beings.

Academia: Experts in physical science now are almost exclusively trained in
universities. Our society financially supports theoretical scientists and facilities primarily through the academic hierarchy.

So there is another reason why it is not sufficient to relate just the new factual results. The current beliefs are the crowning achievement of our research and learning institutions, and if they are so completely wrong—and have been for so long in the face of glaring evidence to the contrary—then we must consider whether there has been an overwhelming breakdown in our academic system. If so, we must find out what went wrong and whether it is possible to fix it.
In order to put the pertinent observations into their proper perspective, I present the following table, which gives a loose outline of modern cosmology:

Introduction

3

TABLE I-1: KEY EVENTS IN COSMOLOGY

1911 W.W. Campbell redshifts of OB stars (K effect)

1922 A. Friedmann solution of Einstein’s field equations

1924- E. Hubble 1930

island universes and redshift relation

1948 J. Bolton

double lobed radio sources

1963 Palomar

Quasars

1970’s G. de Vau-
couleurs
1980- Satellite

Local Supercluster X-rays

1990’s observatories Gamma rays

Cosmic Ray Telescopes

Ultra High-energy Cosmic Rays

Future:

Redshift as a function of age Quantization of redshift Episodic creation of matter Mach generalizes Einstein Mass as a frequency resonance

4

Introduction

Key Events in Cosmology—The Theory
It is currently believed that rigorous cosmology started in the early 1920’s after Einstein wrote down the equations of general relativity. These essentially represented the conservation of mass, energy, momentum, etc. in the most general possible coordinate system. In 1922, the Russian mathematician, A. Friedmann, “solved” these equations, i.e., showed how the system would behave in time. It is interesting to note that at first, Einstein felt this solution was incorrect. Later he said it was correct, but of no consequence. Finally he accepted the validity of this solution, but was so unhappy with the fact that it was not a stable solution, i.e., it either collapsed or expanded, that he retained the cosmological constant he had earlier introduced in order to keep the universe static. (This constant was later referred to as the cosmological fudge factor.)
In 1924, Hubble persuaded the world that the “white nebulae” were really extragalactic, and a few years later announced that the redshifts of their spectral lines increased as they became fainter. This redshift-apparent magnitude relation for galaxies became known as the Hubble law (through lack of rigor, often referred to as the redshiftdistance relation). At this point Einstein dropped his cosmological constant as a great mistake, and adopted the view that his equations had been telling him all along, that the universe was expanding. Thus was born the Big Bang theory, according to which the entire universe was created instantaneously out of nothing 15 billion years ago.
This really is the entirety of the theory on which our whole concept of cosmology has rested for the last 75 years. It is interesting to note, however, that Hubble, the observer, even up to his final lecture before the Royal Society, always held open the possibility that the redshift did not mean velocity of recession but might be caused by something else.

Key Events in Cosmology—The Observations
In 1948, John Bolton discovered radio sources in the sky. Martin Ryle, a reigning pundit, argued furiously that they were inside our own galaxy. Of course they turned out to be overwhelmingly extragalactic. The curious thing was that they tended to occur in pairs, and it was soon noticed that there were galaxies between the pairs. I remember the noted experts of the day assuring us that the pairs had nothing to do with the galaxies.
Then radio filaments were found to connect these pairs (they later came to be called radio lobes) to the central galaxies, which were generally weaker radio sources. Without ever raising a glass of champagne, people began to think that they had always known that the radio sources were ejected in opposite directions by some explosive activity in the central galaxy. This fundamentally changed our view of galaxies: rather than vast, placid aggregates of majestically orbiting stars, dust and gas, it became clear that their centers were the sites of enormous, variable outpourings of energy. Probably this change of concept has still not completely sunk in for many astronomers. It is astonishing to note how closely the X-ray pairs across galaxies now correspond to the ejected radio pairs, and how stubbornly people refuse to accept them as ejected.

Introduction

5

Fig. I-1. The radio galaxy Cygnus A showing ejection of high energy, radio emitting material in opposite directions from the central object. This map was measured at 5 GHz with the Very Large Array at Socorro, New Mexico by Rick Perley.
Figure I-1 here shows a radio image of one of the first double-lobed radio galaxies discovered, Cygnus A. Seeing the thin jets leading out into the swept-back lobes leaves no room for doubt that this is a result of ejection from the central object. Something initially small and associated with radio emission has had to come out from the center of this galaxy. Quasars are also often radio sources, and many examples will be shown in this book of pairs of high redshift, radio- and X-ray emitting objects, obviously ejected from active central galaxies. The reason, of course, for the rejection of the pairing evidence for quasars is the now-sacred assumption that all extragalactic redshifts are caused by velocity and indicate distance. The association must be denied because the quasars are at much higher redshift than the galaxies from which they originate.
Quasars
In 1963, some radio sources which had been identified with apparent stars were being studied spectroscopically. What were puzzling stellar spectra, however, suddenly turned out to be emission line galaxy spectra shifted to very long wavelengths. There was some hesitation at first about accepting these redshifts as due to recession velocities that approached the speed of light, since this would indicate great distance. At their redshift distances, these objects had to be 1000 (and in the end 10,000) times brighter than previously known extragalactic objects. But no other redshifting mechanism was deemed likely, and everyone soon got used to these extraordinary luminosities.
Although the radio positions came from various observatories, the spectroscopic identification was done mostly at the Palomar 200-inch reflector. I was observing at

6

Introduction

Palomar at the time, but the positions were distributed privately. So I instead undertook a multi-year study of peculiar galaxies with the aim of studying how galaxies were formed and evolved. When the Atlas was complete, I discovered that across my most disturbed peculiars were pairs of radio sources. Very nice. Obviously the disturbance had been caused by the ejection of the radio sources. Then came the shock: some of radio sources turned out to be quasars! And the galaxies were not at great distances, but relatively close by.
Suddenly it is 30 years later; I am living in Germany and observing by satellite (computer processing the data) and writing about all the exciting new pairs of X-ray quasars across active galaxies which are being discovered by the ROSAT satellite telescope whose headquarters are located in the institute next door. There is only one little flaw in my idyllic good fortune, and that is that there is a relentless effort to ban all these lovely new observations from conferences and suppress them from publication. The compensation is that a few courageous, and officially disparaged, scientists are meeting and communicating with one another to explore the fundamental meaning this new information holds out to us.

High-energy Radiation and the Local Supercluster
Since the 1980’s not only have satellite telescopes been telemetering down X-ray data, but more recently higher energy gamma-ray data has been gathered, and now ultra high cosmic ray detectors on earth have been reporting even higher energy radiation. In a separate chapter to follow, we will discuss the concentration of this energy at the center of our Local Supercluster and its possible meaning. But first, we should briefly describe the Local Supercluster because, contrary to common belief, this may be the only region of the universe we know much about.
The empirical results of galaxy catalogues were already showing in the 1950’s that galaxies were not distributed uniformly over the sky. Yet the analyses by Gerard de Vaucouleurs showing the distribution along the supergalactic equator and the concentration around the Virgo cluster at the center were privately ridiculed, until suddenly in the early 1970’s everyone discovered that they had known about it all along. It turns out that we will find the oldest galaxies there—and the most energetic radiation—perhaps pointing to current matter creation. Virgo may thus be a very special place in terms of understanding what we can currently see of our universe.

Future Events
At the bottom of Table I-1, some current investigations are listed. The investigation of redshift as a function of age already started in the early 1970’s; quantization of redshifts shortly thereafter; and the creation of matter, perhaps in the 1980’s. Since even the existence of these effects is not accepted at present; we can only say that they are epochal science in the making, if they are someday accepted.
Quantization of redshift and episodic creation of matter combine to offer the most promising empirical understanding of extragalactic objects, as explained in the

Introduction

7

following chapters. As a capsule preview of how galaxies are born, we can say that they are ejected from older galaxies as compact objects with low particle masses. As these newer galaxies age, and grow in size and mass, they in turn eject newer generations in a cascading process. We can actually show in Chapter 8 how groups of a dozen or so active quasars fraction into more and more objects, which in turn eventually evolve into clusters of large numbers of galaxies. The redshifts, which are very high as the newly created matter emerges from its zero-mass state, continue to diminish as the mass of the matter grows. Discrete steps in the redshift values are present throughout, but grow smaller when the overall redshift grows smaller. These aggregates of matter develop into normal galaxies, much like our own and those around us in the Local Group and Local Supercluster. All of this is almost diametrically opposed to the conventional view of galaxies condensing out of some tenuous, homogeneously pervading hot gas. It is a process that is going on in our own Local Supercluster, and, contrary to what is claimed by the Big Bang theorists, we do not know much about what may exist at cosmic distances. It turns out that for what we currently see, but do not understand, the essence is in the changes it is undergoing.
The final possibilities for a more fundamental understanding of the nature of matter as a function of frequency and time will be discussed at the end of the book. A complete understanding might be the ultimate reward for a careful analysis of all the observations. It is clear, however, that if we are to make progress in this area, we cannot wait for establishment science to, perhaps, someday accept the empirical results.

The Stars in 1911
When the first telescopes were being built under clear skies and systematic spectroscopic observations started—for example with the 36-inch refractor at Lick Observatory on Mount Hamilton—it was natural to observe what one could. That meant bright stars. One of the things that could be measured accurately was line shifts in stellar spectra. As the data accumulated, it was noticed that the bright blue (OB) stars, the hot luminous stars, had lines which were slightly, but significantly, shifted to the red. In 1911, the Director of the Observatory, W.W. Campbell gave the enigmatic name “K effect” to the phenomenon. (Actually K represented the expansion term in the formula that described the motion of all the stars measured.)
Since all the other stars in our galaxy moved together in reasonable ways, it was not concluded that we lived at the center of an expanding shell of OB stars. The effect was unexplained until the 1930’s, when Robert Trumpler again found the effect in clusters of young stars in our galaxy. He thought he could explain it with a gravitational redshift at the surface of these hottest, most luminous stars. But that failed when the surface gravity turned out to be too weak. Later Max Born and Erwin Finlay-Freundlich tried to explain it with tired light. But that did not catch on. So the observations were again buried and forgotten.
I think it is a supreme and delicious piece of irony that 85 years after the Director at Lick Observatory announced the K effect Margaret Burbidge, a senior professor at

8

Introduction

the University of California, went up Mount Hamilton on a winter night to that same Lick Observatory. She observed two quasars that all the biggest and most advanced telescopes in the world had deliberately refused to look at, and in so doing, solved the riddle of the K effect—and at the same time laid the last flower on the grave of Big Bang cosmology.
Looking back now, especially from the standpoint of the coming chapters, we can see that if the relativists had heeded the published observations, going back a decade before their theoretical revelations, perhaps they would have decided that the universe was not necessarily exploding away from us in all directions.
My career at the Observatories in Pasadena slightly overlapped Edwin Hubble’s. He personally gave me my first job: to aid in determining the crucial distance scale in cosmology. As a result I lived for two years on Mt. Wilson measuring novae in the Andromeda Nebula (M31). I moved on to observe Cepheid variables from South Africa and, finally, am now presenting evidence for a much different, perhaps truer, distance scale at greater distances derived from quasars and young galaxies.
In his seminal book Realm of the Nebulae Hubble wrote: “On the other hand, if the interpretation as velocity shifts is abandoned, we find in the redshifts a hitherto unrecognized principle whose implications are unknown”. In the ensuing years the evidence discussed in the present book has built up to the point where it is clear that the velocity interpretation can now be abandoned in favour of a new principle which stands on a firm observational and theoretical foundation.
After about 45 years, I now know that if the academic theoreticians at that time had not forced his observations into fashionable molds, we might at least not have started off modern cosmology with the wrong fundamental assumption. We could be much further along in understanding our relation to a much larger, older universe—a universe which is continually unfolding from many points within itself.

Chapter 1

X-RAY OBSERVATIONS CONFIRM INTRINSIC REDSHIFTS

J ust another isolated case. Your eye slid over that phrase because you wanted to see whether the referee was going to recommend publication. The answer was: not for the Astrophysical Journal Letters. The message behind the smooth, assured phrases was clear: “No matter how conclusive the evidence, we have the power to minimize and suppress it.”
What was the evidence this time? Just two X-ray sources unmistakably paired across a galaxy well known for its eruptive activity. The paper reported that these compact sources of high-energy emission were both quasars, stellar-appearing objects of much higher redshift than the central galaxy, NGC4258. Obviously, they had originated from the galaxy, in contradiction to all official rules. Slyly, the referee remarked that “because there was no known cause for such intrinsic, excess redshifts the author should include a brief outline of a theory to explain them.”
My mind flashed back through 30 years of evidence, ignored by people who were sure of their theoretical assumptions. Anger was my only honest option—but stronger than that provoked by many worse “peer reviews” because this was not even my paper. I did not have to stop and worry that my response was ruled by wounded personal ego.
How did this latest skirmish begin? Several years earlier an X-ray astronomer had come into my office with a map of the field around NGC4258. There were two conspicuous X-ray sources paired across the nucleus of the galaxy. He asked if I knew where he could get a good photograph of the field, so he could check whether there were any optical objects which could be identified with the X-ray sources. I was very pleased to be able to swivel my chair around to the bookshelves in back of me and pull out one of the best prints in existence of that particular field. I had taken it with the Kitt Peak National Observatory, 4-meter telescope, about a dozen years previously. It

Halton Arp, Seeing Red: Redshifts, Cosmology and Academic Science (Apeiron, Montreal, 1998)

9

10

X-Ray Observations Confirm

Fig. 1-1. The Seyfert galaxy NGC4258 is known to be ejecting material from an active nucleus. A deep photograph is shown here with contours of X-ray emission superposed (W. Pietsch et al.) The two point sources of X-rays on either side of the nucleus coincide with blue stellar objects (BSO’s).
was very deep, because I had been searching this active galaxy for low surfacebrightness ejection features and associated high redshift objects.
Wolfgang Pietsch quickly found a small pointing correction to the satellite positions and established that his X-ray pair coincided with blue stellar objects at about 20th apparent magnitude. (Figure 1-1) At that instant I knew that the objects were almost certainly quasars, and once again experienced that euphoria that comes at the moment when you see a long way into a different future. In view of the obvious nature of these objects I felt Pietsch showed courage and scientific integrity in publishing the comment: “If the connection of these sources with the galaxy is real, they may be bipolar ejecta from the nucleus.”
Then the dance of evasion began. It was necessary to obtain optical spectra of the blue stellar candidates to confirm that they were quasars and ascertain their redshifts. A small amount of time was requested on the appropriate European telescope. It was turned down. Pietsch’s eyes avoided mine when he said “I guess I did not explain it clearly enough.” The Director of the world’s largest telescope in the U.S. requested a brief observation to get the redshifts. It was not done. The Director of the X-ray Institute requested confirmation. It was not done. Finally, after nearly two years, E. Margaret Burbidge with the relatively small 3 meter reflector on Mount Hamilton, on a

Intrinsic Redshifts

11

Fig. 1-2. Spectra of the two X-ray BSO’s across NGC4258 with the Lick Observatory 3 meter telescope taken by Margaret Burbidge showing the similarity of the quasars.
winter night, against the night sky glow from San Jose, recorded the spectra of both quasars. It was fortunate that mandatory retirement had been abolished in the U.S., because by this time Margaret had over 50 years of observing experience.
Of course, the referee report from which I quoted was directed against her paper, which reported this important new observation. In her firm, but lady-like English way,

12

X-Ray Observations Confirm

Margaret withdrew her paper from the Astrophysical Journal Letters and submitted it to the European journal Astronomy and Astrophysics Letters.
What was particularly appalling about this series of events was that Margaret Burbidge was someone who had given long and distinguished service to the scientific community. Professor at the University of California, Director of the Royal Greenwich Observatory and President of the American Association for the Advancement of Science among other contributions. It seems it was permissible to let her fly anywhere in the world doing onerous administrative tasks, but her scientific accomplishments were not to be accorded elementary scientific respect and fair treatment.
Some would argue that this is a special case, owing to the climate of opinion where the offices of the Astrophysical Journal Letters are located. But, as events in the following chapters make clear, the problem is pervasive throughout astronomy and, contrary to its projected image, endemic throughout most of current science. Scientists, particularly at the most prestigious institutions, regularly suppress and ridicule findings which contradict their current theories and assumptions.
Since scientific research in the end is almost completely supported by public funds, it behooves us as citizens to be aware of whether this money is spent wisely in relation to the real needs of, and possibilities for, the future of the society. The central purpose of this book is to explore this topic, and we will return often to it. But in the case at hand, the greatest progress can be made by discussing the actual observations of how nature works and the ways which science often misinterprets and misrepresents them.

Just another Experimentum Crucis—NGC4258
The referee’s unconscious satire, “just another isolated case” was accompanied by deprecatory remarks such as “the quasars are not that well aligned” and “they are not exactly spaced across the nucleus of the galaxy.” Of course a normal person would simply glance at the pair of X-ray sources across NGC4258 and realize they were physically associated. The average astronomer, however, would look at them and start to argue that they must be accidental, because astronomers now feel compelled to fit the observations to the theory and not vice versa.
Consequently, to head off the derogatory rumors which pass for scientific evaluation, someone had to compute a numerical probability. Basically, this meant computing how dense the associated X-ray sources are on average over the sky at a given apparent brightness. Then I had to ask myself: What is the chance of a source of a given brightness falling this close to an arbitrary point in the sky? Given the chance that the first one falls accidentally as close as the measured distance, then one must multiply by the chance that the second one falls at its observed distance. (I.e. if one out of ten will have a source as close as the real case then only one out of 100 will have two such sources.) For the two sources across NGC4258 it turns out that this chance is 5 × 10–2 (i.e. five chances in one hundred). Of course, this does not include the improbability that they would be aligned across the nucleus of NGC4258 to within 3.3 degrees out of a possi-

Intrinsic Redshifts

13

Fig. 1-3. NGC4258 photographed in the light of hydrogen alpha emission showing excited gas emerging from the nuclear regions (P. Roy et al.).

ble 180. Nor does it include the improbability that they would be so equally spaced across the nucleus. Nor does it include the similar strengths and energy distributions of the two sources (which would not be expected from random, unrelated sources). Altogether the chance of this pairing of X-ray sources across NGC4258 being accidental is only 5 × 10–6 (5 chances in a million).
When it was confirmed that both sources were quasars, their redshifts became available. It was immediately apparent to anyone experienced with quasar spectra that these two were unusually similar. (Figure 1-2) A conservative probability for this similarity can be estimated at 0.08. Therefore the total probability of this association, being accidental, became < 4 × 10–7 (less than 4 in ten million).
Scientists claim that for acceptable scientific rigor, numerical probabilities must be calculated. But no matter how intimidatingly complex the calculation, no matter how small the probability of accident may be, the calculation does not tell you whether the result is true or not. In fact, no matter how significant the number is, scientists won’t believe if they don’t want to. When I submitted the paper with the calculations on NGC4258 which were claimed to be scientifically necessary, it was not even rejected, just put into an indefinite holding pattern and never acted upon to this day.
In the case of NGC4258, however, most astronomers overlooked a very important fact—that this galaxy is not just another object in a sea of identical objects. It is one of the most active nearby spiral galaxies known. In fact, in 1961 when the French astronomer G. Courtès discovered glowing gaseous arms emerging from the center of its concentrated Seyfert nucleus (see Figure 1-3), it led to observations with the Westerbork radio telescope which revealed that these proto spiral arms were also sources of synchrotron radiation (high-energy electrons spiraling around magnetic lines of force).

14

X-Ray Observations Confirm

Fig. 1-4. Spots of water maser emission in the innermost nucleus of NGC4258 (M. Miyoshi et al.) showing approximate alignment in direction of quasars and redshift differences of the order of ± 1000km/sec correlated with the ejection velocities of the quasars.

In the past I had argued with Jan Oort, the discoverer of rotation in our own galaxy, about whether spiral arms were caused by opposite ejections from active nuclei. NGC4258 was the only case in which he ever admitted that proto-spiral arms were being ejected from the center.
The simplest and most obvious conclusion was that the pair of X-ray quasars was also being ejected from this unusually active galaxy. Interestingly enough, shortly after the discovery of the quasar pair, it was discovered that water masered (emission from H2O molecules) spots in the inner .008 arc seconds showed redshift deviations of plus and minus ~1000 km/sec from the redshift of the nucleus. A conventional model explained this to be caused by a rotating black hole of 40 times greater mass than even the largest previously hypothesized. But just a glance at the observations (Figure 1-4) showed what instead looks like entrained material pointing out in either direction roughly toward the quasars. Note the quantitative agreement with the conclusions of van der Kruit, Oort and Mathewson (Astronomy and Astrophysics. 21, 169, 1972): “...clouds expelled from the nucleus in two opposite directions in the equatorial plane about 18 million years ago, at velocities ranging from about 800-1600 km/sec.”
Even if the conventional hypothesis of black holes were tenable, and the hypothesized mechanism of bipolar ejection valid for them, the observation would still testify to the extreme activity of NGC4258, and thus support the association of the quasars. I personally prefer the concept of a “white hole”, a place things irreversibly fall out of rather than into. For me, the whole lesson of the Atlas of Peculiar Galaxies was that galaxies are generally ejecting material. The merger mania seems to be a first guess based on a cursory look at galaxies. But I also think that the observations are not yet detailed enough to suggest a specific mechanism of ejection. Instead, the startling evidence of the association of high redshift quasars with low redshift galaxies needs to be faced. Those observations are more likely to lead to an understanding of the ejection mechanisms when responsibly pursued.
Of course, the evidence of association has been implacably rejected for 30 years by influential astronomers. In the case of NGC4258 just described, the chance of accidental association is only one in 2.5 million! A reasonable response would be to notice such a case and say, “If I see a few more cases like this I will have to believe it is real.” Most astronomers say, “This violates proven physics [i.e. their assumptions] and therefore must be invalid. After all, no matter how improbable, it is only one case.”

Intrinsic Redshifts

15

Then, when they see another case they treat it de novo and reject it with the same argument. Professional scientists, however, have a responsibility to know about previous cases. And they do. When they block them out, it is a clear case of falsifying data for personal advantage—a violation of the primary ethic of science.
In a more general perspective, it can be said that the unique capability of human intelligence is pattern recognition. It is the most difficult task for a computer to perform. When one thinks about it, indeed, the seminal advances in science, and perhaps human affairs in general, were made by recognizing patterns in natural phenomena.

Other Cases of Galaxy-Quasar Associations
In order to forestall this argument about NGC4258 being “just another isolated case,” I realized that it was more or less up to me to try to publish a paper which established its relation to other similar cases. It would also be necessary to calculate numerical probabilities in each case. As mentioned before, this is an obligatory exercise that critics do not like to do themselves, but insist on in any discovery paper. After the ritual argument about statistics is finished—is it 10–5 or 10–6?—the argument is sufficiently abstract that people who wish to disbelieve the result can ignore it, when in fact it would be embarrassing to ignore it as a straight judgment call from looking at a picture.
Of course, the most important purpose was to gather together more examples of the same kind of pairing. That should clinch the issue. Table 1-1 is reproduced here from the paper that never appeared in Astronomy and Astrophysics. Already at that time, it showed five cases involving only X-ray pairs of quasars around lower redshift objects, each with a chance of less than one in a million of being accidental.

PG1211+143
One of the cases in Table 1-1 became available in the following way: During the time it took for the redshift measurements on the NGC4258 quasars to unfold, I traveled to the National Radio Observatory and gave a talk. Afterwards, Ken Keller-

Table 1-1. Some X-ray Pairs Across Galaxies

Central Gal zG

r1

r2

Δθ

Fx,1×10–13 Fx,2×10–13

z1

z2

p1

ptot

NGC4258 .002 8’.6 9’.7 3°

Mark205

.07 13’.8 15’.7 44°

PG1211+14 .085 2’.6 5’.5 8°

3

NGC3842

.02 1’.0 1’.2 33°

NGC4472 .003 4°.4 6°.0 1°

1.4 cgs 2.3 .2
1.0 ~3000

0.8 cgs 2.7 1.4

.40 .64 1.28

.65 .46 1.02

5 × 10–2 < 4 × 10–7 2 × 10–2 connected 1 × 10–2 < 10–6

.3 ~800

.95 .34 7 × 10–5 6 × 10–8 .004 .16 2 × 10–4 < 10–6

Subscript 1 designates nearest source; Δθ represents accuracy of alignment; Fx’s are estimated for .4 – 2.4 keV band except last entries which refer to M87 and 3C273 and are HEAO 1, 2–10 keV band; p1 designates
accidental probability of finding the sources of strength Fx at r1 and r2; 1 – ptot gives estimated probability of physical association

16

X-Ray Observations Confirm

Fig. 1-5. X-ray map of the Seyfert/quasar PG1211+143. The X-ray BSO to the west was confirmed as a quasar by cooperative effort between Beijing Observatory and Indian Astrophysics Institute in Pune. Solid line shows this pair of quasars coincides with line of radio sources which would be conventionally accepted as having been ejected.

mann came up to me and said, “Here is a bright quasar which appears to have a line of radio sources passing through it—and one of the radio sources is a higher-redshift quasar.” As soon as I returned to my office in Munich I asked my friend and computer expert in the next office, H.C. Thomas, to show me how to search the archives for any X-ray observations of this object. (After one year all proprietary observations are put in a public archive—but considering the amount of specialized knowledge one needs to access these records, the term “public” is rather euphemistic.)
An observation was found, and I eagerly reduced the approximately 4 megabytes of data to form an X-ray picture of the field. As Figure 1-5 shows, the central object is strong in X-rays, and the radio quasar to the East is conspicuous. But most electrifying, there is the hoped-for, strong X-ray source just on the other side, to the West. I immediately went to the Sky Survey photographs and found that this latter X-ray source coincided with a blue, stellar appearing object (BSO). Another pair of quasars across an active object! And this one was aligned with radio sources which by now are accepted as ejected from active galaxies!
But now the same old problem, how to obtain a confirming spectrum and get the redshift? Big observatories were obviously out of the question. The quasar candidate was rather bright, however, and it probably could be observed with a smaller telescope. I sent the finding charts to Jayant Narlikar, Director of the Inter University Center for Astronomy and Astrophysics in Pune, India. He interested a young researcher in obtaining a spectrum with the 1 meter Vainu Bappu telescope. The observation was scheduled in April, however, and the monsoon moved in. Despair—it was gone for the year!
Jayant said that they had asked the Beijing Observatory to do it, but I did not take that seriously, because to my knowledge, China did not have adequate equipment. It

Intrinsic Redshifts

17

turns out, however, there is some reason to look forward to e-mail, because a month later I was delighted to receive a message that the spectrum had been obtained by the Chinese. After some normalizing of photon counts, it was possible to derive a redshift of z = 1.015.
The confirmation of this X-ray BSO as a quasar was particularly compelling, because PG1211+143 had been noticed as a result of its having a line of radio sources across it. As we have discussed, flanking radio sources are customarily interpreted as arising in ejection processes. How else could this pair of X-ray quasars, along exactly the same line, have arisen?
The numerical value of this redshift also turns out to be an important result. When included in Table 1-1, it showed that the difference of redshifts between the quasars in the first three, best pairs was .25, .18, and .26. In other words, interpreting the quasars as ejecta, the projected ejection velocities should be .082c, .058c and .060c, in km/sec. The coincidence of three independent determinations giving closely the same ejection velocity is very encouraging for this interpretation. (Velocities can only be added as in (1 + zi)(1 + zv) = (1 + zt) where i = intrinsic, v = velocity and t = total, as described in Chapter 8.) For an average projection angle of 45 deg., this gives an average true ejection velocity of .094c or 28,200 km/sec.

Radio Pairs from 1968
Several months after submitting this result, and in the midst of dealing with the usual hostile referee and nervous editor, I recalled a surprising fact. Back in 1968 I had investigated pairs of radio sources in the sky, some of which had turned out to be quasars.* From the estimated age of conspicuous disturbances in the central galaxy, and the measured separation of the quasars from their galaxy of origin, I had calculated ejection velocities of .1c. In fact, I had calculated ejection velocities only five years after quasars had been discovered by a completely different method which now agreed well with the new measures!
Of course, even so early in the game, such a storm had been raised against local quasars that there was no chance of publishing in a normal journal. As a result, I had published in the Journal of the Armenian Academy of Sciences, Astrofyzika. Viktor Ambarzumian was a hero of science in Armenia. We agreed on his initial insight that galaxies were formed by ejection from older galaxies. He did not believe my evidence at that time that redshifts were not velocity indicators. But as a tribute to his fairness, he did not hesitate for a moment to welcome my paper. The 12 figures in that paper are dramatic proof that the X-ray results of 1994 had been predicted in detail by the radio quasars in 1968. The paper was also testimony to the fact that sensible analysis of observations was being blocked and ignored, while the high profile journals were submerged with a flood of elaborations of incorrect assumptions which prevented anyone from remembering anything important for more than a few years.

* This reversed the original discovery procedure of 1966. Instead of finding pairs of high redshift radio sources across disturbed galaxies, I looked for pairs of radio sources on the sky and then looked to see whether there were galaxies between them. Of course, many of the radio sources in the pairs turned out to be quasars.

18

X-Ray Observations Confirm

Fig. 1-6. The two strongest radio sources in the pictured area fall across the disturbed spiral galaxy IC1767. The redshifts of these radio quasars at z = .62 and .67 are so close as to confirm their physical relation. (H. Arp, Astrofizika, 1968).

Figure 1-6 here shows a quasar pair from that 1968 paper. The pair involves the brightest radio sources in the field, and is so conspicuous that it is difficult to entertain any idea that it is an accident. Then, of course, there is the disturbed galaxy IC1767 falling at the center of the pair: how likely is that to be an accident? I did not know the redshifts when this pair was published in Astrofyzika but they were subsequently determined to be .67 and .62. Finally, out of a possible range for radio quasar redshifts from .1< z <2.4, what is the probability of getting two unrelated quasars to have redshifts within .05 of each other? This result was then published in Astrophysical Journal, but with the same lack of result. In the face of 28 years of accumulated evidence, to go on proclaiming that quasars are out at the edge of the universe seems unpardonable.
Markarian 205
The second entry in Table 1-1, which yields a projected ejection velocity of .058c, is from a famous and controversial association of a quasar-like object (Mark205) with a violently disrupted spiral galaxy (NGC4319). It is featured in color on the cover of my book Quasars, Redshifts and Controversies, and the long campaign to disprove the connection between the two objects is described therein. The connection was first shown in 1971, but as late as August 1995, there was still an exchange of letters in Sky and Telescope, in which one of the original disputants continued to claim the bridge did not exist.
The observations listed in Table 1-1 involve two new quasars and connections which were discovered in 1994. In 1990, the Max-Planck Institut für Extraterrestrische Physik (MPE) launched the X-ray telescope ROSAT (Röntgen Observatory Satellite Astronomical Telescope). Actually the telescope, a superior work of engineering, was launched by Delta rocket from Cape Kennedy for which (plus one instrument) the U.S. claimed 50% of the observing time, leaving Germany with 38%, and Britain with 12% for building a small ultraviolet camera. Observing time was assigned to proposals from each country by allocation committees in that country. By an enormous stroke of good

Intrinsic Redshifts

19
Fig. 1-7. X-ray filaments emerging from the Seyfert galaxy Markarian 205 and ending on quasars of redshift z = .46 and .64 (very similar to the z = .40 and .65 pair across NGC4258). This observation is shown on the front cover of this book and also in color Plate 1-7.

fortune, I was then a member of a German Institute and could submit proposals to the German selection committee. Even though I had been in Europe for four years, I still heard from friends in the U.S. how my previous requests for time on ground based telescopes and current space telescope requests had fared. Some secondhand accounts reported intense anger and ridicule expressed from the select group of the most reputable (but generally anonymous) astronomers who comprise the U.S. allocation panels.
My proposals to the German committee were rated very low, but at least the case was not hopeless. On the first few schedules I only received time for one very harmless proposal. But some of the experts at the MPE, located next door, were very helpful in preparing the proposals in the most acceptable possible format, and in later scheduling periods I received time on some “hot” objects. One of these was Mark205.
The proposal was to see if the connecting bridge from NGC4319 to Mark205 showed up in X-rays. As is often the case, the major aim failed. (I think now the connection to the galaxy is too old to show well in high-energy). But what did show up was two X-ray filaments coming out of either side of Mark205 and ending on point-like X-ray sources (Front cover of this book, Figure 1-7 and color Plate 1-7)! I immediately got out the sky survey prints and superposed scaled X-ray maps to see if they were optically identifiable. Lo and behold! They were not only blue stellar objects, but unusually bright in apparent magnitude.
Of course, they were quasars: but how to get the requisite spectroscopy which would give the redshifts? The same old problem, all the telescopes were occupied studying distant high redshift objects. Then a routine check of catalogued quasars bore unexpected good fortune. It turned out that a team of researchers had previously investigated fields around strong X-ray objects and found an excess of sources around Mark205. The excess sources turned out to be mostly higher-redshift quasars, but they rejected the significance of this on the basis that Mark205 had been previously known to be in an active region (!??!). I could almost forgive them for that inverted logic,

20

X-Ray Observations Confirm

because I was so happy to see the spectroscopy of the sources in the field. It turned out that there were three(!) confirmed quasars in the X-ray filament that I had discovered coming out of Mark205 in the ROSAT observations. The two major ones at the ends of the two filaments are listed in Table 1-1 with their redshifts, which yield a projected velocity of .06c each (average deprojected velocity .08c). As previously remarked, this now becomes a very important confirmation of the ejection velocities computed for radio quasars 27 years earlier.
But, of course, the stunning aspect of the ROSAT observations was that two quasars of redshift .63 and .45 are actually physically linked by a luminous connection to a low redshift object of z = .007. When I showed this to the local experts, there were alarmed stares followed by annoyance. “Of course, if you go faint enough you will find noise features or instrument imperfections which connect everything together.” The frightening aspect of this reaction was that they were saying: “If the connection between these objects cannot be attributed to noise, there must be something wrong with the instrument.” The latter possibility, even the mention of it, is enough to freeze any member of a well-funded project in his tracks.
Of course, I made the argument that since the filaments from Mark205 were sufficiently broad, coherent features, they obviously could not be noise. I also reduced an exposure of a bright X-ray star in the same way as Mark205, and showed that the faintest levels exhibited no imperfections, but just broke up into random noise as expected. Any non-expert would simply have reasoned that instrumental defects would not likely originate just from an active object, and certainly there would be no reason for them to end on the quasars in the field.
Nevertheless, it was clear that the best possible presentation of the data needed to be communicated. The communication was not easy. Both initial collaborators opted out, because I mentioned the word “ejection” in connection with the filaments ending on quasars. This was just before the word was mentioned in connection with the pair of X-ray sources across NGC4258, which later turned out to be quasars. Actually I became somewhat worried that the pair across Mark205 was not better aligned. In attempting to account for this I pointed out that the connecting filaments started out from Mark205 in initially opposite directions, but that the N one then curved over to the quasar in the NW. It was not until a few years later that I realized the Narlikar/Das model of ejected quasars, which required the increasing mass of the ejected object to slow its initial high velocity, fitted the X-ray observations around Seyfert galaxies very well. Then the light went on: the N quasar on its way out had been gravitationally attracted to the companion galaxy NW of NGC4319 which had swung it around in the observed direction.
But the referee complained because the data tables were not arranged in a certain order, and the objects were not discussed in a certain sequence, and it had not been “proved” that the connections and extensions were not noise. The inevitable ritual was upheld, and the paper was stalled indefinitely.

Intrinsic Redshifts

21

The IAU Symposium
Fortunately, the International Astronomical Union (IAU) was holding its threeyearly meeting in Holland in August 1994. A four-day symposium on Examining the Big Bang and Diffuse Background Radiation had been appended. Now my participation was always a matter of doubt, but this time not enough members of the organizing committee spoke against it to prevent my being invited to give a short paper. I realized I could cram most of the important new observational data on the new cases of X-ray quasars associated with low redshift galaxies into the five pages of a camera ready paper. Even though it would take more than a year to appear in the little-read Proceedings, it was at least a publication to which interested researchers could be referred to see the vital pictures of the actual X-ray data.
Returning early from the peace of the family vacation in the French Alps, I picked up my transparencies and diagrams and headed off to entertain the power elite with deliciously forbidden “crackpot ideas.” (The establishment always confuses data with theories.) There were a few other dissidents in attendance to whom it was very important to communicate the new observations. Jayant Narlikar gave a rigorous presentation of how, near mass concentrations, new matter could be “created” in the vicinity of old matter. Geoff Burbidge gave his usual pungent update of the evidence that some quasars were much closer than their redshift distance.
The symposium relentlessly advanced toward one of its high points. The customary authority on extragalactic theory was scheduled to give the inevitable summary of the present state of knowledge. It always pained me that even though everyone knew what was going to be said, it was given the better part of an hour, whereas the new observations which destroyed the premises and conclusions of the talk never had enough time to be presented in 15 or 20 minutes (and usually not at all). Clearly, the main purpose of these “review of the theory talks” was to fix firmly in everyone’s mind what the party line was so that all observations could be interpreted properly.
The reviewer of choice was naturally Martin Rees—recently having glided effortlessly from Plumian Professor to Astronomer Royal of England. After the standard defense of the Big Bang (even though it did not need defending) the only substantive comment from the audience was from the perceptive veteran Prof. Jean-Claude Pecker. He pointed out inconsistencies in the use of galaxy evolution as an adjustable parameter in order to avoid unexpected behavior with redshift in the Big Bang.
The final day consisted of a panel of about 9 members picked to represent the range of topics covered during the symposium. Facing the audience on the extreme right was Martin Rees, middle-left Geoffrey Burbidge and on the extreme left, myself. Rees opened up with a strong attack on the observations I had shown in my short talk a few days previously. When it came my turn to make an opening statement, I showed even more startling observational images that contradicted conventional models. The discussion was then thrown open to the rather large audience and a Dutch journalist, Govert Schilling, rose to ask Martin Rees a question.

22

X-Ray Observations Confirm

The question, roughly paraphrased, was: “In view of the evidence Dr. Arp has shown, why have not major facilities been used to further observe these objects?” Martin turned toward me and erupted in a vitriolic personal attack. He said I did not understand the evidence from superluminal motions, that I did not believe the age of galaxies, plus a number of other elementary failings. I was rather stunned by the vehemence of this response, and I suppose the audience was also. After a moment or so, I replied that superluminal velocities were not a problem if you put things at their correct distances, and that I, of all persons, should believe the ages of galaxies, because as a graduate student I had measured the countless stars in globular clusters which helped establish the only age we have for galaxies. But most important of all I said, “I feel it is the primary responsibility of a scientist to face, and resolve, discrepant observations.”

An Amateur Observes Mark205
What apparently set off Rees in response to the journalist’s question was that it had been mentioned that an amateur had observed the NGC4319-Mark205 connection with the Hubble Space Telescope. Since 1971 this had been considered a crucial object in the proof of discordant redshifts of quasars, and in the symposium I had shown new evidence for the association of further, higher-redshift quasars with the same system. Because the Space Telescope was reputed to be able to answer all questions, many people had urged us to observe this key object again. Jack Sulentic, a long time collaborator on this project, and I prepared a complex, time consuming observing proposal— the kind that automatically sifts out the outsiders. It was not only turned down, but savaged by the allocation committee. So much for that exercise in futility. I was informed later in a letter that “it was NASA’s policy not to release the names of scientific assessment panels.” My first image was of my colleagues in false beards and dark glasses sneaking into the meeting room. Then a less humorous thought occurred to me—that large amounts of public money were being handed out by a secret committee.
It was not long before a delightful story started to circulate. The Space Telescope administrators had decided to make 10% of the time available to the community of amateur astronomers. This is actually a well-deserved acknowledgment of an enthusiastic, knowledgeable and important community. The rumor was that they had asked to observe Mark205. I did not really believe this until several years later when the author of the proposal himself walked into my office and put the observations down on my desk. He was a well-informed and able high school teacher who had quite competently confirmed the bridge between the low-redshift NGC4319 and the high-redshift Mark205. I urged him to publish, but to this day I have not seen it in print and I do not know what difficulties he may have encountered.
As a side note: Someone observed the galaxy NGC1073 with the three quasars in its arms with the William Herschel Telescope in La Palma. I thought I saw some filaments associated with the quasars, but I have seen nothing published yet. Finally, one amateur was assigned time to observe spectroscopically the quasar which is attached by a luminous filament to a galaxy called 1327-206. But NASA set the Space

Intrinsic Redshifts

23

Telescope on the wrong object! Shortly thereafter, the Space Telescope Science Institute announced it was suspending the amateur program because it was “too great a strain on its expert personnel.”
An even greater embarrassment was, however, that all these objects were drawn from my book Quasars, Redshifts and Controversies, the contents of which the NASA allocation committees had been avoiding at all costs. As we will have occasion to mention a number of times during this book, amateurs have a much better grasp of the realities of astronomy because they really look at pictures of galaxies and stars. Professionals start out with a theory and only see those details which can be interpreted in terms of that theory.
This is some of the background behind the sensitive point which the journalist raised with Martin Rees in the final discussion panel. The reason the point is so sensitive is that the influential people in the field know what the observations portend, but they are too deeply committed to go back. The result will surely be to inexorably push academic science toward a position akin to that of the medieval church. But if that is the evolutionarily necessary solution, then perhaps we should hasten the process of replacing the present system with a more effective mode of doing science.

X-Ray Observations of Galaxy-Quasar Pairs
In addition, I reported for the first time in IAU Symposium 168, the results of pointed ROSAT observations on four additional galaxy-quasar pairs that fell conspicuously close together on the sky. (See Table 1-2). The probability of these associations being accidental was already very small back in the 1970’s, and when the observations revealed X-ray extensions from the galaxies toward the quasars, it not only clinched the physical association of these objects of vastly different redshift, but it also confirmed the ejection origin of the quasar from the galaxy.

Table 1-2. Galaxy-quasar associations investigated in X-rays through 1995

Galaxy

Quasar

Redshift Separation Probability

Mark474/NGC5682 NGC4651 NGC3067 NGC5832 NGC4319

BSO1 3C275.1 3C232 3C309.1 Mark205(3)

z = 1.94 .557 .534 .904 .070

1’.6(1) 3.5 1.9 6.2 0.7

5 × 10–3(2) 3 × 10–3 3 × 10–4 7 × 10–4 2 × 10–5(4)

1 Separation of quasar measured from NGC5682 nucleus. 2 Probability from Burbidge et al. (1971). 3 On cosmological hypothesis Mark205 is 0.5 mag. Less luminous than the definition of a
quasar . 4 Probability that a Seyfert galaxy would fall within 0’.7 of an arbitrary point in the sky.

24

X-Ray Observations Confirm

Fig. 1-8. The optical jet, spiral galaxy NGC4651 showing an X-ray jet emerging from its nucleus directly to the quasar with redshift z = .557. See also Fig. 7-11 for larger area view around the galaxy.
One of these pairs, NGC4319-Mark205, has already been discussed here, but the others are mentioned below because of the understanding they add to the nature of the galaxy-quasar relation. The two most compelling cases are discussed first.
Fig. 1-9. An X-ray map of the area around NGC4651/3C275.1 showing lines of X-ray sources from the quasar. Source no. 4 is a catalogued quasar of z = 1.477.

Intrinsic Redshifts

25

NGC4651/3C275.1
The radio-bright quasar 3C275.1 is situated only 3.5 arcmin from the bright apparent magnitude spiral galaxy NGC4651. The probability that this would occur by chance is only about 3 in 1000. But what no one ever calculated was the compound probability that the galaxy it fell so close to would be the one spiral galaxy in the brightest 7000 that had the most conspicuous jet emerging from it. That reduced the accidental probability to less than 1 in a million. Now Allan Sandage, who had photographed this galaxy in 1956, nervously grasped the implication, but immediately pressed the argument on me that the galaxy jet was not pointing at the quasar, which proved that it had nothing to do with the quasar. Of course, it was only pointing 20 degrees away from the quasar, and subsequent deeper plates revealed that there was material filling in under the jet, down to within a direction only 6 degrees away in position angle from the quasar. (See Figure 7-11 in a later chapter). But by that time the configuration had been relegated to the category of disproved associations.
Actually, there is an amusing story about the statistical association of the whole group of radio bright 3C quasars with bright apparent magnitude galaxies that G.R. Burbidge, E.M. Burbidge, P.M. Solomon and P.A. Strittmatter (B2S2) established. They found a less than 5 in 1000 chance of accidental association for the whole sample. When I showed the X-ray extension from the nucleus of NGC4651 almost to the position of the quasar to Prof. J. Trümper, the Director of the X-ray section of the Max-Planck Institut für Extraterrestrische Physik (MPE), I mentioned that this was one of a class of galaxies known statistically to be associated with quasars. He was very skeptical until I remarked that the B2S2 result had been confirmed by Rudi Kippenhahn (former Director of the Max-Planck Institut für Astrophysik). After that he wanted analyses only in the latter form! But as I brought more and more results to him he said, “Well I know you can’t be right, but I will help you where I can.” I had to ruefully admit that was not completely discouraging—in fact, it was about as as much encouragement as I ever got.
Figure 1-8 shows that X-ray material stretches from the nucleus of the galaxy toward the position of the quasar, where the quasar material extends almost to meet it. If the 10.5 kilosec exposure had been just a little bit longer, it might have shown the bridge to be continuous. But does that really matter, considering the low probability of accidental contiguity, the low probability of such an active jet galaxy being accidentally involved, and the vanishingly small probability that an X-ray jet would accidentally be coming out of the nucleus of the galaxy and pointing directly at the quasar? It would seem to me that a healthy science would eagerly recall all the other cases which pointed to the same conclusion and get on with the job of finding out why.
Figure 1-9 shows another characteristic tendency of these active objects, viz., to exhibit lines of sources emanating from them. Also shown in this picture is a tendency of the lines to be nearly at right angles to each other—something we will see many times. The former is easy to picture in a model where the sources are ejected from active galaxies and quasars. A cause for the latter is difficult to imagine, but when we

26

X-Ray Observations Confirm

Fig. 1-10. The NGC5689 group, a typical association of active objects around a large, low redshift galaxy. X-ray contours show Markarian 474 to be a very active Seyfert with an X-ray filament leading out to a quasar with a redshift of z = 1.94. The companion galaxy, NGC5682, is just to the upper right of Mark474.

get a mechanism that gives such ejections, it may be a sign we are approaching understanding. (Robert Fosbury, the ESO expert on Seyfert galaxies, tells me the optical ejection cones from these active galaxies have de-projected opening angles of about 80°.)
Mark474 and the NGC5689 Group
This is a prototype of the groups which I believe represent the building block units which make up our known universe. Like the groups of galaxies we know the most about, such as our Local Group and the next nearest large group, the M81 group; the NGC5689 group has as a spiral galaxy like that of type Sb as its dominant galaxy. Actually NGC5689 is classified as an Sa; but it is the same morphological type of massive rotating galaxy with a large central bulge of old stars. (Figure 1-10)
My attention was first called to it by Edward Khachikyan, an Armenian astronomer friend. B.E. Markarian, another Armenian astronomer, had found this very high surface brightness, ultraviolet rich galaxy now called Mark474. Next to it was the lower surface brightness galaxy NGC5682, which turned out to be a companion to the large NGC5689 and, characteristically, had a redshift about 100 km/sec greater. Mark474 had a redshift about 10,000 km/sec greater. I felt that the companion should have an associated quasar, and looked on the Palomar Schmidt prints for a blue object in the neighborhood. I found it, but it was a little too faint for the poor spectrograph on the 200-inch telescope. I asked Joe Wampler at Lick Observatory to get the spectrum, and it turned out to be a quasar of redshift z = 1.94. (Maarten Schmidt criticized me for going outside the Hale Observatories to get this spectrum on a smaller telescope, but I replied that Joe was the only one who had built a good enough spectrograph—the Wamplertron—to observe the object.)

Intrinsic Redshifts

27

Fig. 1-11. A closer view of the Seyfert galaxy Mark474 showing the X-ray material connecting to the quasar at the upper right (small dot inside smallest contour). Note material extending from the quasar in a direction away from the Seyfert.
Now I had a close triplet of unusual objects which were almost certainly associated, in spite of their vastly different redshifts. While I was sitting in an MPE working group, my ears perked up when I heard that this Markarian object had been discovered on the survey to be a copious source of X-rays. Arguing that such a strong source deserved to be observed in a pointed observation, I was able to obtain a 12,862 sec exposure in the low resolution mode. The initial reduction showed everything I had hoped for. The quasar was well visible in X-rays, and was connected back to, and elongated away from, the strong X-ray Seyfert. (Figure 1-10). (Actually it is unusual to see such a faint apparent magnitude, high redshift quasar detected in X-Arays.)
Figure 1-10 also shows that X-ray emitting material is being ejected along the minor axis of the “parent” galaxy in the system, NGC5689. The interesting implication here is that even though the presently active galaxies in the group probably evolve rapidly into more quiescent entities, the original galaxy in the group is capable of subsequent ejection episodes. It is also apparent that there are other relatively strong Xray sources aligned in an “X” pattern across Mark474. Most of them are identified with blue stellar objects (BSO’s), and clearly represent additional quasars associated with this active group.
The optimally smoothed X-ray contours are shown in enlargement in Figure 1-11. Skeptics immediately argue that if one puts two unrelated distributions of photons close to each other, they will meld together to form an apparent connection. Yet if one thinks about it for a moment, one realizes that they intermingle only in the outer isophotes to form an hour glass-like shape. Figure 1-12 here shows an isophotal contouring of two adjacent instrumental point-spread functions. It is clear that only the outermost isophotes merge into an hourglass shape, and all the inner isophotes immediately return to

28

X-Ray Observations Confirm

Fig. 1-12. Instrumental spread of photons around two unrelated point sources. Only outer isophotes hourglass together with inner contour lines returning quickly to symmetry.

circularity. Real elongated inner isophotes, filamentary connections and jets look conspicuously different, regardless of added noise.*
In Figure 1-11 one can see that the connection between Mark474 and the quasar passes close to the companion galaxy NGC5682, the galaxy which I had originally felt was the origin of the quasar. Now, however, a collaborator in the office next to mine, H.G. Bi, applied a deconvolution program to the data in order to sharpen the resolu-
a

Fig. 1-13. X-ray source (b)

discovered by H.G. Bi by

deconvolving strong image of

Mark474. The important

result is the almost exact

alignment of this blue,

peculiar, X-ray object across

the nucleus of the Seyfert

with the quasar (a). What

would a spectrum reveal

about (b)?

b

* Because of unfamiliarity with low surface brightness detection techniques coupled with non expectation of extended features, almost no use of such information has been made by X-ray observers. The X-ray archives are presently a gold mine of untapped information waiting for someone with access and computing power to harvest the data.

Intrinsic Redshifts

29

Fig. 1-14. Non-equilibrium configuration near Mark 474 of galaxies plus one X-ray BSO. What is redshift of the quasar candidate? What is nature of the extremely low surface brightness dwarf just to the north?

tion, and discovered a rather strong X-ray source buried in the outskirts of the Markarian galaxy. That X-ray source, was readily identified with a compact but deformed blue object. The decisive aspect then emerged—this new quasar-like object was almost exactly aligned with the known quasar across Mark474! (Figure 1-13). In view of the close pairings of quasars across Seyfert galaxies which have now emerged, this appears to be just another confirmatory pair across a Seyfert galaxy!
The NGC5689 group is typical also in the pattern of redshifts of the objects in the group. The largest galaxy has the lowest redshift; the smaller companion has a higher percentage of younger stars and a redshift hundreds of km/sec higher. There is a very active galaxy with thousands of km/sec higher redshift, and finally very high redshift quasars emerging from the latter. This theme is repeated over and over again. In later chapters, when we consider the redshifts decreasing as the objects age, we will try to suggest some possible reasons for this hierarchical, fractal pattern.
It would be helpful, however, to know the redshift of the compact blue object which is on the other side of Mark474 from the quasar, as shown in Figure 1-13. There is also an intriguing region situated midway between the dominant galaxy and Mark474. It consists of a string of red galaxies (a string being a non-equilibrium configuration which cannot last the age of the galaxies) containing an X-ray BSO. A peculiar dwarf galaxy is less than 1 arcmin away. The picture of this latter group is shown in Figure 114 and also in the publication of the proceedings of IAU Symposium 168 (ed. M. Kafatos and Y. Kondo). How long will it be before some of the numerous large telescopes around the world are used to observe these curious and intrinsically informative objects?

30

X-Ray Observations Confirm

Fig. 1-15. Kitt Peak 4 meter photograph of NGC3067 showing high surface brightness and shattered appearance of absorption filaments.

NGC3067 and the Quasar 3C232
This galaxy-quasar pair has had an absolutely amazing history. Back in 1971, Burbidge et al. derived a probability of accidental association of less than one in three thousand. A. Boksenberg and W.L.W. Sargent found absorption lines of the galaxy in the spectrum of the quasar in 1978 and assumed it was a distant, background quasar shining through the galaxy, a chance coincidence. In 1982, Vera Rubin et al. went further and attributed the spectral shift of the galaxy absorption lines to rotation around a massive galaxy taking place at an unusually large distance from its nucleus. Naturally, the latter calculation produced a mass of “dark” (undetectable) matter some 16 times the estimated mass of visible matter. Despite the enormity of this factor, it was hailed as proof of the existence of enormous amounts of unseen matter in the universe. But the galaxy was patently not an ordinary galaxy. It was a sharply bounded, very high surface brightness “star burst” galaxy—a rare and active kind of galaxy, which would make the
Fig. 1-16. Palomar 200-inch photograph of NGC3067 in light of hydrogen alpha emission showing ejected, hot gaseous filaments

Intrinsic Redshifts

31

Fig. 1-17. Radio map of neutral hydrogen in NGC3067 showing filament leading from disturbed galaxy to quasar. Map from Carilli, van Gorkom and Stocke.

accidental association with a quasar hundreds of times less likely. Moreover, pictures of the galaxy revealed a shattered, explosive morphology with emission line filaments issuing from it (Figures 1-15 and 16). Under no circumstances could it be a normal galaxy in equilibrium rotation, which would be required in order to derive a meaningful mass. The huge derived mass was a complete fiction! Why didn’t they look at the galaxy? (Actually I sent pictures, but to no avail).
An even more startling development occurred in 1989 when C.L. Carilli et al. found a filament of neutral hydrogen leading from the west end of the galaxy directly to and beyond the quasar! (Figure 1-17). The hydrogen had clearly come from the active galaxy—how else other than being pulled out by the ejection of the quasar? And notice that the quasar falls just at the densest point of the hydrogen distribution with contours of less dense gas trailing back towards the galaxy.
This extraordinary result should have cemented the association beyond any doubt, but later it was claimed that the configuration was merely accidental. J. Stocke et al. argued that the neutral hydrogen at the redshift of the galaxy absorbed continuum light from the quasar, but did not show excited optical emission lines, proving the quasar was quite far in back of the hydrogen filament. Because the other arguments are so overwhelming that we are dealing with another physically associated galaxy and quasar, I reread very carefully the complex calculations they had made. There it was: a “short” extrapolation. The photons they needed to ionize the hydrogen in the filament and make it fluoresce were at shorter wavelength than those in the spectrum. So they extrapolated to an unobserved portion of the spectrum. I extrapolated and got half their value. But regardless of how much quasar radiation was extrapolated to be shorter than

32

X-Ray Observations Confirm

Fig. 1-18. Integrated, low surface brightness X-ray emission around the galaxy/quasar pair NGC3067/3C232. This represents another “cross” extension of X-ray material from active objects.

this wavelength, the actual amount would be determined by the amount of hydrogen at redshifts intermediate between the quasar and the hydrogen filament, the degree to which the filament was composed of small, dense clouds, and the relative beaming angle between the ultraviolet and radio wavelengths from the quasar. If the conventional paradigm had instead required the quasar and filament to be adjacent, which of these plausible configurations would have been announced as a new “discovery”?
The X-ray fun had only just begun. When the Einstein Laboratory Satellite went up, it observed the quasar because it was quite a bright object. At a workshop at the European Southern Observatory (ESO), I pointed out that there was an X-ray tail coming off the quasar in a direction opposite to the galaxy. Martin Elvis from the Cambridge Center for Astrophysics (CFA) jumped up and said, “That’s noise.” I argued that you could see that it was not noise. He said, “I’ll look into it and report what I find.” He never reported back.
When I got the relatively short 5600 sec exposure on it with ROSAT, there was the X-ray extension north of the quasar! In fact, there was another cross extension of X-rays (Figure 1-18)—quite similar to the configuration around Mark474. But the most exciting result was that there was a double-sided X-ray jet coming out of the nucleus of the starburst galaxy, NGC3067 (Figure 1-19). How many galaxies does one find with such conspicuous bipolar X-ray jets? When I showed this to my MPE colleagues they became angry at me for saying that I thought the jet was curving slightly around as it extended NE, even more toward the quasar, and that a longer exposure might show it leading directly to the quasar. Others said they thought the X-ray extensions from the quasar were just noise. Further X-ray observations on the object were rejected by the allocation committee.

Intrinsic Redshifts

33

Fig. 1-19. A bipolar X-ray jet from the nucleus of NGC3067E, one side of which extends in the general direction of 3C232. This is a short, 5000 second exposure with ROSAT. A longer, higher resolution exposure was turned down by the allocation committee.

NGC5832 and 3C309.1
The last of the five pointed observations that I got with ROSAT was a very short exposure of 4300 sec on one of the Burbidge et al. galaxy/radio quasar pairs. Only the quasar registered, and the galaxy, relatively far away at 6.2 arcmin distance, did not. The distribution of X-ray sources in the field, however, was very interesting. As (Figure 120) shows there is again a strong line of sources running NE to SW through the quasar
Fig. 1-20. A very short, 4300 second X-ray exposure shows only the quasar 3C309.1 but not the nearby galaxy NGC5832 (plus sign). Small X-ray sources in the field, however, form a line and possibly a cross through the quasar.

34

X-Ray Observations Confirm

and the suggestion of a line coming off in almost an perpendicular direction. This configuration was criticized because some of the sources had only 3 to 6 counts. I argued in return that if the background is low enough, just a few counts make for significant sources, as can be well judged visually.

Chapter 2

SEYFERT GALAXIES AS QUASAR FACTORIES

Evidence that quasars were physically associated with low redshift galaxies had been amassing since 1966 (See Quasars, Redshifts and Controversies for the history through to about 1987). The following years saw further proofs accumulate, mostly from X-ray observations, and they are reported now in the previous chapter. But the stronger the evidence, it seemed, the more attitudes hardened against accepting these observations.
With the discovery of the pair of quasars across NGC4258, however, a new level of proof suggested itself. If more such striking pairs across active galaxies could be found, it would be hard to resist the ultimate conclusion. What more obvious class to inspect than those like NGC4258, namely Seyfert galaxies?
Seyfert Galaxies
The American astronomer Karl Seyfert discovered this class of galaxies in the 1950’s by looking at photographs and noticing that some galaxies had brilliant, sharp nuclei. The emission line spectrum of such a galaxy signified that large amounts of energy were being released into its nucleus. For a long time, no one was worried where this energy came from. When the problem was finally realized, “accretion disks” came to the rescue—a kind of cosmic equivalent of throwing another log on the campfire. But the conspicuous emission lines did enable astronomers to do something they are good at—systematically classify and catalogue these objects.
Since Seyfert galaxies produced strong X-ray emission, by 1995 most of the brighter ones had been observed with the ROSAT satellite. This presented an opportunity to investigate a class of active galaxies which had been previously defined and more or less completely observed. The existing observations could be analyzed to see

Halton Arp, Seeing Red: Redshifts, Cosmology and Academic Science (Apeiron, Montreal, 1998)

35

36

Seyfert Galaxies

Fig. 2-1. Cumulative number of X-ray sources brighter than strength (S) around a nearly complete sample of bright Seyfert galaxies. Dashed line represents counts in non-Seyfert control fields.
whether there existed more cases of pairs of quasars across active nuclei such as had been observed in the Seyfert galaxy NGC4258.
A search through the archived X-ray observations revealed that, of all Seyferts known, observations were 74% complete to 10th apparent magnitude and 50% complete to 12th. After some contaminated fields had been weeded out, a total of 26 fields were available.
Now came the formidable task of accessing and analyzing these fields. As mentioned previously, an enormous amount of specialized knowledge is required to enter the “public” archives. I found the perfect candidate to collaborate with me on this job, a German astronomer named Hans-Dieter Radecke. He had just finished doing a very important and courageous piece of work on the gamma ray observations in the region of the Virgo Supercluster which we will discuss later. But he was out of a job—and the problem was to find him some funding. It seemed hopeless, but as a last resort I asked Simon White, our new director at Max-Planck Institut für Astrophysik (MPA) if he could help. To our delight he found support for 6 months, and this made possible what I hope will be recognized as a crucial step forward in our understanding of physics and cosmology.
Hans-Dieter produced lists of sources, their strengths and positions, for each of the 26 Seyfert fields. Then, using exactly the same detection algorithm, he reduced 14 control fields. The control fields were within the same range of galactic latitudes and treated identically to the Seyfert fields. Therefore, when a significant excess of X-ray sources was found around the Seyfert galaxies, there was no question that these X-ray

as Quasar Factories

37

Fig. 2-2. X-ray sources around Seyferts of two brightness classes showing how associated sources become less bright as Seyferts become more distant.
sources belonged to the active galaxies. The sources were 10 to 100 times more luminous than sources usually found in galaxies, such as binary stars or supernova remnants, and they lay far outside the confines of the galaxy (generally from 10 to 40 arcmin away or several hundred kiloparsecs at the average distance of the Seyfert.) Practical experience guaranteed that these kinds of X-ray sources would be confirmed as quasars. The beautiful feature of this result was that any astronomer could simply look at this one plot of X-ray strength versus number per square degree (as shown here in Figure 2-1), and realize that when these excess sources—which manifestly belonged to the Seyferts—were measured, they would almost all turn out to be quasars.
With one economical diagram we had proved that Seyfert galaxies as a class were physically associated with quasars! This added enormously to the significance of the pairs across Seyferts such as NGC4258, because now the data were just telling us in what way the pairs were related to the active galaxy. In later sections we will discuss the obvious relation of the quasar pairs to pairs of radio sources which, since the 1950’s, are acknowledged to have been shot out of the nuclei of active galaxies. Of course, in this sample of 24 Seyferts (omitting the brightest two as being too close to fit into the average sample—see Figure 2-2), many more pairs of X-ray sources were found. All told, there were 21 pairs of X-ray sources involving 53 BSO’s (some pairs or alignments involved multiple X-ray sources). Almost every Seyfert had a pair of BSO’s, most of which were certain to turn out to be quasars! Before we discuss some of these new pairs, however, it is interesting to comment on how these developments were received.
Spreading the Good News
Astronomers are always holding meetings, and as the journals become choked with papers, the meetings are increasingly the forum where new results are communicated (except for press releases, which are so hyped that they have to be heavily discounted). The meetings are traditionally the places where power relations are straightened out. It is painful for me to attend them because there is almost total conformity to

38

Seyfert Galaxies

obsolete assumptions. But I am old-fashioned enough to believe that when truly important new results come along, the conference organizers have a moral obligation to see that the results are presented.
With the new results in hand, I became optimistic that when they were communicated, they would finally persuade the researchers to at last begin to reappraise the fundamentals in the field. Many of the new results discussed in Chapter 1 were available at the time of the well known Texas Symposium on Relativistic Astrophysics, which was held in the adjacent institute in Munich in December of 1994. I submitted an abstract of a paper I wanted to give. The schedule was released, but my name was missing. About 14 years previously the Texas Symposium had been held in Munich, and I travelled all the way from California to give a summary paper on the evidence for associations of quasars with low redshift galaxies. The paper had made some impression then. But now, I was sad to say, after all this time, when the evidence had grown so much stronger, the newest evidence was not to be allowed. Sometime later, an international X-ray conference was held in the nearby town of Würzburg. Again I was excluded.
Finally, in 1996 I was awarded an Eminent Scientist invitation to come to Japan for three months. It was suggested that I could time my visit with an international conference on X-rays that was going to take place in Tokyo. The new results on the families of quasars around Seyferts were just out, so I sent in an abstract and arranged to come during that period. I was really joyous at the thought that this important information could be communicated in these circumstances, and that some sort of reconciliation could take place between people who were really interested in advancing knowledge. Just as I was packing, the conference schedule came out without my name on it.
Now, I am experienced enough to know how organizing committees pick speakers for conferences. And I have a rough idea of who, particularly in the most advanced countries, exerts pressure to keep what they consider rival research off the programs. But I am extremely saddened to realize that the members of the local organizing committees give in to such imperialistic pressure.

A Striking New Pair
It was exhilarating just to scan through all the new, good looking pairs of X-ray sources across the Seyferts in the maps obtained from the archives. A sample of these maps is shown here in Figures 2-3,4 and 5. Note that the X-rays are plotted just as received, and not averaged to give their mean position, which is on average accurate to a few seconds of arc. Even though the images enlarge as they occur further off axis, their disposition with respect to the central Seyfert and the relative X-ray brightness of the sources are conveyed very clearly.

as Quasar Factories

39

Fig. 2-3. A sample of the pairs of X-ray sources discovered across bright Seyfert galaxies. X-ray photons are plotted as received so that spreading of images with increasing distance from field center is conspicuous. Numbers represent counts per kilosecond. Lines in NGC1068 field represent direction of the distribution of water maser sources in the inner nucleus

40

Seyfert Galaxies

Fig. 2-4. A new pair of quasars across the Seyfert NGC2639—the most similar in redshift so far found! From measures by E.M. Burbidge.

An outstanding pair was immediately noticed across NGC2639. The two X-ray sources were both very strong (26 and 38 counts per kilosec). The identifications with BSO’s were accurate and unambiguous (actually, one was a catalogued quasar that I had discovered near a companion to NGC2639 in 1980). Again, there was the need to get a spectrum of the other member of the pair—again Margaret Burbidge to the rescue. That pair of redshifts turned out to be a lift-you-out-of-your-chair thrill! As Figure 2-4 shows, the redshifts were z = .307 and .325, a difference of only .018. This was the closest that any of the pairs had been in redshift. What was exciting about this, of course, was that two unrelated X-ray quasars had only about one chance in 100 of accidentally falling this close in redshift. That probability, times the vanishing probability of finding such a strong pair of X-ray sources across an arbitrary point in the sky, made the whole computation a waste of time—here was clearly another physical pair across a Seyfert.
Fainter sources can be seen in Figure 2-4 aligned opposite to the z = .307 quasar and extending toward the z = .325 quasar. With fainter isophoting on an enlarged view of this region, four BSO’s are optically identified, and clearly will represent a trail of quasars leading out to the z = .325 quasar when confirmed (see later Figure 3-26).
Another Water Maser
While the quasar redshifts were being measured, word arrived that stimulated emission from H2O molecules had been observed in the nucleus of NGC2639—the same water masering that had been observed in the nucleus of NGC4258. This meant that the two best known pairs of quasars across a Seyfert fell across the Seyferts known to have the strongest “black hole” activity. The water maser lines in NGC2639 were particularly variable, showing velocity drifts of about 7 km/sec in a year.

as Quasar Factories

41

Fig. 2-5. Very strong (268 and 119 cts/ks) X-ray sources across the Seyfert NGC4235. Catalogued identifications as a quasar and a BL Lac object are labeled. Plus sign indicates the position of a Seyfert 1 of z = .080 identified previously but not registered in the present ROSAT X-ray map.

I had pointed out that the patches of water maser emission in the nucleus of NGC4258 were distributed approximately along lines in the direction of the two quasars (Figure 1-4 in preceding Chapter). So I was very pleased when Margaret and Geoff Burbidge wrote a short paper gathering together all the evidence for NGC4258 ejecting material in roughly these directions (in contradiction to the conventional interpretation, which had the rotation axis of the black hole at 90 degrees to this direction). Following this, another water maser was discovered in the center of NGC1068. As the lower right hand panel in Figure 2-3 shows, the orientation of the water masering spots (full line) again points in the direction of a strong pair of X-ray BSO’s aligned across the nucleus of NGC1068. Now it may turn out that masering activity is common in Seyferts, as is ejection activity, but it also appears to be correlated with the strength or direction of the major ejection activity in the galaxy.
At the moment, the best guess as to what excites the water molecules is radiant energy in the beam associated with the ejection of the quasars. Why such a “cool” molecule is present in the very inner regions of such active galaxies may be a more challenging question.
A New Pair of Enormously Strong X-Ray Sources—NGC4235
When I first saw the X-ray map of the field around NGC4235, I was sure the pair of sources belonged to the Seyfert, because the chance of accidentally encountering such strong sources is only about 4 in 1000. Taking into account the alignment and equal spacing gives a total chance of the pair being accidental of only 6 in 100,000!
But I made a hasty assumption—namely that they were so strong in X-rays that they would be X-ray galaxies. So I only checked catalogues of known X-ray galaxies, and when I did not find them, I assumed they would have to be measured. After the paper was submitted I stumbled across the two sources catalogued; one as a very bright quasar of z = .334 and the other as a characteristically X-ray bright BL Lac object of

42

Seyfert Galaxies

z = .136. (See Figure 2-5.) The discovery of BL Lac objects in associated pairs is extremely important. We will show in later sections that BL Lac’s, because of their rarity, offer a powerful proof of associations, and therefore of intrinsic redshifts. They will also play an important role in the discussion of galaxy clusters in Chapter 6.

The Question of Ejection Velocity
In the first chapter we stressed the fact that the observations of pairs of quasars allowed us to compute a projected ejection velocity of about .07c. The NGC4235 pair just discussed would support this by giving a projected ejection velocity of .08c. (That is the intrinsic redshift of the quasar would be z = .235 but the velocity towards us would subtract z = .099 and the velocity away from us would add z = .099.) In Chapter 1, however, we showed one case where the redshifts in the pair were z = .62 and .67, yielding a projected velocity of only .015c. The separation on the sky for this case was about 1.3 deg., about 50% greater than other pairs associated with galaxies at this approximate distance from us. This made it plausible to argue that we were viewing the inevitable occurrence where the ejection was across our line of sight and the toward and away components of velocity were much reduced.
It was amusing to note that when the NGC4258 pair was first being discussed in a colloquium, Günther Hasinger challenged me to predict the redshifts of the probable quasars. The conventionalists clearly wanted a way to wriggle out of having to accept the association of the quasars. When they were measured at z = .40 and .65 I was encouraged that they were that close, and the conventionalists were relieved that they were not closer. But they should not have been relieved, because if quasars had been much closer, there would have not been enough velocity to get them out of the galaxy nucleus.
The pair across NGC2639 at z = .307 and .325, however, represents a more interesting situation. That would only allow .007c velocity component in the line of sight. For an ejection at 0.1c, we might expect to get such an orientation across the line of sight only about 9% of the time, a figure that can only be checked by measurements of many more pairs.
But there is another very interesting aspect to this problem. Do the ejection velocities represent a constant velocity of escape which will allow the quasars to pass out into the space between the galaxies; or do they decelerate as they reach larger distances from the ejecting galaxy? Do they keep going or stop?

The Narlikar-Das Model for Ejection of Quasars
By the 1980’s I had produced strong statistical evidence that quasars were in excess density around younger companion and active galaxies. Jayant Narlikar and P.K. Das set out to make a dynamical model which could explain this. The problem was, assuming reasonable properties for the quasars, to find a way to keep the quasars in the spatial vicinity of the ejecting galaxy. Their model did this very nicely (Astrophysical Journal. 240,401).

as Quasar Factories

43

Fig. 2-6. All quasar candidates in a region around NGC1097 identified by X.T. He from a Schmidt, objective prism plate. The size of the PSPC and HRI fields investigated with X-rays are shown dashed.

One quantitative prediction of their model was that a quasar would reach a maximum apogee from the galaxy of about 400 kpc. Now it is very interesting that at the redshift distance of NGC2639, the two quasars are just about 400 kpc from the Seyfert. This would mean that the ejection velocity would have been lost and the quasars would be moving very slowly. Therefore quasars at larger distances from their galaxies of origin might be expected to have more closely matching redshifts regardless of the orientation of their ejection direction to the line of sight. Another aspect, which will be discussed in later chapters, is that quasars probably evolve to lower intrinsic redshifts as they age. In that case quasars of lower redshift would generally be expected to have smaller components of ejection or “peculiar” velocity—more like the galaxies into which they are evolving.
The Seyfert Galaxy NGC1097
NGC1097 has the most extensive, low surface brightness optical jets of any galaxy known. Plate 2-7 shows true color compositing by Jean Lorre of a set of the deepest 4 meter telescope plates ever taken at Cerro Tololo, Chile. On one side, just between the brightest optical jets, is a concentration of 5 or 6 bright quasars. These have been shown to represent an excess of a factor of 20 over expected background values, and about 40 quasars have been demonstrated to be concentrated around the galaxy (Quasars, Redshifts and Controversies pp 48-53 and 64). Figure 2-6 shows all the quasar candidates in the inner 2.85 × 2.85 degree center of an objective prism plate taken by the U.K. Schmidt telescope in Australia. The Chinese astronomer X.T. He picked these out by the appearance of their spectra; and in a two year observing program in Chile, I was able to verify with individual spectra that his accuracy of quasar identification was an amazing 94%. It is important to note that when this considerable work by a number of people was published in 1983 and 1984, it already established, at that time, the association of

44

Seyfert Galaxies

Fig. 2-7. Enhanced, star removed, composite (by Jean Lorre) photographs of NGC1097 showing luminous, crossed jets. Below center is the PSPC X-ray map of the field with known X-ray quasars numbered 24 through 28. Note faint, unidentified X-ray sources on the other side from the Xray bright quasars. Bottom right shows an enlarged map in radio wavelengths with the two strongest jet directions marked.

quasars around this one Seyfert galaxy that we are now finding to be characteristic of Seyferts as a class.
In 1993 and 94, however, I received X-ray results of my own on this most exciting galaxy-quasar association. The data came from all three ROSAT modes, the low and high resolution pointed and survey modes. (The size of the fields covered by PSPC and HRI is shown in Figure 2-6.) When I first reduced the X-ray data, I was at once struck by the large number of X-ray sources in the field. Brighter sources in the NGC1097 field were more than 50% in excess of average control fields. The X-ray sources detected by ROSAT confirmed the earlier observations by the Einstein X-ray observatory and, in particular, confirmed that the brightest quasars fell just between and along the strongest optical jets. Since it is difficult not to believe that the optical jets are ejected, it is obvious that the quasars are also ejected from NGC1097.
These observations also showed lines and pairs of fainter X-ray sources coming out of the nuclear region of the Seyfert (Figures 2-7 and 2-8 and Plate 2-8). There was a large excess of X-ray sources around the disk region of the galaxy, and evidence for strong absorption of the soft X-ray component of many of the faint sources. Since it is known from optical studies of the galaxy that there is strong absorption in the disk of

as Quasar Factories

45

Fig. 2-8. The high resolution X-ray map (HRI) of NGC1097. Note material filling in toward the bright quasars 26 and 27. Note also the new point X-ray sources (6) and (a) aligned across the nucleus in the direction of the brightest jet.

the galaxy (Figure 2-9), it was reasonable to suppose that the metals in this mixture of dust and gas were also dimming the soft band of the X-ray sources. Unabsorbed, the Xray sources would be bright enough so that they could reasonably only be quasars. The upshot is that these observations suggest many higher redshift quasars are being ejected, and that many may be encased in thick cocoons. Evidently, this is a busy quasar factory, and an interesting place to investigate in the far red and infrared.
We will see later that high redshift quasars (z = 2) are generally fainter than qua-
Fig. 2-9. Photograph of the barred spiral NGC1097 showing interior regions with opaque filaments and clouds of dust. A technique of emphasizing contrast in high surface brightness regions while preserving faint surface brightness features has been applied. (This technique was originally called automatic dodging but is now called unsharp masking.)

46

Seyfert Galaxies

Fig. 2-10. An X-ray map from the all sky survey by ROSAT shows a very bright X-ray source about 1.9 degrees SW, roughly along line of the counterjet (dashed) to the brightest jet (full line). Spectrum by Tony Fairall identifies the source A as a BL Lac object.

sars in the z = .3 to 1.5 range. We have, however, already mentioned that the quasars appear to be born with high redshifts which decrease with age. Since many bright quasars with these smaller redshifts are associated with NGC1097; it is reasonable to suggest the fainter X-ray sources are high-redshift, younger quasars, many of which are just emerging from the central, dusty regions of the Seyfert.
For example, the high resolution X-ray observations (ROSAT HRI) show a pair of point sources paired across the NGC1097 nucleus (designated as 6 and a in Figure 28). These are not optically identified on deep, blue sensitive plates. They would probably be identifiable with the penetrating power of the infrared techniques on the new large aperture telescopes. What a useful project for these expensive facilities!
Figure 2-8 also shows low surface brightness material extending out from the nucleus of NGC1097, between the jets, to the location of two of the brightest quasars. It is not clear that this is X-ray material, because it does not show in the more sensitive PSPC observations. It is more probably ultraviolet light leaking in through an imperfectly blocked filter. (This possibility was doubted when I first published the evidence, but later a leak was verified in a measurement of the filter.) In any case the important aspect of this material is that it must arise from some form of hot gas which has been ejected along with the quasars! (One attempt to get spectra with the satellite ultra violet explorer failed on an administrative error, and the other attempt failed when a stabilizing gyro died.)
In the wide field of the ROSAT survey mode shown in Figure 2-10, there is a very strong X-ray source (A) identified about 1.9 deg to the SW. In fact, it is stronger than the very strong NGC1097 itself. It is identified with a bright (16.5 apparent mag.), probably stellar appearing object. Tony Fairall with the 74-inch South African telescope took a spectrum which demonstrated that it had a blue, continuous energy distribution, thus identifying it as a BL Lac object. This important kind of quasar-like object will be

as Quasar Factories

47

Fig. 2-11. An enlarged portion of the previous Xray survey map shows sources along the general line to the BL Lac object. Perhaps of even more importance, the brightest quasar no. 28 lies along the line of the brightest jet and its X-ray contours are extended in both directions along the line of this jet!

discussed immediately below. But first I would like to point to an important discovery in this survey X-ray field.
As the enlarged Figure 2-11 shows, the BL Lac object is on a line of sources SW from NGC1097, which coincides very closely with the counterjet of the strongest optical jet to the NE. Along that major optical jet to the NE, however, is one of the brightest X-ray quasars belonging to NGC1097. The X-ray isophotes of this quasar extend both backward and forward along the line of the strongest optical jet. Since this alignment is obviously not an accident, and since the optical jet obviously originated by ejection from the active nucleus, this is another proof that the quasar must also have originated by ejection from the nucleus! Moreover, the strong X-ray BL LAC on the other side of the nucleus must then represent the other component of the ejected pair.
The Empire Strikes Back
Since the NGC1097 paper contained tables full of new source identifications from the analysis of the three different field sizes centered on the important Seyfert galaxy NGC1097, I thought it would be routine to publish in the journal which was carrying most of the European X-ray results of archival value. How wrong I was! The referee’s report came back accusing me of “manipulating the data” and trying to claim an association of quasars with galaxies, which had “long ago been disproved.” The editor forwarded these comments and rejected the paper on the grounds that he saw no need to reopen the debate. The extraordinary aspect was that four papers in addition to my own had just appeared in the same journal giving strong additional evidence for just such associations! The figures appear here for the first time, and the tabular X-ray data is still unpublished.

48

Seyfert Galaxies

Fig. 2-12. X-ray map of the large southern spiral NGC1365. The strong source SSW of the Seyfert is a BL Lac object of redshift z = .308.

BL Lac Objects
These objects are so named because the prototype was originally classified as a variable star within our own galaxy. But then it was discovered that, in many cases, faint, redshifted lines could be detected on the strong continuum spectrum. Often these objects also showed faint nebulous edges to their images. The BL Lac’s are now known to be strong radio and X-ray emitters, and are strongly variable.
They are also rather rare, and when they showed up in a ROSAT Seyfert field, they were very conspicuous because of their strong X-ray emission. Figure 2-12 shows an example of a BL Lac object close to the grand design spiral Seyfert NGC1365. While I was inspecting the 26 Seyfert archival fields discussed earlier, it was clear to me that the number of such objects encountered was significantly higher than would be expected in non-Seyfert fields. There is no real need to compute probabilities—but it can be done simply enough! The probability of encountering X-ray BL Lac’s this bright, this close to the Seyfert ran from about 10–2 to 10–4. (See Table 2-1). Therefore, the chance

Table 2-1 Current summary of BL Lac Objects in Seyfert fields

Seyfert

X-ray BL Lac

R

IPC

FX

P(BL)

V

z

Cen A

(570)ctsks–1

114’

168 × 10–13 1.5 × 10–3 17.0 mag. .108

NGC 1365

89

12

6.7

4.2 × 10–3

18.0

.308

NGC 4151

257

4.5

14.8

4.1 × 10–4

20.3

.615

NGC 5548

1213

35

(88.5)

4.1 × 10–3

16

.237

NGC 4235

268

36

(19.6)

2.2 × 10–2

16

.136

NGC 3516

156(Sl)

22

13.6

—

16.4

.089

as Quasar Factories

49

Fig. 2-13. a) (top left) Optical photograph of the barred spiral NGC1365. b) (bottom left) Neutral hydrogen map by Jörsäter and van Moorsel with faint optical arms overlaid. (scale in units of arc seconds from nucleus) c) (top right) Faintest hydrogen contours with distance and direction to the BL Lac object indicated.
of encountering the first five was about 3 × 10–5, and if we count the object near NGC3516 as a BL Lac, a chance of only about 3 in ten million! A referee argued that due to the uncertain density of bright BL Lac’s, this probability was uncertain—thus disproving the association! But even if it were only one in a million, the result is overpoweringly significant. Moreover the finding is restricted to just the five clear-cut BL Lac’s encountered so far, and more are indicated in the full sample of 26.
“Ridiculous!” snorted the conventional astronomer. Who would believe a probability that small? Right! What is wrong? Well it’s a posteriori, computed after you found the effect. So let’s throw it out! Ah, but along came a great stroke of good fortune. In

50

Seyfert Galaxies

1979, Jack Sulentic and I had tested the proximity of then known BL Lac’s to bright apparent-magnitude, Shapley-Ames Catalogue galaxies and found an excess at a separation of about 1 deg. (the same as in the Seyfert fields). So the Seyfert result was not a posteriori, but a confirmation of a previously predicted result. The cautionary lesson here seems to be that no matter how significant the result, it is customary to try to invent a reason to discard it if it doesn’t fit expectations. The game here is to lump all the previous observations into one “hypothesis” and then claim there is no second, confirming observation.
But most important of all: Does the result make sense? It does, and in fact it is expected on empirical grounds. Consider one of the quasars in the pair associated with NGC2639, which we just discussed. Its apparent magnitude was V = 18.1 mag. and its redshift z = .307. Compare that to the BL Lac object within 12 arcmin of the Seyfert NGC1365. That BL Lac object had an apparent magnitude of V = 18.1 mag. and z = .308. The X-ray flux from the quasar was strong, but the flux from the BL Lac object was 3.5 times greater—undoubtedly the signal of the strong non-thermal continuum which reduces the spectral lines characteristic of BL Lac objects to low contrast.
But BL Lac objects are notoriously variable. The implication, then, is that a BL Lac can turn into a quasar quite easily, and vice versa, since they are already very similar. The key point here is that BL Lac’s are a rare kind of quasar which can be easily recognized because of their strong X-ray emission. Therefore, they are easily proved to be associated with active galaxies—confirming the proofs that the related kinds of objects, the quasars, are also associated.
It is interesting to inspect the neutral hydrogen maps of the grand-design, barred spiral NGC1365. Figure 2-13a shows the optical photograph. Figure 2-13b shows how the hydrogen concentrates in the spiral arms to the southwest of the galaxy. (One can see the multiple, ejected arms to the north of the galaxy which, at a glance, disposes of several decades of density wave theory for the formation of the arms.) But Figure 2-13c shows how this hydrogen is extended closely in the direction of the nearby BL Lac. In the following case of NGC4151 we will actually see a connection to a BL Lac.

The Seyfert Galaxy NGC4151
Another famous and extremely active Seyfert galaxy is NGC4151. A map of it and its surrounding companions is shown in Figure 3-18 in the next chapter. Presented here is the X-ray map in Figure 2-14, which shows that there is a line of X-ray sources stretching through the central active galaxy to the NNW and SSE. The two to the NNW are rather strong, at 16.0 and 16.2 counts per kilosecond, but they are relatively defocused, at 33.1 and 33.9 arc minutes from the center of the field. They therefore appear rather spread out. Now they, and the sources opposite them with 14.3 + 9.1 and 35.1 counts/ksec, are all identified with blue stellar objects (BSO’s). Therefore, we have a case of two highly probable pairs of quasars aligned across this Seyfert, both are aligned fairly closely in the same direction. (One might also consider this one ejection with a narrow-opening cone angle.)

as Quasar Factories

51

Fig. 2-14. An X-ray map of a 1.1 × 1.1 degree region centered on the large Seyfert NGC4151 (see Fig. 3-18). A strong BL Lac (257 counts/kilosec) is situated 4.5 arcmin N of the galaxy. Outer X-ray sources are also distributed generally along this line.

Along this line, about 4.5 arc min NNE from NGC4151, is a very powerful X-ray source, measured at 257 counts/ksec (compared to the Seyfert itself at 570 counts/ksec). This is a BL Lac object. Like the one falling next to the Seyfert NGC1365, it is very unlikely to have been encountered by chance. In this case, the probability is only 2 × 10–5 (see Table 2-1). But the object is also very unusual, in that it was first discovered in a radio mapping of the environs of nearby bright galaxies by the Westerbork telescope. Jan Oort had urged this project on me in collaboration with two Leiden radio astronomers, Tony Willis and Hans de Ruiter. We had identified this rather strong radio source in the NGC4151 field with a very faint stellar object. It was so faint that I had to use the multichannel spectrometer on the 200-inch at Palomar for several long exposures in order to try and determine its redshift. I only could register one line, and guessed that it was Lyman alpha at a redshift near z = 2. That turned out to be wrong, as John Stocke and collaborators later measured the object to have a redshift of z = .615.
The puzzle is this: What kind of object would be so faint optically and have such strong radio and enormous X-ray emission? As mentioned, it was highly probable that it belonged to NGC4151, and from Figure 2-14, it could be seen to lie in the apparent channel of ejection from that active Seyfert. But would there be any interaction in Xrays due to the spatial proximity of this strong BL Lac and the Seyfert? By searching the archives, Radecke and I found some HRI exposures of this field, and I set about looking at outer contours of the two images. The outer, lower surface-brightness regions of NGC4151 revealed a filamentary extension which connected directly to the BL Lac object, as shown in Figure 2-15.
Identifying luminous connections between objects of greatly disparate redshifts is a decisive way to establish their non-velocity character, as we have seen in the previous connections to quasars from Mark205, Mark474 and NGC4651. There could be a rich

52

Seyfert Galaxies

Fig. 2-15. High resolution Xray map (HRI) showing low surface brightness connection between NGC4151 at z = .003 and the BL Lac at z = .615.

harvest of additional information if X-ray astronomers were to recognize the increased detection to be gained from integrating their data over extended surfaces. This is related to the old art of surface brightness photometry, but would require hiring people who were either experienced or motivated.
Finally, we call attention in Figure 2-16 to the innermost radio structure of NGC4151. At the high frequency of 5 GigaHerz, the resolution is so good that objects of less than ¼ arc sec can be seen emerging on a line on either side of the central nucleus, C4. X-rays cannot yield such high resolution, but show extension in the same direction. Some compact, high-energy objects are being ejected in opposite directions from this compact nucleus—what else could they be but proto quasars? This ejection direction obviously rotates with time, so only older ejection tracks would be pointing to outer, associated quasars. We will later grapple with the question of what state the matter is in when it first starts its journey, but the important inference for now is that small entities are ejected from the nuclei of active galaxies and evolve into high redshift quasars and allied objects.
Fig. 2-16. A high resolution (5 Gigaherz) radio map of the nucleus of NGC4151 by A. Pedlar et al. The condensation C4 is considered to be the central source.

as Quasar Factories

53

Fig. 2-17. The Seyfert galaxy ESO 416-G002 is about z = .03 redshift. The two aligned X-ray sources are identified with blue stellar objects whose spectra have only recently been observed. (Observations by Wolfgang Pietsch.)

ESO416-G002
During the time rumours were flying about Wolfgang Pietsch’s pair of X-ray sources across NGC4258, the inevitable reaction was “Well that’s just a curiosity; there won’t be any more.” But in addition to all the other cases described earlier in this chapter, in his own programs he had observed several other Seyfert galaxies which turned out to be just as devastating.
One is shown here in Figure 2-17. There are only three strong X-ray sources in the field. The source in the center is a Seyfert of z = 10,000 km/sec, and the two others, almost exactly paired across it, are accurately centered on stellar-appearing objects. Somehow, after more than two years of constant effort, it was never possible to obtain their spectra. Perhaps that speaks more eloquently than any further comment that could be made. (Recently Pietsch, with collaborators, confirmed the weaker of the pair as a quasar of about z = .6, and the stronger as a BL Lac object.)
Other Examples
Now that Seyfert galaxies have been identified as quasar factories, it is easy to look back and recognize all the other Seyferts which, in the past, were found to be the origin of associated quasars. Of the first two quasars to be associated with companion galaxies (see Quasars, Redshifts and Controversies pp 22-23), the quasar in the NGC5689 group turned out to be associated with the Seyfert Mark474 (see Chapter 1 here), and the quasar in the NGC7171 group turned out to be associated with a Seyfert. Mark205 is technically a Seyfert, although it is often called a quasar, and PG1211+143 is arbitrarily called a quasar, but very similar to Mark205 (redshifts z = .070 and .085 respectively).
There is a quasar GC0248+430 which—if you are ready for this—is described in the literature as “a possibly microlensed quasar behind a tidal arm of a merging galaxy.”

54

Seyfert Galaxies

The galaxy turns out to be a Seyfert 3, and the quasar has a redshift of z = 1.31. Many more of these galaxies associated with quasars may well turn out to be Seyferts when people get around to measuring them. That is not to say that only Seyferts eject quasars. Some good examples of “starburst” galaxies that give rise to quasars are NGC520, M82 and NGC3067. But then starburst galaxies are closely allied to Seyferts, and the classes may evolve rather rapidly. There is also the probability that outbursts occur intermittently, and after a galaxy has released some quasars it may become quiescent.
An example of an unworked gold mine is the starburst galaxy NGC7541. Described by Arp in the 1968 Astrofyzika article as being between a pair of bright radio sources, it subsequently turned out to have quasars of z = .22, .62, 1.05 and 1.97 around it. From the ROSAT survey, it has a pair of X-ray sources across it, and radio measurements to fainter levels show additional radio sources closely grouped around it. The main galaxy has a straight spiral arm, which looks like an ejection and has an early type stellar absorption spectrum. A close companion galaxy, NGC7537 appears active and might well be a Seyfert or allied type. This is the kind of region which requires a thorough observing program—the kind of program that used to be possible in the era of small telescopes, but is unthinkable in the era of big telescopes.

Summary of Empirical Evidence
In spite of a deliberate effort to avoid them, a large number of cases of quasars undeniably associated with much lower redshift galaxies have accumulated. Based on the discussion of the first two chapters of this book, the unavoidable conclusion, stated as simply as possible, is this:
It is clear that, spectroscopically, a quasar looks like a small portion of an active (Seyfert-like) nucleus. That supports the conclusion, from their ubiquitous pairing tendency across the active nuclei, that they have been ejected in opposite directions from this central point, which shows similar physical conditions. As explained in the introduction, starting in about 1948, it has become an article of firm belief that galaxies eject radio emitting material in opposite directions. The quasars often show radio emission, as well as the other attributes of matter in an energetic state, such as X-ray emission and excited optical emission lines. The only possible conclusion from this observational evidence is that quasars are energized condensations of matter which have been recently ejected from active galaxy nuclei.
We will see later, however, that it will be necessary to consider the quasar to be made of more recently created matter in order to account for its higher intrinsic redshift.

Terminology
It is interesting to recount how the current confusion between some Seyferts and quasars came about. When the luminosities of quasars were computed on the assumption that they were at their redshift distance, it turned out there was continuity with galaxies in that parameter as well as other properties. Maarten Schmidt decided that

as Quasar Factories

55

MV = –23.0 mag. was about bright enough for a galaxy; and that anything brighter than that should be called a quasar. Of course, it has turned out quasars are actually fainter than galaxies, and should be classified on the empirical criteria of compactness and spectral excitation.
Another example of the penalty people pay for not using operational definitions is the term “AGN” (active galaxy nuclei). Once, as I was stepping onto a plane in Santiago, heading back to Pasadena, I met Bruce Margon coming the other way for an observing run in Chile.
“Oh Chip”, he enthused, “I have just decided the new terminology for all these objects: we’ll call them AGN’s.” He was using his theoretical knowledge that quasars were enormously bright nuclei of enormously distant galaxies.
“Absolutely terrible”, I replied, “If you do that you will wreck the empirical classification.”
Eventually, everyone came to believe that quasars had host galaxies. John Hutchings, Susan Wyckoff, Peter Wehinger and others found host galaxies. Assuming the quasars were at their redshift distances, they found host galaxies that were too big— and some examples that were too small. Taking a mean, they reported that their sizes were just right. When the Space Telescope started taking high-resolution pictures of quasars, John Bahcall called a press conference to report that a number of them did not have any host galaxies at all! Gasp! Naked quasars!!
The community was horrified. What was going to sustain the enormous luminous output of distant quasars if they did not have a host galaxy to fuel them? Private meetings were held immediately, and it was rumored that incorrect image reduction was involved. The judgment of doom! The irony here was that Bahcall had been coming on like Ghengis Khan, Tammerlane and Vlad the Impaler to anyone who doubted the redshift distance of quasars. Bahcall then produced some quasars with “host” galaxies, and everyone decided to paper over the issue in public.
There was no need for this chaos because the first quasar discovered (3C48 by Matthews and Sandage; 3C273 was only the first to have its redshift determined) had a nebulous fuzz around it, about 12 arc sec in extent. At a conventional distance corresponding to its redshift of z = .367, this translated into a diameter of 35-70 kiloparsecs, depending on the choice of Hubble constant. That is bigger than the big galaxies we know the most about, e.g. M31 and M81. But many quasars with a z around .3 were observed to have central brightness 3 or more magnitudes fainter than the 16.2 mag. of 3C48. Observed with seeing better than 1 arc sec., many showed no fuzz at all, so any host galaxy would have had to be abnormally small. Figure 2-18 shows a long exposure of 3C48—not with Space Telescope, but with a relatively small aperture 2.2 meter telescope in Hawaii and some image processing. It shows that the quasar has slipped completely out of the alleged host galaxy! What a way to fuel a quasar! What is worse, anyone who bothered to look would see that a huge low surface-brightness envelope surrounds the pair. The galaxy looks very much like a nearby dwarf!

56

Seyfert Galaxies

Fig. 2-18. The nearby quasar 3C48 as registered on coadded photographs with the Hawaiian 2.2 meter telescope by Allan Stockton et al. Note the quasar slipping out of the nucleus of the “host” in the blow up on the right. On the left there appears an extended, low surface brightness envelope around the system which looks like a dwarf galaxy.

If scientists had only heeded the words of Percy Bridgeman on the necessity of scrupulously using operational definitions in science. It would now be natural to describe an empirical sequence of quasar development, from initially point-like objects at relatively faint apparent magnitudes, gradually transforming into lower-redshift compact objects with “fuzz” around their perimeters, then into small, high surface-brightness galaxies with more material around them and, finally, normal, quiescent galaxies.
Trying to Stop the Stampede
When just the most prominent members of operationally defined classes are known, it is usually easier to see the overall relations between them. Figure 2-19 shows the Hubble diagram which I published in June 1968 (Astrophysical Journal 152,1101). The diagram showed that compact galaxies (morphological transitions between galaxies and point-like quasars) had active, Seyfert-like spectra and formed an obvious physical continuity between Seyfert galaxies and quasars. But, as Figure 2-19 shows, this class of objects clearly violated the Hubble redshift-apparent magnitude relation.
Nevertheless, this very Hubble relation is assumed in order to calculate luminosities for these objects. Then the luminosities are used to reclassify them on the basis of a theoretical property, which leads to the chaos described above. I followed the June paper with an expanded version in July 1968 (Astrophysical Journal. 153, L33) in which I showed more members of these classes which were continuous in color properties as well, and even more conspicuously violated the slope of the Hubble line. But my desperate effort did not even slow down the rush to express all measured quantities in terms of great distances in an expanding universe. The juggernaut has continued to gather momentum to the present day.

as Quasar Factories

57

Fig. 2-19. The Hubble diagram for objects with Seyfert like spectra published by Arp in 1968. The solid circles represent nearby Seyfert galaxies, the crosses represent compact galaxies with Seyfert-like spectra and the triangles represent quasars known at the time. The class of objects obviously violates the dashed Hubble line which objects of the same luminosity in an expanding universe must obey.

3C48 as a Key to the Paradigm
We will show in Chapter 5 that 3C273 (the first quasar to have its redshift measured and, on the redshift-distance assumption, discontinuously the most luminous) is an important member of the relatively nearby Virgo Cluster. But the first quasar discovered was 3C48, and from it one could correctly deduce that it was a very strong radio source and a bright apparent-magnitude, stellar-appearing object. One might also suppose that of all members of this class of objects, it was among the nearest to us. Then, if the preceding Chapters have any meaning, one would expect a very bright, lowredshift galaxy to be identifiable as its progenitor at not too great a distance from it on the sky.
Now, one of the brighter galaxies in our Local Group of galaxies is M33, a companion to the dominant M31. M33 is a companion galaxy with a rather young stellar population, and just the kind of galaxy first associated with quasars (see Quasars, Redshifts and Controversies). The quasar 3C48 is only about 2.5 degrees away—exceptionally close for such bright objects! Figure 2-20 shows the configuration with another bright quasar in the region. If M33 were removed to the distance of the Virgo Cluster, the angular separation of 3C48 and paired quasars would be 7.1 and 12.9 arcmin from the galaxy. This is just the range of separations we were finding for quasars at the beginning of this Chapter, around galaxies which were on average at just about the distance of the Local Supercluster center.

58

Seyfert Galaxies

Fig. 2-20. This Figure illustrates the proximity of the first discovered quasar, 3C48, to the bright, Local group companion, M33. On the other side of M33 is the exceptionally bright, high redshift quasar, PKS 0123+25 with V = 17.5 mag. and z = 2.353 (see text).

What about the quasar on the other side of 3C48? Its redshift is z = 2.353, and it is a strong radio source with very bright apparent magnitude V = 17.5 for such a large redshift. (Further out in this region in this direction we see an extension of more highredshift quasars, which apparently belong to M33, as shown in Quasars, Redshifts and Controversies pp. 72-73). But we also know from the just referenced work that the highredshift quasars are less luminous than the lower-redshift quasars. This supports the surprising result that quasars of redshift up to about z = 1.5 can be seen out to the distance of the Virgo Cluster, but quasars of greater than about z = 1.8 are generally not seen much beyond the bounds of the Local Group.
Actually, the PKS quasar in Figure 2-20 is probably a secondary ejection. The candidate for the counter ejection from 3C48 would be a bright BL Lac (15.7 mag., redshift unknown) at 1h 09m 24s and 22d 28m 44” (1950). Because of the rapid evolution of high redshift quasars (z around 2 or greater), we would expect them to be seen rather close to their galaxy of origin. The latter prediction is forcefully born out by the 7 high-z quasars around the Seyfert 1 galaxy 3C120, which appears to be the closest active galaxy to our own in the Local Group. (See page 130 in Quasars, Redshifts and Controversies: That book also contains a chapter on the distribution of high-redshift quasars in space (Chapter 5), which shows their locations in the Local Group, with the strongest concentration southwest of M33 (lower right in Figure 2-20).
Way Back in the Beginning
In about 1951 I was choosing a Ph.D. thesis topic. I had been captivated by the early reports of Karl Seyfert’s discovery of galaxies with brilliant compact cores. I was

as Quasar Factories

59

particularly intrigued with the fact that these cores were rich in ultraviolet light. I guess I sensed this was where there was some action, some mystery. For a thesis, I proposed to photograph these galaxies in ultraviolet light, and see what connection the nucleus had to the galaxy, and whether there were any other ultraviolet objects around.
Rudolf Minkowski, who was Walter Baade’s right hand man, said that was a terrible thesis that would yield nothing. I wound up measuring thousands of little clumps of silver grains (photographic images of stars in globular clusters) in order to calibrate distance indicators in which Baade was vitally interested. Twenty years later, I was finding quasars around active galaxies by photographing them in ultraviolet and blue light and taking spectra of those candidates with ultraviolet excess. Occasionally, I would think on those nights: If I had done that thesis, maybe I would have discovered quasars ten years before they were discovered from radio positions. What difference would it have made to the course of cosmology? Then again, maybe I wouldn’t—and then I would not have gotten the chance later.
Even though the globular cluster thesis-work helped lead to derivations of the age of the oldest stars, and hence to the age of our galaxy, Baade was suspicious of my reliability and did not recommend me for a staff position. It was Allan Sandage who successfully pressed for my appointment, because he thought I would be a great help in determining the Holy Grail of the distance scale that was the keystone of cosmology. But when I started having independent opinions about stellar population types that proved too competitive for Allan, and he wanted to get rid of me. When that did not happen, he refused to speak to me for ten years. Later he began to feel lonely, and we were close confidants for a while. One day he sat in my office and said, “Chip, you’re the only one I can talk to.” Well it was up and down a lot after that, too. But in the end, regardless of everything else, I feel close to him—like someone you have been together with through a tough war. It transcends the issues, and even the opposite sides, because no one else quite understands.

Chapter 3

EXCESS REDSHIFTS ALL THE WAY DOWN

There is a story about a cosmologist giving a public lecture. Afterwards a lady stood up and said, “The real universe rests on the back of a turtle.” He quickly shot back, “Well what is the turtle standing on?” “Young man,” replied the lady, “it’s turtles all the way down.”
For those astronomers who are willing to consider quasars much closer than their redshift distances, there is one great stumbling block. That block is the instilled certainty that “normal” galaxies can only have velocity redshifts. When it comes to intrinsic redshifts in galaxies, most astronomers would consider that to be “turtles all the way down.”
Yet we have already seen signs that quasars are not the only objects in the universe to have intrinsic redshifts. This would almost have to be the case just from considerations of continuity. There is a very obvious continuous progression of empirical characteristics from unresolved high-redshift quasars, through lower redshift compact objects, and finally into normal galaxies. We can argue that this is simply evolution in age, because the compact objects must be young—both from their tendency to expand due to the outward pressure of the concentrated energy, and the fact that the high energy tends to decay unless strongly fueled. Actually though, I was led to look for intrinsic redshifts in companions to large galaxies by an empirical series of results.

Companion Galaxies

The Atlas of Peculiar Galaxies contained a very interesting class of galaxies called spirals with companions (smaller galaxies) on the ends of arms. How had they got there? Certainly not by accidental collision or by the beginning of a merger process, which is fashionably used to “explain” everything in the extragalactic realm. (I actually

Halton Arp, Seeing Red: Redshifts, Cosmology and Academic Science (Apeiron, Montreal, 1998)

61

62

Excess Redshifts

Fig. 3-1. No. 49 in The Atlas of Peculiar Galaxies shows a compact object trailing a wake of material behind it as it passes out through the disk of the galaxy.
read in the Astrophysical Journal once that double galaxies are galaxies in the process of merging and single galaxies are galaxies that have already merged.) I had argued that since galaxies characteristically eject material which eventually forms new galaxies and that if ejection took place in the galactic plane, then it would pull material out in the form of a spiral arm attached to the companion. Figure 3-1 here is No. 49 in my Atlas of Peculiar Galaxies, and it suggests quite plainly what is going on.
Whether or not that is true, I decided to look at the redshifts of the companions to see if, by any chance, they were systematically greater than the larger galaxy. They were, and that started another long running battle which eventually led to a quantitative proof of the dependence of redshift on age.
The clues begin in the Local Group of galaxies centered on our giant Sb spiral M31, historically known as “the Andromeda Nebula.” M31 is the most massive galaxy in our group, and is classified Sb by virtue of its extensive central bulge of old, red stars. Every major companion (by inference, including our own Milky Way galaxy) is positively redshifted, as seen from M31. The next nearest major group to us, the M81 group, is centered on the same kind of massive Sb galaxy and, again, every major companion is redshifted with respect to it!
By 1987, there had been a dozen different investigations, every one of which showed companion galaxies were systematically redshifted (see Table 7-1 of Quasars, Redshifts and Controversies). By 1992, there were 18 different references to studies which showed this effect in the published literature. In spite of all this, a paper then appeared in the Astrophysical Journal. which interpreted companion redshifts as velocities to derive masses of parent galaxies—and referenced none of the 18 papers which showed that

all the Way Down

63
Fig. 3-2. The Local Group (M31) and the next nearest major group (M81). The smaller, companion galaxies are shown to be all of higher redshift, a distribution having one chance in 4 million of being accidental.

the velocity assumption was untenable. I got pretty heated up over this, and after a lengthy battle managed to get an answering paper published in the same journal (Astrophysical Journal 430,74,1994). Figure 3-2 here is taken from that paper.
One interesting development that had taken place was that a new member of the Local Group of galaxies had been found, IC342. This dwarfish spiral was at low galactic latitude (Figure 3-3), and an accurate absorption and distance had only been determined recently. It then became a member of the Local Group at about 1.2 Mpc distance on the other side of M31 from us. At +289 km/sec redshift with respect to M31, it had the largest excess redshift. (Actually this redshift was very close to four times the basic redshift quantization of 72.4 km/sec, a matter that will be discussed further on.) This discovery brought the count to 22 out of 22 of the major companions, all of which had

Fig. 3-3. A spiral of large apparent diameter seen close to the plane of the Milky Way. IC342 is the newest and most distant member of the Local Group.

64

Excess Redshifts

higher redshifts than the dominant galaxy, in the two best known, nearest groups. The chance of this arrangement of galaxies randomly orbiting their central galaxies with equal numbers of approaching and receding velocities was only one in 4 million!

Major Clusters of Galaxies
If companion galaxies in groups have systematically larger redshifts, what about companions (less luminous) galaxies in clusters? One could logically argue that great clusters, like the Local Supercluster, were made up of many groups like the M31 and M81 groups. In fact, it is true that if one looks at the Virgo Cluster (i.e. the center of the Local Supercluster), one finds all the usual morphological types of galaxies. And the smaller galaxies are systematically redshifted with respect to the larger!
One can see this effect in two ways. First one can calculate the mean redshift of the galaxies in the Virgo Cluster by weighting the redshift of each galaxy by the brightness of the galaxy. If luminosity is proportional to mass, then one gets the redshift of the average mass of the cluster, the only dynamically meaningful quantity. This calculation gives a mean redshift for the Virgo Cluster of +863 km/sec. Now the value calculated by assuming all the galaxies have the same mass comes out to between 1000 and 1200 km/sec. Why this striking difference? It is simply because the smaller galaxies have systematically higher redshifts.
The second way to see this effect is to note that late-type galaxies (spirals and young spirals) are systematically redshifted in clusters. Since spirals are generally less luminous than giant E’s, and, further, since their mass-to-luminosity ratios are lower; this shows in a different way that companion (lower-mass) galaxies in clusters are systematically redshifted.

The Redshift of the Virgo Cluster and the Hubble Constant

Sometimes I think that Astronomy is not so much a science as a series of scandals. One of the most egregious is the derivation of the value of the Hubble constant from the Virgo Cluster. There have been innumerable headlines about new distance determinations to the cluster in the past decades, and most recently from Space Telescope press releases. The debate swings between the “long” distance scale (a little more than 20 megaparsecs) and the “short” distance (about 16-17 Mpc). The longer distance is used by the proponents of Ho = 50 km/sec/Mpc. The shorter distance is used by proponents of Ho around 80, the latter having the drastic consequence that the universe is then younger than the oldest stars it contains. (Unless one brings back the cosmological constant etc., etc.)
Although both sides use different mean redshifts for Virgo (ones that favor their preferred value: see Astronomy and Astrophysics 202,70,1988), neither side pays the slightest bit of attention to the fact that they have both made an elementary mistake in computing that mean. In physics, we learn to compute the center of mass of an ensemble of particles by weighting each particle. How can we compute the mean redshift of

all the Way Down

65

the center of mass of a cluster of galaxies without weighting the mass of the galaxies? Of course, astronomers insist on assuming that the low-mass and high-mass galaxies have the same average redshift. If that were so, they would get their usual answer, and they should have no objection to making the more rigorous calculation. In fact, if they defined the dynamical center as the most luminous and massive galaxies (which should not drift away from the rest of the cluster), they would not be able to change the mean redshift of the cluster by adding or not adding negligibly small galaxies over which there is obvious disagreement as to membership.
Another “adjustment” which pushes the derived Hubble constant to higher values is the notion that the mass of the Virgo Cluster attracts our own Local Group, and its consequent “infall velocity” must be added to obtain the true cosmic recession velocity of the Virgo Cluster. The “infall velocity” is the supposed result of the gravitational attraction of the Virgo Cluster on the Local Group. But if masses of galaxies have been generally overestimated, or if peculiar velocities between groups are very small—both of which will be argued later—, then this adjustment cannot be used to increase the Hubble constant, as in the conventional derivation. Moreover, if galaxies on the near side of the Virgo Cluster were falling toward its center, then the brightest galaxies would have the more positive redshifts. The opposite is actually true. Therefore, the 1400 km/sec systemic redshift used for the much-publicized Hubble constant calculations is far from the 863 km/sec actually measured. (In fact, 863 km/sec is an overestimate because luminosities measured in red wavelengths should be used and also the spirals weighted less.)

Late type Spiral Galaxies as Younger Companions
From the beginning, we have noticed the excess redshift of companions around massive central galaxies which had large components of old stars. The implication was that these old stars had been around from the beginning of the group, and that smaller, younger companions had been ejected intermittently as time passed. These central galaxies had morphological types mainly of Sa, Sb and giant E. The smaller companions ranged over the remaining morphological types, but featured dwarf E’s (showing spectroscopic indications of an admixture of a population of stars younger than in the giant E’s) and later-type spirals (SBbc, Sc, Sd and Im). The latter types are marked by conspicuous numbers of bright, young stars. These late-type spirals were measured to have low masses from their rotation characteristics, and low mass-to-luminosity ratios indicative of relatively recently formed stars. Empirically then, the smaller nuclear bulges and open spiral structure of the late-type spirals came to mark them as lowermass, younger “companion” type galaxies.
A special kind of supposedly high luminosity spiral, designated ScI, will be discussed later as really being of low luminosity because of large excess redshift due to its younger age. But for the purpose here of investigating the redshift behavior of companions in major clusters of galaxies, it will be useful to identify companions by their morphological classification as late type spirals.

66
Fig. 3-4. The excess redshifts of the companions are shown as a function of their morphological type.

Excess Redshifts

Late Type Spirals in Major Galaxy Clusters
Figure 3-4 shows the excess redshift of companions as a function of their morphological type in the two nearest groups, M31 and M81. The later type spirals are clearly systematically higher redshift. Figure 3-5 shows the same diagram for the entire Virgo Cluster, and the same pattern is evident. This enables us to check other major clusters, as shown in Figures 3-6 and 7. The end result is that the younger spirals in the nearest groups, as well as the 4 or 5 major clusters of galaxies, all show systematic positive redshifts. There seems to be no escape from this result.
Fig. 3-5. Redshifts of galaxies in the Virgo Cluster as a function of morphological type. The full line is the luminosity weighted mean and the dashed line the number mean. Symbol sizes are proportional to the apparent magnitudes. Note that, as in the nearby groups, the galaxies around type Sb are the lowest redshift and tend to be among the brightest.

all the Way Down

67

Fig. 3-6. Summary of redshift-galaxy type relations for major galaxy clusters from Giraud (1983).

Fig. 3-7. A plot of redshift versus galaxy type for galaxies brighter than 15th magnitude in the cluster Abell 262 (from Tifft and Cocke 1987).
Back to the Virgo Cluster
In Figures 3-4 and 3-5, what is really most apparent is the minimum redshift exhibited by the brightest Sb’s. In the Virgo cluster, galaxies of this morphological type are predominantly low or even negative redshift. One could obtain a very low redshift for the cluster if one were to accept them as the dominant galaxies in the cluster.

68

Excess Redshifts

Fig. 3-8. Redshift-apparent magnitude diagram for all of the SO (open circles) and Sb (filled circles) galaxies which are the most certain members of the Virgo Cluster. The mean redshift for the cluster by various authors is indicated along the right ordinate. The Huchra value includes an infall velocity.
The S0’s (a kind of disk galaxy without bright young stars), which are the most numerous kind of galaxy in the cluster, actually show a continuous gradient of redshift from the brightest to the faintest. Figure 3-8 shows their apparent magnitude-redshift relation. Again one could pick almost any redshift for the Virgo Cluster one wanted, depending on the apparent magnitude of the S0 which one chose to be representative of the mean mass. On the conventional assumption, the S0’s are supposed to define a horizontal line in Figure 3-8! In view of this uncertainty, the best procedure seems to be to make a luminosity weighted integration over the galaxies in the cluster and hope that this averages close to the age of our own Milky Way galaxy, so that there is no age induced differential redshift.
It is encouraging to note that the +863 km/sec derived in this way is very close to the largest, brightest and apparently oldest galaxy at the geometrical center of the Virgo Cluster, M49 (also known as NGC4472). That seems the best bet to be the currently dominant galaxy and it has a redshift of +822 km/sec. If we take the short distance scale to the Virgo Cluster of 16-17 Mpc (in my opinion the more correct one) we obtain a Hubble constant, H0, close to 50 km/sec/Mpc. We will see in Chapter 9 that this fits quantitatively with a non-expanding universe in which the redshift is a measure of the age of a galaxy.
What about the negative redshifts in Virgo (i.e. blueshifts)? People often ask: If intrinsic redshifts are a function of age, can there be negative redshifts? The answer is: Yes, it is required if the galaxy is older than we are, as we see it. Aside from the Local Group where M31 is the parent and we see it as negatively redshifted by –86 km/sec, there are only six major galaxies of negative redshift in the sky. All six are in the Virgo Cluster, and are obviously members. They are chiefly the big Sa’s and Sb’s which we

all the Way Down

69

Fig. 3-9. A pie diagram for all galaxies listed as Virgo Cluster members in the Revised Shapley Ames Catalog by Sandage and Tamman, plotted as a function of their redshifts. Crosses are spirals and later types, pluses are remaining types. Symbol size decreases with decreasing apparent brightness.
have learned to associate with the originally dominant galaxies. Hence, these are probably somewhat older than any of the galaxies in our Local Group, and may represent the original galaxies in Virgo. It is even possible that our Local group originated from them. [It is touching to speculate that when we look at the Andromeda galaxy, we are looking at our parent. Perhaps in Virgo we can gaze at our grandparents.]
Later, we will discuss aggregates of numerous faint smudges which are called distant galaxy clusters. But we will argue that they are generally something different from the great clusters of galaxies like to our own.
Pie in the Sky Diagrams
An enormous amount of modern telescope time and staff is devoted to measuring redshifts of faint smudges on the sky. It is called “probing the universe.” So much time is consumed, in fact, that there is no time at all available to investigate the many crucial objects which disprove the assumption that redshift measures distance. Still, one has to do something with these redshifts after they are measured. What is done is, an area on the sky is selected and all the available measures plotted as a function of their redshift.
As an example, the plot in Figure 3-9 shows what the well-known Virgo Cluster looks like. What a shock! There is a great “Finger of God” pointing directly at us, the observer. Of course, this is hastily explained as due to high orbital velocities for the galaxies in the center of the cluster which invalidate their use as distance criteria. But it

70

Excess Redshifts

Fig. 3-10. The same galaxies as in the preceding diagram but now plotted as a distribution function of redshifts which enables negative redshifts to be included. The luminosity weighted mean is indicated by an arrow.
is not just the center of the cluster which shows these “peculiar” velocities; the whole cluster is strung out. Moreover—and this is the telling point—, the brightest galaxies are preferentially at the lowest redshift. This is shown even more clearly in Figure 3-10, where the negative redshifts in Virgo can be plotted. The fainter galaxies and late type spirals trail asymmetrically away to much higher redshifts. If the elementary precaution of plotting these points in proportion to their brightness had been taken, it would have been obvious that the fainter galaxies had intrinsic redshifts.
Another obvious feature of Figure 3-9 is that the higher-redshift tail drifts off in a different direction from the center of the Virgo Cluster. That cannot be due to velocity dispersion in the center of the cluster. These must be smaller galaxies in a somewhat different part of the cluster, but with a continuity of increasing intrinsic redshift. This one feature, by itself, is disproof of the redshift-equals-velocity hypothesis.
Nevertheless, region after region of the sky has been presented in journal articles and public lectures that present the Fingers of God as velocity dispersions and show how the universe is made up of bubbles and voids. When people occasionally question

all the Way Down

71

Fig. 3-11. Illustration of what happens when one takes a spherical galaxy cluster with the brightest, lowest redshift galaxies in the center and then plots them in a pie diagram with redshift assumed to be a measure of distance.
this orgy of Swiss cheese universes, the answer is always the same: anyone who does not believe redshifts are measures of distance is termed a “psychoceramic artifact.”
Blowing Bubbles and Digging Voids
Considering what we know about a group or cluster of galaxies, let’s look for a moment at how plotting them in a pie diagram would distort the picture. Figure 3-11 shows the large, low redshift galaxies at the center with the smaller, intrinsically redshifted galaxies distributed around them. As soon as we plot with redshift as a distance indicator, the large low-redshift galaxies pull out of the center, leaving a ring or bubble.
There are many ways one could elaborate on this picture. If the fainter galaxies have plunging orbits, that would elongate the ring along the line of sight. If there is a component of rotational orbiting, that would fatten the ring toward the edges. An approximate mixed velocity dispersion is shown in the right hand panel of Figure 3-11. Of course, this is all under the usual default assumption that the cluster or group is in equilibrium. One could find a variety of forms if groups of younger galaxies were moving away from the central galaxy.
Since we know that the central, larger galaxies have the lowest intrinsic redshifts, it will now be necessary to go back and carefully correct the inferred distributions of galaxies in different directions in the sky.
Further Evidence for Excess Redshifts of Companion Galaxies
Shortly after the publication of my 1994 paper on the subject of companions described earlier, I was walking past the journal rack in the library when a paper on “Arp 105” caught my eye. Curious, I skimmed it and quickly ascertained that, as with so many other objects from my Atlas of Peculiar Galaxies which were prime examples of

72

Excess Redshifts

Fig. 3-12. Photograph of Arp 105 (NGC3561B). Ambarzumian’s knot is seen ejected due south from this active elliptical and an opposite ejection northward appears to be puncturing the disturbed spiral. Redshifts of galaxies measured by Duc and Mirabel and quasar of redshift z = 2.19 discovered by Alan Stockton are indicated.

ejection, this was also being presented as an example of collision and merger. As Figure 3-12 and the color presentation on the back cover of this book testifies, this was a particularly inappropriate interpretation, because it was one of Ambarzumian’s finest examples of protogalaxies being ejected, jet-like, from an active elliptical galaxy. Exactly opposite this was the counter jet, a magnificent straight plume punching through a disrupted spiral. Fritz Zwicky, after looking at his spectra of the knots in the jet, had remarked that these were the only galaxies he knew that were not resolved with the 200-inch telescope. Allan Stockton had discovered a quasar of redshift z = 2.2 so close to this ejecting galaxy that the chance of accidental occurrence was less than one in a thousand.
I was about to return the paper to the stand with exasperation when I noticed that the authors had measured the redshifts of most of the companions. What they had overlooked, and what leaped off the page, was that they were all positively redshifted with respect to the dominant galaxy. Since the authors claimed these companions were colliding with what they termed a “giant E”, there was no question that they believed the galaxies they had measured were bona fide companions at the same distance as Arp 105. It did not matter whether they were orbiting the central galaxy, falling in, or being ejected outward—one should roughly expect just as many relatively plus as minus

all the Way Down

73
Fig. 3-13. Distribution of redshifts of companion galaxies around the “massive elliptical” Arp 105 as measured by Duc and Mirabel.

velocities on the average. As Figure 3-13 reveals, all 9 of the lowest redshifts (actually 10 if one were to include the one at +1100 km/sec, relative redshift) are higher than the central galaxy redshift. One had here another case, like the Local and M81 groups, where the intrinsic excess redshift of the companion had overcome the smaller plus and minus velocity dispersion.
One of the reasons this was a particularly satisfying confirmation was that this was a somewhat different kind of central galaxy, much rarer, caught in the act of ejecting. It had a much higher mean redshift than the more local groups that had been tested. In addition, there was an unusually large number of companions.
While I was writing this result up for communication, a preprint crossed my desk. A new investigation of the Hercules Cluster of galaxies had shown that in every subsection of the cluster, the late-type spirals (companions) had conspicuously higher redshifts than the early-type galaxies in the same sector. This was impressive, because it was a detailed confirmation of the results for companions in clusters.
Finally, simultaneously with the above, a student in Holland sent me one of the secondary findings in his thesis. While investigating galaxies in the Bootes void, he had discovered that 78% of the companions around his galaxies had positive redshifts relative to the dominant galaxy. Figure 3-14 shows this very strong confirmation in a large sample of galaxies.

Fig. 3-14. Excess redshift for companions in a sample of galaxies in the Bootes void and comparison fields as measured by Arpad Smozuru. Plotted as a function of the distance from the central galaxy.

74

Excess Redshifts

Fig. 3-15. For Paul Hickson’s catalog of compact groups of galaxies, the amount by which the brightest galaxy is brighter than the second brightest is designated Δmag. The redshift of each galaxy in the group minus the redshift of the brightest in the group is called Δcz. The histogram shows that for mag. > .2 mag, the fainter companions are systematically redshifted.
Trying to Publish Further Results
Putting all these results together, they seemed to me to offer decisive proof of excess redshifts in companions. But the author who used the companion redshifts as velocities, without referencing the contrary evidence, wrote an angry letter to the editor complaining that I had been rude in my manner of pointing this out. Another pair published a rebuttal paper claiming complex orbits could explain the preponderance of positive companion redshifts! When the “Further Evidence” paper went to the referee, he suggested the interacting companions around Arp 105 belonged to another galaxy outside the pictured area. There were hints that the thesis student who found the excess companion redshifts would be in big trouble. After holding the paper for three months, one referee sent a Xerox from a 1902 book on celestial mechanics plus a graph showing the moon orbiting around its barycentre. Another referee said a study of weak galaxy clusters showed the largest galaxies to have the same redshift as their cluster. When I analyzed that data, the same result turned up—the brightest galaxies had –355 km/sec lower redshift. The referee replied to the editor: “Perhaps the author did not understand that I have rejected the paper.”! The editor rejected it.

all the Way Down

75

At a conference, one of these referees gave a rather startling (for conventional beliefs) lecture on how Fourier analysis could not be trusted, and mentioned that ergodicity did not ensure that the ensemble average was equal to the time average! He said he was eager for more data on this important subject of companions!! But after four years, this further evidence had not been published in a major journal. The only result is a stack of insulting letters from referees and editors.
One thing has been accomplished, though. I now understand what should be called the statistics of nihilism. It can be reduced to a very simple axiom: “No matter how many times something new has been observed, it cannot be believed until it has been observed again.” I have also reduced my attitude toward this form of statistics to an axiom: “No matter how bad a thing you say about it, it is not bad enough.”

Compact Groups of Galaxies
The first compact group was discovered with the Marseille telescope in 1877 by M. E. Stephens. In 1961, Margaret and Geoffrey Burbidge measured redshifts of the five galaxies and showed they were 800, 5700, and three at 6700 km/sec (see Figure 319). Now the 5700 and one of the 6700 galaxies were entwined together. If redshifts were interpreted as velocity, this meant they were separating at 1000 km/sec. Even in conventional terms, galaxies don’t move that fast; and even if they did, the chance of catching two at just the moment of collision would be very small. And, of course, the gas would not keep two separate velocities. From that time forward it should have been clear we were dealing with non-velocity redshifts.
But as you might suppose, the picture has become increasingly muddied with mergers, dark matter and gravitational lenses, while any redshifts which do not fit a conventional theory are placed in the foreground or background. Is there anything new? Well, an observational advance has been made by Paul Hickson, who catalogued, photographed and measured redshifts in a sample of 100 compact groups. (A compact group is defined as four or more galaxies crowded together by a factor of 10-30 more than their local surroundings.) The Catalogue made it possible to test the following proposition: Since compact groups are in many cases denser versions of normal groups in which companions have excess redshifts, do compact groups with a dominant galaxy have systematically redshifted companions? Figure 3-15 answers this question by showing that, as the difference in apparent magnitude between the brightest and next brightest galaxy becomes larger, the number of positively redshifted companions becomes larger. This makes sense, because if the galaxies are all the same brightness, one does not know which is dominant and the effect is untestable. But the fact that when one galaxy becomes clearly dominant the effect emerges—this demonstrates that non-velocity effects are present in the compact group galaxies, just as in every other group tested.
This point is strikingly illustrated in Figure 3-16 where the distribution of companion redshifts in compact groups with the most dominant galaxies is compared to the Local Group. Actually most groups have companions with up to about 800 km/sec higher redshift, and it is obvious that the Local Group is missing companions above

76

Excess Redshifts

Fig. 3-16. This figure compares the excess redshifts of companion galaxies in the compact groups versus those of the accepted members in the Local (M31) group, then the fainter companions in the M31 line and finally to the companions in the small Sculptor group which is between the M31 and M81 group.

about 300 km/sec. The reason is quite simple—namely that astronomers are just unwilling to call any galaxy more than 300 km/sec higher than M31 a member of the Local Group because that makes the preponderance of positive redshifts embarrassingly obvious.
If one examines the brightest galaxies as they fall on the sky, however, it is immediately apparent that there is a loose string of them running out of M31, through M33 and ending close to 3C120 near the disk of our galaxy. (See Figure 8-1 in a later chapter). These galaxies have redshifts up to 900 km/sec and are obviously members of the Local Group. A group of later-type spirals called the Sculptor Group is located closer to us than the M81 group. As the last panel in Figure 3-16 shows, it also has higher redshift companions. (Details are available in Quasars, Redshifts and Controversies page 131 and Journal of Astrophysics and Astronomy. (India) 1987, 8, 241.)
An earlier study of what I had called “multiply interacting galaxies” comprised the most striking examples of what later came to be called the “compact groups.” What I pointed out in that original study was that these multiply interacting groups preferen-

all the Way Down

77
Fig. 3-17. Number of discordant (Δcz > 1000 m/sec) redshifts as a function of Δcz for compact groups. Lines of plus signs show expected distribution for background interlopers. Arrow points to preferred redshift peak of Δz=.061 found in all sky measures of quasar and quasar-like objects.

tially occurred near large, low redshift galaxies. In some cases, for example NGC3718, the high redshift, interacting group could be seen actually bending back the spiral arms of the larger galaxy. (Consult the picture on page 94 of Quasars Redshifts and Controversies.) This result made it clear that the compact and interacting groups were just a more concentrated ensemble of young, non-equilibrium companion galaxies which had been ejected more recently from the parent galaxy, and were composed of material of higher redshift. Aside from being empirically true, this interpretation solves all the conventional paradoxes of the failure of the galaxies to merge into a single galaxy on a cosmic time scale, and also explains the unbearable presence of “discordant” redshifts. Of course, none of this is conceded by the conventional army.
Large Excess Redshifts in Compact Groups
We have just seen that the so-called accordant group members (defined as having redshifts different from the group by less than 1000 km/sec) demonstrate again that the fainter members have the higher redshifts. But most shocking of all, there are a number of (mostly) fainter galaxies that fall in these compact groups which have redshifts thousands and tens of thousands of km/sec greater than the group (Figure 3-17).
The consternation caused by the apparent membership of these highly discordant galaxies has led to a blizzard of papers arguing that, despite appearances, they were just projected background galaxies. Just in case, it was also argued that they were gravitationally lensed background objects. To be triply safe, it was also argued that we were seeing filaments of galaxies end on—like looking down a straw with a galaxy stuck on the far end. The only trouble is that in the famous case of Seyfert’s Sextet, the length of the straw had to be about 26,000 times its diameter (see Astrophysical Journal. 474, 74, 1997)!

78

Excess Redshifts

Fig. 3-18. Schematic representation of features of interest around the active Seyfert NGC4151. Note especially the companion galaxies at 6400 and 6700 km/sec.

Figure 3-17, however, shows with small plus signs how background galaxies should increase sharply in number with increasing redshift. The numbers of discordant redshifts actually decreases precipitously in this direction. It seems to me that at a glance, it is clear the discordants are not background galaxies.
Companion Galaxies of High Redshift
This is a very important link in the argument that material is ejected from large galaxies, initially with very high intrinsic redshift, and then ages and expands into compact, active galaxies of moderately high redshift, and finally into normal companions with only slightly excess redshift. So far we have shown that the extensive evidence which already existed has been enormously strengthened by new evidence that “normal companions” belonging to dominant galaxies have excess redshifts in the hundreds of km/sec. Companions with excess redshifts of thousands to tens of thousands of km/sec establish a compelling continuity to the quasars which start at about 20,000 km/sec excess redshift and go up to nearly the velocity of light (if they were really velocities).
Unfortunately, there is not much in the way of new results on this group. In 1982, a list of 38 (yes, thirty-eight) of these high redshift discordant companions was published. They were discussed in two Astrophysical Journal. papers and in a chapter starting on page 81 of Quasars, Redshifts and Controversies. Yet despite the fact that almost every one of these objects is a fascinating study in itself, no further study of these key objects has been made! Certainly these crucial, discordant redshift galaxies have been deliberately avoided by the world’s biggest and most expensive modern telescopes.
To mention just two examples in order to reemphasize the importance of these kinds of companions, I show in Figure 3-18 a schematic of the large, active Sb, NGC4151. (Deep photographs of this galaxy can be seen in Quasars, Redshifts and

all the Way Down

79

Fig. 3-19. Region around the large Sb, NGC7331 and Stephan’s Quintet. Line contours represent radio emission. Note especially the companion galaxies at 6300 to 6900 km/sec around NGC7331 and the 6700 km/sec companions around NGC7320, the low redshift galaxy in the Quintet.

Controversies, pp. 91 and 92). We saw in Chapter 2 how this Seyfert was flanked by two pairs of quasar candidates in an apparent ejection cone, with a strong X-ray BL Lac inside this cone. But also around NGC4151 are associated large companions with redshifts between 6400 and 6800 km/sec. Two of these companions, NGC4156 and G1, lie at either end of the two major spiral arms. With their similar redshifts, are they not like the pairs of quasars discussed in Chapters 1 and 2? With the material of the arms trailing behind them, are they not reminiscent of ejection in the plane, as conjectured at the beginning of this chapter?
But perhaps equally striking is the numerical value of the excess redshifts of these major companions to NGC4151. If one refers to Figure 3-19, one sees that three galaxies in Stephan’s Quintet also have 6700 km/sec redshift and the three galaxies roughly on the other side of the large Sb galaxy, NGC7331, have 6300, 6400 and 6900 km/sec redshift. Galaxies come in groups, and there is no other group leader for these ~6700 km/sec companions to belong to other than the big galaxies at their center. In later chapters we will show that galaxies and quasars tend to occur at certain preferred redshifts. This quantization implies that galaxies do not evolve with smoothly decreasing redshifts, but change in steps.
ScI Spirals as Young, Low Luminosity Galaxies
The companion galaxy NE of NGC4151 has the sharply defined spiral arms which define it as an Sc spiral of luminosity class I. This highest luminosity class is assigned because these galaxies characteristically have moderately high redshifts, which are taken to indicate large distances and high luminosities. Since it is attached to the low-redshift (978 km/sec) NGC4151, it in fact must have an intrinsic redshift and a low luminosity. The same is true of NGC7319, a high redshift ScI galaxy in Stephan’s

80
Fig. 3-20. The excess of the redshift distance over the Tully-Fisher distance plotted as a function of the luminosity of Sc galaxies. Filled circles represent redshifts > 1000 km/sec. This graph demonstrates that for the highest redshift, supposedly most luminous Sc’s, the redshifts give distances too large by huge amounts.

Excess Redshifts

Quintet, which must be at the distance of the low redshift NGC7331 (1114 km/sec) (see Figure 3-19).
How can we check this result? There is a method of estimating distances to galaxies, called the Tully-Fisher method, which uses the rotation of a galaxy to judge its mass, and thus its luminosity, and then its distance by how faint it appears. In Figure 320 we see the difference between the redshift distance and the Tully-Fisher distance plotted as a function of the supposed luminosity of the galaxy. We see that for normal spirals, the two methods are calibrated to give the same distance. But for the highluminosity spirals (ScI’s), the redshift distance is too great by up to almost 40 Mpc! This huge error demonstrates that the redshifts of the ScI’s are too high.
A vivid illustration of how wrong astronomers’ estimates of the sizes of ScI galaxies are is shown in Figure 3-21. The large galaxy is the ScI spiral NGC309 at its sup-

Fig. 3-21. The Sc, luminosity class I, NGC309, if it were at its conventional redshift distance would be so huge that it would swallow one of the largest galaxies of which we have certain knowledge, the Sb M81 (shown as an insert in the lower right between the arms of NGC309).

all the Way Down

81

Fig. 3-22. The Sc galaxy NGC450 has a redshift of 1,900 km/sec and the smaller galaxy to the NE which is apparently interacting with it has a redshift of 11,600 km/sec. The three HII regions in NGC450 near the point of contact with the high redshift galaxy are unprecedentedly luminous and could only reasonably be explained by interaction.
posed redshift distance. The small oval insert shows the size of one of the largest galaxies for which we know an accurate distance, M81. The giant M81 is swallowed like a knot in the arm of the unbelievably large ScI. This picture was published in the April 1991 issue of Sky and Telescope, and the professional astronomers who saw it gasped in astonishment.
But the paper with the analysis was thrown out of the Astronomical Journal with great prejudice. When published in Astrophysics and Space Science 167, 183 it detailed a number of other cases where ScI’s could be shown to be low luminosity, intrinsically redshifted galaxies. I speculate that the sharp, well-formed arms are young ejections before they have had time to be deformed and to spread out. But most astronomers are willing to suppress this observational evidence in order to protect the key assumption about extragalactic redshift from re-examination.
The “Non Interacting” Companion to NGC450
One case that was further investigated is the peculiar Sc spiral NGC450, shown in Figure 3-22. It has a redshift of 1,900 km/sec, and the apparently interacting companion has a redshift of 11,600 km/sec. Just at the point of interaction there appear three enormous HII regions at the redshift of the Sc galaxy. These were so gross that the expert who first spotted this system just assumed they were foreground stars. These glowing regions of excited hydrogen gas are so exceptional that I frankly cannot see how anyone with reasonable common sense and good judgment would not immediately realize that they are a result of the unusually close interaction with the companion.
Nevertheless, a pair of astronomers measured some rotation curves in the system, pronounced them “normal,” and published a paper proclaiming “Non-Interacting” in the title. There would be nothing new to report if it was not for the Spanish astronomer Mariano Moles, who had long been intrigued with this system, and unknown to me, had conducted an extremely thorough observational project of photometry, spectroscopy

82
Fig. 3-23. The HII region which is at the SW end of the high redshift companion galaxy has a redshift that indicates it is falling from NGC450 into the companion. This picture shows a short luminous tail, supporting that interpretation.

Excess Redshifts

and imaging on it with the moderate aperture telescope at Calar Alto. His analysis demonstrated six different observational results, all of which led to the conclusion: “ ... one would have to invoke an enormous conspiracy of accidents in order to avoid the conclusion that [the companion] is a moderately low luminosity galaxy interacting with NGC450.”
One particular aspect was especially pleasing to me. It involved the circumstance that on one of my last runs on the 200-inch telescope at Palomar, I had measured the redshifts of the bright HII regions on the companion side of NGC450. In particular, I had gone after the fourth and faintest HII region, which was just at the end of the high redshift companion, where the companion spread out in an apparent interaction effect with the lower redshift galaxy. It was a difficult observation, and I had to use the Oke multichannel spectrophotometer (commonly called the gold Cadillac). But the emission lines were strong and I got good measures, which I reduced before leaving California for Europe. The redshifts showed larger than normal differences of about 100 km/sec, but the faintest, near the end of the companion, showed a plus redshift of 400 km/sec, well in excess of escape velocity from NGC450.
That 400 km/sec measure enabled a very satisfying model of the interaction to be constructed. It was simply that NGC450 was rotating clockwise, and the companion was coming up from behind it. As the spiral arm of NGC450 approached the companion, its gas was being pushed back and was accumulating and forming the very large HII regions. The companion, which is wedging itself in between the spiral arms of NGC450, was close enough so that the nearest HII region was actually beginning to fall into the near end of the companion. The unexpected confirmation of this came from the hydrogen emission image, which showed a trail of excited gas as this fourth HII region fell toward the high-redshift companion (See Figure 3-23).

all the Way Down

83
Fig. 3-24. The bright galaxy NGC4448 at a redshift cz = 693 km/sec surrounded by non equilibrium companions having redshifts from 5,200 to 36,000 km/sec.

The Referees Go Ballistic
When this new paper, with six authors, was sent to the Journal it elicited furious rejections by two referees in a row. Anonymous messages such as “ludicrous” and “bizarre conclusions based on an extreme bias of the authors wishing to find noncosmological redshifts” were forwarded. One referee suggested that since we knew from the redshifts that the galaxies could not be interacting, the system should be adopted as a control for testing interaction evidence in other groups.
The principal author was so appalled he considered giving up research. But by a great stroke of fortune he asked for a third referee, who turned out to be a breath of sanity. Carefully enumerating all the ways in which this new study presented better observations than the previous ones, the last referee showed how the conclusions were properly drawn from the new data and also commented that the second referee seemed too angry to give a fair assessment of the worth of the paper.
Jubilation that the paper was finally published has to be tempered with the cold experience that much fewer than 1/3 of the referees in this field are objective. Disappointing also is the fact that even though this observational paper was published in a major journal, no notice was taken of it. I relate this story in detail because I think it reveals in the most telling way what the situation is in this particular branch of science. The facts can be consulted in Astrophysical Journal. 432,135 and references therein.
The Environs of the Average Bright Spiral
When pressed, skeptics usually complain that the examples are selected, that they don’t represent a complete sample. But when a complete sample is carried out, for example a survey of 99 bright spiral galaxies carefully compared to non-galaxy control fields, and it shows interacting and peculiar companions are significantly associated with the central bright spirals (Astrophysical Journal. 220,47)—then the results are ignored. Not enough observing time was allocated to complete the measurements of the redshifts,

84

Excess Redshifts

Fig. 3-25. The dusty, starburst, barred galaxy, NGC1808, is imbedded in a dense cloud of fainter galaxies which are undoubtedly of much higher redshift.
but that was really not necessary, as anyone could tell by looking at the galaxies that they were medium-high redshifts.
An example of such a galaxy is shown in Figure 3-24. The galaxy is NGC4448, and the analysis in Astrophysical Journal, 273,167 shows that the numerous, peculiar faint galaxies have redshifts ranging between 5,200 and 36,000 km/sec, while the central galaxy is at 693 km/sec. Another galaxy just embedded in a dense cloud of fainter, certainly higher redshift galaxies is the starburst, dusty NGC1808 in the southern skies. That is shown here in Figure 3-25.
The Origin of Companion Galaxies
The ejection of quasars from active galaxies documented in Chapters 1 and 2 leads to an extraordinarily important synthesis which I did not at first fully appreciate. It was not until after the later chapters on evolution of clusters of galaxies from clusters of quasars that I realized what the data did was to establish the origin of companion galaxies as the end point of the evolution of quasars!
To understand how we come to this result, one must go back to 1957 when Viktor Ambarzumian, from just looking at galaxies on Sky Survey photographs, proposed that young galaxies were born from material ejected from older, active galaxies. Independently I reached the same conclusion from my Atlas of Peculiar Galaxies in 1966. By 1969, the much respected Swedish astronomer, Erik Holmberg, was visiting the Mt.

all the Way Down

85

Fig. 3-26. X-ray map of the Seyfert galaxy NGC2639. A line of X-ray sources coming out exactly along the minor axis to the NE has four identified BSO’s and Bcg’s (quasars to be confirmed— see text for positions). The major pair of quasars measured by Margaret Burbidge is indicated at z = .323 and .305.

Wilson and Palomar Observatories in Pasadena. After 20 years of studying groups of galaxies, he was in possession of some startling evidence—namely, that companion galaxies were preferentially distributed along the minor axis (rotation axis) of the dominant galaxy. As a young researcher at the Observatories, I discussed with him Ambarzumian’s evidence for linear ejection of new galaxies, my evidence for ejection of radio quasars and pairs of objects across disturbed galaxies, and my most recent evidence (1969) that proto companions ejected in the plane of a spiral were stopped very close to the ejecting parent.
He agreed that his alignment of companions along the minor axis was strong evidence for the ejection origin of companion galaxies. But he would not utter a word of this at the Observatories for fear of being ridiculed. I was disappointed, because I badly needed support for my findings. Some time after he had gone back to Sweden, however, his paper appeared in his country’s Arkiv. f. Astronomie To my delight, he forthrightly stated: “ ... physical satellites of spiral galaxies are apparently concentrated in high local latitudes and ... favor systems which have [blue nuclear colors] and contain large amounts of gas. The results seemingly point to one interpretation: that the satellites have been formed from gas ejected from the central galaxies.” (Italics added for emphasis.)
What the X-ray quasar data showed in 1996, and what I did not immediately grasp, was that the quasars were also preferentially ejected out along the minor axis! This was first apparent in NGC4258 where the quasars were only 13 and 17 degrees away from the minor axis (Fig1-1). Then came NGC4235 (Figure 2-5) where the pair were only 2 and 12 degrees away from the minor axis of a clearly defined, nearly edge-on spiral. Finally NGC2639, pictured here in Figure 3-26, shows a group of seven X-ray sources coming out exactly along the NE minor axis. These latter, closer sources are apparently most

86

Excess Redshifts

recently ejected. The outer pair of quasars may represent earlier ejection when the minor axis was rotated in a somewhat different position. In general, such minor axis rotation could account for the greater spread in minor axis alignment of the older companion galaxies, as summarized in Table 3-1.
The fainter X-ray sources coming out along the NE minor axis of NGC2639 contain four optically identified BSO’s or Bcg’s. These blue stellar objects and compact galaxies are predicted to be less luminous, higher-redshift objects on their way out of the nucleus of NGC2639. In the naïve hope that they might someday be spectroscopically observed, I give their exact positions in Table 3-2.

It is not always possible to obtain cases of ejecting galaxies where the major axis (and hence the minor) is well defined. An example is NGC1097, a barred spiral, where the position of the major axis is necessarily uncertain by about 10 degrees (see Figures 2-7 through 2-9). Given this uncertainty, however, the four nearest quasars are within

Table 3-1 Companion Objects around Spiral Galaxies

No.
2 2 + (4) 2 4 6 218 96
115 12

Companions
quasars across NGC4258 quasars across NGC2639 quasars across NGC4235 quasars nearest NGC1097 quasars nearest NGC3516 compns around 174 spirals distbd. compns around 99 spirals compns around 69 spirals compns of M31

ΔΘ1 ΔΘ2
13° 17° 0° 13°(31°)
2° 12° ~ 20° ±20° ~35° ~60°
~35° ~0°

r1 ~ r2
25-30 kpc 10-400 500-600 100-500 100-400 40 kpc 150
500 (700)

Reference
Pietsch et al. 1994 Figure 3-26 Figure 2-5 Arp 1987 Chu et al. 1997 Holmberg 1969 Sulentic et al. 1978
Zaritsky et al. 1997 Arp 1987

Table 3-2 Properties of X-ray Sources in the NGC2639 fields

Name

X-ray (ctsks–1)

R.Α.

Bright X-ray sources in Figure 3-26

RX J08443+5031

37.8

08h44m1930

NGC2639

13.5

8 43 37.9

NGC2639 U10

25.7

8 42 30.0

X-ray sources NE of NGC2639

2.4

8 44 46.1

4.1

8 45 04.4

2.0

8 44 25.3

1.3

8 44 48.7

1.4

8 44 31.8

2.6

8 44 07.2

1.2

8 44 17.0

Dec.
+50°31’36” 50 12 19 49 57 51
50 22 54 50 21 30 50 20 37 50 20 34 50 16 50 50 16 28 50 15 09

Off axis (arcmin)
20.6 0
17.7
14.9 16.4 11.0 13.8 9.5 6.0 6.7

Ident.
QSO z=.323 Seyfert z=.011 QSO z=.305
BSO 19.2 mag. no ident. ambiguous BSO 19.9 mag. BSO 18.3 mag. BSO 18.8 mag.
——

all the Way Down

87
Fig. 3-27. Distribution of companion galaxies and quasars along the minor axes of ejecting disk galaxies. The companion galaxies are at angles of approximately ±35 degrees (from Holmberg 1969 and Zaritsky et al. 1997). The quasar distributions are ±20 degrees from recent data discussed in this book. The observed size of galaxy groups is about 1 Megaparsec (3.26 millon light years).

about ±20 degrees of the estimated minor axis, as noted in Table 3-1. This places them just between the long, luminous optical jets which emerge from the nucleus and must represent some form of ejection.
The most compelling evidence for the origin of companion galaxies is certainly their coincident alignment with quasars. Figure 3-27 shows how the quasars and companion galaxies occupy the same volume of space along the minor axis of the ejecting galaxy. Together with the smaller, but systematic, excess redshifts of the companions, there seems to be no alternative to the conclusion that the quasars are ejected as more recently created matter, and that their intrinsic redshifts decay with time. The morphological and spectroscopic evidence shows them to be evolving into more normal galaxies. (Ejection along the minor axis involves no rotational component of motion and hence the objects remain on radial orbits as they age.) It will be discussed in Chapter 8 how the intervals of quantization of the quasar redshift values also decay into the smaller quantization values observed in companion galaxies.
We will discuss in Chapter 9 how the Narlikar/Arp application of the mass creation theory predicts initially rapid ejection of low-luminosity, high intrinsic-redshift objects, followed by a slowing and final stop out at about 400 kpc—just the range within which quasars and companion galaxies are found. As they continue to increase in luminosity, they slowly start to fall back, roughly (if not perturbed) along the line of original ejection. They also continue to diminish in intrinsic redshift as they evolve into normal galaxies, as shown in Figure 9-3. All these properties are observed—and cannot be explained on the assumptions of the Big Bang theory.
Spectacular Confirmation
As this book was being finished, word was received from Prof. Yaoquan Chu that he had measured with the Beijing telescope the new X-ray candidates around the extremely active Seyfert NGC3516. Fig. 11 of A&A 319,36,1997 shows the X-ray map derived by Arp and Radecke from the archive observations. There are five X-ray sources marked there, which Chu confirmed as quasars. Figure 9-7 shows their redshifts. A quick check of the minor axis direction revealed they all lay within about ±20 degrees of the minor axis. Together with the bright BL Lac type object to the NW, that

88

Excess Redshifts

Fig. 3-28. Atlas of Peculiar Galaxies #65 showing companion galaxies on the ends of two long straight arms strongly suggesting an ejection origin.

meant six quasars coming out along the minor axis of this violently active Seyfert, all more closely aligned than the average of the many Holmberg companions.
But this did not exhaust the dynamite in Chu’s observation. It turned out that the quasars were ordered, with the highest redshifts closest to NGC3516 and the smallest redshifts furthest away. Moreover, the redshifts were all very close to the quantized redshifts to be discussed in Chapter 8. I think the reader can already sense the exultation with which we received this news. Here was an observation which fulfilled every prediction as discussed in Chapter 9, was an incontrovertible confirmation of the sum of past observations, and which we knew eventually would ensure that 30 years of struggle would be of value.
Companions Ejected in the Plane
In the beginning of this chapter, I mentioned that it was my initial idea from studying photographs that if protogalaxies were ejected in the plane of their originators, they would pass through a phase of being companions on the ends of spiral arms. This idea was abetted by my belief that spiral arms were the result of ejection processes and that companions on their ends were related to the large “knots” one often saw along spiral arms.
Figure 3-1 shows a compact, luminous object emerging from the center of a galaxy trailing material behind it. Figure 3-28 shows two small companion galaxies on the ends of two long straight arms. Both of these pictures are from the Arp Atlas of Peculiar Galaxies. This means that already in 1966, we had pictures which showed at a glance that galaxies ejected compact objects which evolved into companion galaxies. Because knots in spiral arms were usually dominated by glowing HII regions, they were pre-

all the Way Down

89

Fig. 3-29. The barred spiral NGC1672 has strong X-rays coming from its Seyfert nucleus. The diametric pair of X-ray sources across its nucleus suggests a pair of objects has been ejected in the plane of the galaxy and slowed down by interaction. (Picture from W.N. Brandt, J.P. Halpern and K. Iawasawa)
sumed to be excited by hot, recently formed stars. Was there something faint, of higher redshift inside, that was masked by this gas of the parent galaxy? Or had the bullet passed on out, leaving star condensation to take place in the entrained gas of the galaxy (perhaps constrained in the magnetic tube of the spiral arm)? To answer such questions required observational hard work, which was obviously not forthcoming.
But the broad thrust of the observational inferences was helped by the X-ray observations reported in the first two chapters. There we saw that the newly created quasars which passed far outside the bounds of the galaxy had a strong tendency to lie along the rotation axis—or at least not in the plane. Were there any examples where the X-ray ejection had gone off in the plane? There may be some which have not yet been recognized, but one clearly probable case was called to my attention by the Japanese astronomer Awaki.
Figure 3-29 shows the barred spiral NGC1672. This galaxy has strong X-rays coming from its Seyfert nucleus, as well as X-ray sources coming from two diametrically opposite points, just at the ends of its bar where the curved spiral arms begin. We know that X-ray sources are ejected from the nuclei of active galaxies. What happens when they are ejected in the plane of the galaxy? Whatever their nature, they will be slowed down more going through the material in the plane than if they were ejected out of the plane. That means they will go through their rather rapid initial evolution closer to their galaxy of origin. If they evolve completely into a companion galaxy, they can then become higher-redshift companions connected to, or still interacting with, their galaxy of origin.
What do we see at the position of the two X-ray sources in NGC1672? Not much on routine low-resolution images—just the high surface brightness of the bar. But galaxies typically contain a lot of obscuring dust in the plane, particularly barred spirals

90

Excess Redshifts

Fig. 3-30. Space telescope picture of the Seyfert 2 galaxy Mark573 by Wilson, Falcke and Simpson. Contour lines represent positions of radio sources. Notice Hydrogen alpha gas of the galaxy is drawn out along line of radio ejection.
that often have thick lanes of dust running out along the bar (See picture of NGC1097 in Figure 2-9). The Japanese satellite telescope, ASCA, which detects higher energy Xrays, registers the western source much stronger than the lower energy X-rays of ROSAT. This implies very strong dust absorption. If there were a highly obscured BL Lac object at the position of X-3 in Figure 3-29, how would we detect it? Even with advanced infrared equipment on large-aperture telescopes, we could have trouble identifying a faint object and getting a definitive spectrum. But that is not to say we should not try—eventually we should succeed in identifying what those strong X-ray sources are.
Another example of what I would take to be ejection in the plane is shown in Figure 3-30. The Space Telescope photograph of the Seyfert 2 galaxy, Mark573, shows a pair of radio sources ejected in opposite directions from a radio nucleus. Hydrogen alpha gas seems to form bow shocks around these ejected sources. But material from the galaxy is clearly drawn out in these ejections.
It Almost Never Happened
As important as I believe the intrinsically redshifted companion galaxies are to understanding the nature of cosmic redshifts, I must recall that I almost did not have the chance to publish or follow up the implications. It was 1967, and I had just finished The Atlas of Peculiar Galaxies. I had used my staff member’s observing time to study the best examples of companions on the ends of spiral arms, and I submitted a paper, the abstract of which is reproduced above:
It was well understood at that time that the journal in which important papers were published was the Astrophysical Journal. The long-time editor of that journal was Subrahmanyan Chandrasekhar, a theoretician of great renown and generally considered

all the Way Down

91

a tough but fair guardian of its reputation. I don’t know now how I ever could have imagined that he would have been pleased by these interesting new observational results. He returned the paper with a handwritten message scrawled across the top: “This exceeds my experience.”
It took a little while for it to penetrate my stunned senses that he had rejected the paper without ever sending it to a referee. I suddenly felt a cold shudder of apprehension as I realized that my prospects in astronomy were not very bright if I had alienated the editor of the Astrophysical Journal. What to do? First I felt I had to safeguard the observations by getting them published somewhere they could be read and referenced. The only possibility seemed to be the journal that was just starting up in Europe called Astronomy and Astrophysics. With some trepidation, I submitted the paper there. After some anxious weeks the paper came back. A new jolt of panic hit when I saw it had been refereed by another renowned and conservative astronomer, Jan Oort.
Forcing myself to read on I was overjoyed to find that, although he did not agree with the interpretation, he found the observations valuable and interesting and accepted the paper for publication. In the ensuing years I came to know Oort better and found him to be an extraordinarily polite and gracious man. Underneath, however, he had opinions of steel, and apparently would never for a moment entertain a solution which violated the usual assumptions of astronomy. Many years later when he was nearing 90, after a warm dinner at his house, he wrote me a letter urging me to give up my radical ideas and once again participate in the privilege of doing mainstream astronomy. I thanked him and answered him with a quote from my wife: “If you are wrong it doesn’t make any difference, if you are right it is enormously important.”
A most vivid memory I have, however, comes from the time I was sitting next to Oort in the Krakow meeting of the International Astronomical Union. Ambarzumian was chairing the session and Oort leaned over to me and whispered: “You know, Ambarzumian was right about absolutely everything!” Many times since then I have wondered whether, if Oort had said that out loud, and backed it with his enormous

92

Excess Redshifts

influence, the paradigm of astronomy today might not be much different. And I wondered too whether this was not his real, intuitive intelligence slipping for an instant out from behind the secure conformity of accepted dogma. At any rate, although it pained him very much to see an interpretation given which was contrary to his own, it never occurred to him to prevent another genuine observer from speaking or publishing that opinion. To say that this was the ethics of an old-fashioned gentleman is to emphasize that ethics have changed today.
But that did not solve my problem with the Astrophysical Journal. The Director of the Mt. Wilson and Palomar Observatories called me down to his office. To my horror there was a copy of my paper sitting on his desk. Chandrasekhar had sent him a copy of my paper complaining that I had been caught up in a “phantasmagoria” (who could forget that word) and suggested that my Director do something about it. He did. He told me that my appointment would not be renewed next year.
Stunned disbelief and fright was my reaction. My understanding was that, though unwritten, my tenure at the Observatories was permanent. And yet what could I do if it were not? I could only mutter weakly that I would wait for his notification in writing. As the weeks and then months went by with no letter arriving, my terror began to subside and I began to think the crisis had passed. But I felt hunted, and there loomed the question of how I would handle the publishing of future observations.
In the height of the storm, there seemed only one principle to cling to—that was fairness. I knew the observations were good and the interpretation was based on scientific reasoning. The Astrophysical Journal had a responsibility to communicate them to other astronomers. Even though it would exacerbate my position, I decided I must protest to the Editorial Board. Almost a year passed and one day I heard that Chandrasekhar, after long and honorable service, had decided finally to relinquish the onerous burden of the editorship of the Astrophysical Journal. By then I was concentrating on further observational programs and I remember thinking: “…well, it is a hard job and he has been at it a long time, I suppose this had to come sooner or later.”
A few months after that I came down to the Friday afternoon astronomy luncheon at Cal Tech. There was a seat open next to Fred Hoyle at the middle of the long table. I sat down next to him and started chatting happily about new observations. After a while Chandrasekhar, there on a brief unannounced visit, slowly entered the room and proceeded to the only empty seat at the table, directly opposite me. After finishing the subject with Fred I found myself looking directly across at a silent Chandrasekhar. Merely to make polite conversation I remarked: “You must be enjoying the respite from your arduous duties as Editor.”
Suddenly there was one of those complete silences, as all conversation stopped and the whole long table turned to stare directly at us. Chandra rose up a few inches from his chair and said angrily:
“How could I continue to be Editor when people like you complained about me?” I was stricken with embarrassment, but for the first time before or since, managed to come up with an immediate reply:

all the Way Down

93
Fig. 3-31. A galaxy from the ESO Catalogue of Southern Galaxies, ESO 161-IG24. Companion galaxies appear to be attached to the ends of three spiral arms! Detailed spectroscopic observations would be extremely interesting.

“I would hope, in spite of our professional differences, to remain cordial in public.”
The table went back to conversation and we did not speak to each other for the rest of the meal. Come to think of it, we have never had an occasion to speak since then. In fact, these were the only words we spoke to each other in our entire lives.
Of course Chandrasekhar went on to be awarded the Nobel Prize for his work on structure of stellar interiors and allied subjects. About that time I was at an astronomical meeting and attended a lecture he gave. I was amazed that he spent a great part of the lecture talking about his relationship with his erstwhile teacher, Sir Arthur Eddington. In those remarks, Chandrasekhar stressed the emotional hurt that he had received when Eddington had strongly rejected his ideas on the degenerate cores of white dwarf stars. He emphasized what a debilitating effect it had on his outlook for a long time afterward. I was surprised, but I admired him for being able to talk about it publicly. Although at the same time it was sad to realize that he had then turned around and passed on a similar blow to someone else.
A Chance Galaxy
One day I was passing the photographic laboratories at the European Southern Observatory and I saw a pile of photographs they were discarding. I picked out the object shown in Figure 3-31. It turned out to be an object in the ESO Catalogue of Southern Galaxies, ESO 161-IG24. Just a chance galaxy. But it is so eloquent. Three spiral arms with a companion on the end of each arm. And what is more, the longest arm has a series of large knots along it, which look simply like nascent companions. Of course, it would be fun to examine this system with high resolution and spectra. But is

94

Excess Redshifts

it really necessary in a broader sense? The two major companions are obviously not falling in, and from what we know about ejection in many other galaxies, it is just inviting us to fill in the evolutionary links. Someone who knows galaxies will someday identify and observe it. Meanwhile, we can move on to investigate the questions of fundamental physical processes.

Chapter 4

INTRINSIC REDSHIFTS IN STARS!

If we are to believe the previous three chapters, then most extragalactic objects have intrinsic redshifts—ranging from large values for high redshift quasars and continuing right down to small values for low redshift galaxies. It was important to discover that low and medium redshift galaxies also had non-velocity redshifts because it meant that the effect pertained not only to quasars, which could be argued to be exotic and not well understood. Now the phenomenon could also be studied in nearby galaxies having gas, dust, and stars which could be resolved individually—all components about which we thought we knew most of what was important.

The Magellanic Clouds
The two nearest galaxies to us were reported as faintly luminous clouds in the Southern Hemisphere by early explorers. Even with the 74-inch telescope in South Africa in 1955, I was able to measure 10 magnitudes fainter than the brightest stars in the Small Magellanic Cloud (SMC). But it was not until 1980 that the brightest supergiant stars in both Clouds were measured with high spectroscopic dispersion.
Now the Magellanic Clouds are members of the Local Group of galaxies and therefore have an intrinsic redshift relative to the oldest galaxy in the Group, M31. (They also have positive redshifts with respect to our own Milky Way Galaxy, which would mark them in turn as our younger offspring). One could not help wondering whether the gas, dust and stars in these smaller neighbors all shared this same excess redshift—particularly the supergiants which are short-lived and must be, in some sense, younger than the rest of the galaxy. I remember vividly when, long ago, I first checked the companion galaxies to see if they were redshifted with respect to the dominant galaxy. It was with the same sense of not-daring-to-hope that I now approached the necessity of checking the supergiants in the Magellanic Clouds to see if they were, by any chance, redshifted with respect to their own galaxy.

Halton Arp, Seeing Red: Redshifts, Cosmology and Academic Science (Apeiron, Montreal, 1998)

95