Frontiers of Fundamental Physics Frontiers of Fundamental Physics Edited by Michele Barone Nuclear Research Centre Demokritos, Greece and Franco Selleri Universita di Bari Bari, Italy Springer Science+Business Media, LLC Library of Congress Cataloging-In-Publication Data Frontiers of fundaNental physics I edited by Michele Barone and Franco Sellerl. p. CN. ·Proceedlngs of an International conference on Frontiers of fundnentl physics, held September 27-30, 1993, In Oly.pla, Greece"- -T.p. verso. Includes bibliographical references and Index. ISBN 978-1-4613-6093-3 ISBN 978-1-4615-2560-8 (eBook) DOI 10.1007/978-1-4615-2560-8 1. Physics. 2. Astrophysics. 3. GeophysiCS. I. Barone, Michele. II. Sellerl, Franco. QC21.2.F76 1994 500.2--dc20 94-38843 CIP Proceedings of an International Conference on Frontiers of Fundamental Physics. held September 27-30, 1993, in Olympia. Greece © 1994 Springer Science+Business Media New York Originally published by Plenum Press, New York in 1994 All rights reserved No part of this book may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording, or otherwise, without written permission from the Publisher Preface The Olympia conference Frontiers of Fundamental Physics was a gathering of about hundred scientists who carryon their research in conceptually important areas of physical science (they do "fundamental physics"). Most of them were physicists, but also historians and philosophers of science were well represented. An important fraction of the participants could be considered "heretical" because they disagreed with the validity of one or several fundamental assumptions of modern physics. Common to all participants was an excellent scientific level coupled with a remarkable intellectual honesty: we are proud to present to the readers this certainly unique book. Alternative ways of considering fundamental matters should of course be vitally important for the progress of science, unless one wanted to admit that physics at the end of the XXth century has already obtained the final truth, a very unlikely possibility even if one accepted the doubtful idea of the existence of a "final" truth. The merits of the Olympia conference should therefore not be judged a priori in a positive or in a negative way depending on one's refusal or acceptance, respectively, of basic principles of contemporary science, but considered after reading the actual new proposals and evidences there presented. They seem very important to us. The confrontation between different lines of research has accompanied science from its birth. Galileo's scientific ideas were heretical, not only with respect to the dominant religious and political powers of his times, but with respect to the academic establishments of the universities as well: Well known is the example of the astronomy professor who refused to look in the telescope, but many were the centers where the heliocentric ideas were rejected. The great results obtained by Kepler, Newton, and many others, slowly transformed Galileo's heresy into the orthodoxy of modern physical science. Atomism had existed as an idea cultivated by few isolated people for about 2300 years when, at the end of the XIXth century, Ludwig Boltzmann presented his conception of an objectively existing atomic structure of matter. Almost all scientists surrounding him seemed to reject atomism, and the bitter struggle that went around this question probably contributed to the dramatic ending of his life (1906). In the same years, however, Albert Einstein and Jean Perrin obtained an atomistic description of Brownian motion and shortly afterwards atomism was fully accepted in physics, owing also to the discoveries made by Ernst Rutherford and Niels Bohr. In this way the isolated ideas of Boltzmann became the new orthodoxy. v The geophysicist Alfred Wegener was much laughed at for his 1912 proposal that the continents had shifted relatively thousands of km, that the Atlantic Ocean had opened as the Americas split from Africa and Europe, and that all the continents had once been united as a single supercontinent, Pangaea. Only after the confirmations found by Warren Carey in 1954 Wegener's discovery started to be accepted in the scientific community. Today there are so many independent proofs that continents have been united in the past that it seems impossible to doubt it. Here the new frontier has become the conjecture that the Earth radius has considerably increased in the past. In spite of these well known examples science is of course not reducible to an endless confrontation between opposite ideas, since it deals with the material reality surrounding us and uses powerful methods that allow sometimes the scientists to understand true properties of the real world. Therefore part of the orthodoxy can also be considered as valid knowledge. Such are for example the following statements: the Sun is just a star; the Milky Way is our galaxy seen from inside; in outer space there are hundreds of millions of other galaxies; there is a molecular and atomic structure of matter, and a nuclear structure of atoms; there exist subatomic entities called electron, proton, pion, etc. Many other examples of valid knowledge could easily be given. It is a fact however that today's physics seems to contain more than just valid knowledge. In books dealing with astrophysics and cosmology one often finds statements like: "the astronomer Edwin Hubble established beyond all reasonable doubt that the Universe is expanding", but Hubble himself wrote in several different occasions statements like the following one of 1939: "... the results do not establish the expansion as the only possible interpretation of redshifts". Moreover quasars are the objects with the largest observed redshifts, and should therefore be considered at the margins of the visible universe, but many independent pieces of observational evidence indicate that some of them are actually associated with nearby galaxies and that their redshifts cannot therefore be due to recessional motion. A recent amazing discovery is the so called "redshift quantization" phenomenon for spiral galaxies, and this is so difficult to explain within the standard cosmology that most people prefer to forget about it - a predictable reaction of modern scientific thinking confronted with radically new evidence. Important astrophysicists and cosmologists (Hannes Alfven, Halton Arp, Geoffrey Burbidge, Fred Hoyle, Jayant Narlikar, ... ) have repeatedly argued that the observed redshifts of quasars and galaxies could well have an explanation radically different from the standard one based on big bang. In spite of all this the dominant view remains the idea that the only possible explanation of galaxy and quasar redshifts is based on the universal expansion In relativity most people believe that the "luminiferous ether" of the XIXth century has been ruled out by Michelson-type experiments and by the development of the theory of special relativity. The situation is very different however, since vi Poincare and Lorentz were both defenders of the existence of ether, and Einstein himself after 1916 radically modified his previously negative attitude. For example in 1924 he wrote: "According to special relativity, the ether remains still absolute because its influence on the inertia of bodies ... is independent of every kind of physical influence." The minority group of people working today in the foundations of special relativity seems to be almost completely ether-oriented, and there are many proposing a reformulation of the theory along the lines dear to Lorentz: Simon Prokhovnik and the late John Bell are two examples. It has also become clear how such a reformulation should be carried out, after the 1977 realization that the conventional nature of the clock synchronization procedures opens the door to theories which are different from, but physically equivalent to special relativity. Also general relativity has problems of fundamental nature, in particular those connected with the right-hand side of Einstein's field equations, where only the matter stress-energy tensor, but not the field stress-energy tensor, contributes to space-time curvature. This goes against the very fundamental conclusion of special relativity that all forms of energy are completely equivalent, and gives rise to a curious conservation law of rest mass, but not of energy-momentum. A very large number of theoretical physicists seem to be happy with calculations performed strictly within the standard formulation, in spite of the fact that it has been shown that Einstein's field equations do not lead to interactive N-body solutions, if N > 1. General relativity can be considered as a test-particle theory, and as such it explains the three classical tests, but in other respects seems sometimes not to be quite satisfactory. More about this can be found in these proceedings. In our century the interplay between science and ideology has become more important than ever and the hystorians of physics have produced detailed reconstructions of the true scientific/cultural processes leading to the development of what we call "modern physics". From this work evidence has emerged for the existence of common cultural roots with philosophers such as S. Kierkegaard, M. Heidegger, A. Schopenhauer, and W. James. It is therefore not surprising that these philosophers developed ideas similar to some now prevailing in modern physics, in particular concerning the negative attitude toward the possibility of a correct understanding of the objective reality. In fact in quantum physics the standard teaching (after 1927) is that one cannot understand the atomic world in "classical" terms, that is by employing causal space-time descriptions. People active in the foundations of quantum physics believe instead that no good reason for such a pessimistic conclusion has ever been presented, and recall that Einstein, Planck, Schr6dinger and de Broglie could not accept it. A group of participants in Olympia try accordingly to find new space-time models of elementary particles and/ or to develop new rna thema tical tools useful for this task. Bell's theorem states that any theory of the physical world based on the rather natural point of view of local realism must disagree at the empirical level with the vii predictions of quantum mechanics by as much as 42%. Experiments performed in the seventies and early eighties have produced results compatible with the existing quantum theory, but Bell's theorem has actually not been checked due to the introduction in the reasoning of arbitrary (but unavoidable, given the efficiency of the used apparata) additional assumptions. In this way a confusion has been produced between Bell's original inequality and the much stronger inequality violated in those experiments, forgetting that the latter owes the very possibility of being violated by the quantum theoretical predictions to the mentioned additional assumptions. In spite of the fact that Bell's theorem could allow in principle to decide who was right in the Einstein/Bohr debate, we still do not know the answer thirty years after the formulation of the theorem. The confrontation between different points of view goes on, but a strange mutation seems to reduce its effects, since new ideas in fundamental physics find invariably difficulties in being accepted by the majority, no matter how well formulated and important they could be. While the ruling of the majorities is a fundamental feature of every democracy, it certainly does not apply to science where the great steps forward have always been made by isolated individuals. This dogmatic hardening risks today to make the scientific majorities impenetrable to a critical understanding of the foundations of contemporary scientific theories. The existence of such attitudes within modern science has been observed by many physicists and also by the best epistemologists of our century. Thomas Kuhn, for example, wrote about the education of young physicists: "Of course, it is a narrow and rigid education, probably more so than any other except perhaps in orthodox theology." Karl Popper was worried about the poor standards of scientific confrontations and stated: "A very serious situation has arisen. The general antirationalist atmosphere which has become a major menace of our time, and which to combat is the duty of every thinker who cares for the traditions of our civilization, has led to a most seri0us deterioration of the standards of scientific discussion.... But the greatest among contemporary physicists never adopted any such attitude. This holds for Einstein and Schrodinger, and also for Bohr. They never gloried in their formalism, but always remained seekers, only too conscious of the vastness of their ignorance." The understanding of the vastness of our ignorance was generally present in Olympia, but in all fairness we must add that one could also get glimpses of what our science could become in the future: in all cases these were very exciting moments. The choice of Olympia for helding the conference was not casual: this is the place where the Olympic games of ancient times were held for something like 1,200 years. Wars were stopped when the games started and activities included reading of poems, and discussions about science and philosophy. Olympia is not only one of the most beautiful and "interesting spots of the world, but also a positive symbol of the modern civilization. viii The generous efforts of many people have made our conference possible. First of all we wish to thank Attanassios Kanellopoulos for his encouragment and for many useful suggestions. The elected member of the Parliament Crigno Kanellopoulos-Barone has generously helped us in establishing fundamental contacts in Olympia and elsewhere. The constant help of Georges Kanellopoulos has been of tremendous importance for the success of the meeting: we thank him warmly. We are also very grateful to the physics students Rossella Colmayer, Francesco Minerva and Gabriella Pugliese, who formed an efficient and charming secretariat. Our thanks go also to the International Olympic Committee, and to its president Prof. X. Yzezezez, for allowing us to use, free of charge, the wonderful structures of the Olympic Academy where the conference was held. Mr. A Bababab, representative of the Greek government, brought us welcome greetings and encouragment, and Prof. R. Rapetti, president of the Istituto Italiano di Cultura in Athens, stressed the European nature of the conference. The words of Mr. X. Kosmopoulos, major of Olympia, made the participants feel at home in his marvellous town. Last but not least our gratefulness goes to the generous sponsors: the Greek Ministry of Culture, the General Secretary of Research and Technology of the Greek Ministry of Industry, the UniversWI di Bari and, independently, the Physics Department of the Universita di Bari, the National Tourist Organization of Greece, the Commercial Bank of Greece, the Ionian Bank of Greece, the Ellenic Industrial Development Bank S.A, and Glaxo AE.B.E. Without their concrete help the Olympia conference would not have taken place. M. Barone and F. Selleri ix Contents ASTROPHYSICS: ANOMALOUS-REDSHIFTS Empirical Evidence on the Creation of Galaxies and Quasars 1 Halton Arp Periodicity in Extragalactic Redshifts 13 William M. Napier Quasar Spectra: Black Holes or Nonstandard Models? 27 Jack W. Sulentic Configurations and Redshifts of Galaxies 37 Miroslaw Zabierowski Isominkowskian Representation of Cosmological Redshifts and the Internal Red-BIue-Shifts of Quasars ....................................................... 41 Ruggero M. Santilli The Relativistic Electron Pair Theory of Matter and its Implications for Cosmology ................................................................................................... 59 Ernest J. Sternglass Are Quasars Manifesting a de Sitter Redshift? 67 John B. Miller and Thomas E. Miller xi What, if Anything, Is the Anthropic Cosmological Principle Telling Us? 73 Silvio Bergia Large Anomalous Redshifts and Zero-Point Radiation 83 Peter F. Browne Theoretical Basis for a Non-Expanding and Euclidean Universe 89 Thomas B. Andrews Light Propagation in an Expanding Universe 99 Alexandros Paparodopoulos Fornax - The Companion of the Milky Way and the Question of Its Standard Motion ........................................................................................ 105 Miroslaw Zabierowski Cosmological Redshifts and the Law of Corresponding States 107 Victor Clube RELATIVITY: ENERGY AND ETHER Did the Apple Fall? ................................................................................................. 115 Hiiseyin Yilmaz Investigations with Lasers, Atomic Clocks and Computer Calculations of Curved Spacetime and of the Differences between the Gravitation Theories of Yilmaz and of Einstein 125 Carrol O. Alley Gravity Is the Simplest Thing! 139 David F. Roscoe Fourdimensional Elasticity: Is It General Relativity? .................................. 147 Angelo Tartaglia Universality of the Lie-Isotopic Symmetries for Deformed Minkowskian Metrics ........ ...... ....... ...... ..... .... ..... .......... ... .... .... ......... ..... ...... .... ..... ........ ....... 153 Ascar K. Aringazin and K.M. Aringazin xii Hertz's Special Relativity and Physical Reality 163 Constantin I. Mocanu From Relativistic Paradoxes to Absolute Space and Time Physics 171 Horst E. Wilhelm Theories Equivalent to Special Relativity 181 Franco Selleri The Physical Meaning of Albert Einstein's Relativistic Ether Concept.... 193 Ludwik Kostro The Limiting Nature of Light-Velocity as the Causal Factor Underlying Relativity ..................................................................................................... 203 Trevor Morris The Ether Revisited .............................................................................................. 209 Adolphe Martin and C. Roy Keys What Is and What Is Not Essential in Lorentz's Relativity 217 Jan Czemiawski Vacuum Substratum in Electrodynamics and Quantum Mechanics Theory and Experiment ........................................................................... 223 Horst E. Wilhelm The Influence of Ideal ism In 20th Century Science 233 Heather McCouat and Simon Prokhovnik GEOPHYSICS: EXPANDING EARTH Creeds of Physics .................................................................................................... 241 S. Warren Carey Earth Complexity vs. Plate Tectonic Simplicity 257 Giancarlo Scalera An Evolutionary Earth Expansion Hypothesis 275 Stavros T. Tassos xiii Global Models of the Expanding Earth 281 Klaus Vogel An Orogenic Model Consistent with Earth Expansion 287 Carol Strutinski Earth Expansion Requires Increase in Mass 295 John K. Davidson Principles of Plate Movements on the Expanding Earth 301 Jan Koziar The Origin of Granite and Continental Masses in an Expanding Earth 309 Lorence C. Collins The Primordially Hydridic Character of Our Planet and Proving It by Deep Drilling ........................................................................................... 315 C. Warren Hunt Possible Relation between Earth Expansion and Dark Matter 321 Stanislaw Ciechanowicz and Jan Koziar Earth Expansion and the Prediction of Earthquakes and Volcanicism 327 Martin Kokus Tension-Gravitational Model of Island Arcs 335 Jan Koziar and Leszek Jamrozik FIELDS, PARTICLES: SPACE-TIME STRUCTURES Electromagnetic Interactions and Particle Physics 339 Asim O. Barut Isotopic and Genotopic Relativistic Theory 347 Asterios Jannussis and Anna Sotiropoulou A Look at Frontiers of High Energy Physics: From the GeV(10geV) to PeV(1015eV) and Beyond ..................................................................... 359 Michele Barone xiv An Approach to Finite-Size Particles with Spin 369 Bronislaw Sredniawa A New High Energy Scale? 377 Vladimir Kadyshevsky On the Space-Time Structure of the Electron 383 Martin Rivas Physics without Physical Constants 387 Edward Kapuscik The Relation between Information, Time and Space Inferred from Universal Phenomena in Solid-State Physics 393 Gerhard Dorda Quantum-Like Behaviour of Charged Particles in a Magnetic Field and Observation of Discrete Forbidden States in the Classical Mechanical Domain ......................................................................................................... 401 Ram K. Varma Unipolar Induction and Weber's Electrodynamics 409 Andre K. T. Assis and Dario S. Thober Impact of Maxwell's Equation of Displacement Current on Electromagnetic Laws and Comparison of the Maxwellian Waves with Our Model of Dipolic Particles ............ ..... ......... ................ ......... .............. ......................... 415 Lefteris A. Kaliambos Direct Calculation of H and the Complete Self Energy of the Electron from Fluid Models ....................................................................... ...... ..... ............. 423 William M. Honig Interbasis "Sphere-Cylinder" Expansions for the Oscillator in the Three Dimensional Space of Constant Positive Curvature 429 George S. Pogosyan, A.N. Sissakian and S.l. Vinitsky Pancharatnam's Topological Phase in Relation to theDynamical Phase in Polarization Optics ..................................................................................... 437 Susanne Klein, Wolfgang Dultz and Heidrun Schmitzer xv On the Connection between Classical and Quantum Mechanics 443 Andrzej Horzela Discrete Time Realizations of Quantum Mechanics and Their Possible Experimental Tests 449 Carl Wolf Heraclitus' Vision - Schrodinger's Version 459 Fitter Griiff OUANTUM PHYSICS: DUALITY AND LOCALITY Is It Possible to Believe in both Orthodox Quantum Theory and History? 465 Euan J. Squires A New Logic for Quantum Mechanics? 475 Eftichios Bitsakis Dangerous Effects of the Incomprehensibility in Microphysics 485 Jenner Barretto Bastos Filho Classical Interpretation of Quantum Mechanics 493 Vladimir K. Ignatovich Rabi Oscillations Described by de Broglian Probabilities 503 Mirjana Bozic and Dusan Arsenovic A Test of the Complementarity Principle in Single-Photon States of Light 511 Yutaka Mizobuchi and Yoshiyuki Othake Experiments with Entangled Two-Photon States from Type-II Parametric Down Conversion: Evidence for Wave-Particle Unity...................... 519 Carroll O. Alley, T.E. Kiess, A. V. Sergienko and Y.H. Shih Note on Wave-Particle Unity H. Yilmaz Correlation Functions and Einstein Locality 529 Augusto Garuccio and Liberato De Caro xvi Optical Tests of Bell's Inequalities. Closing the Poor Correlation Loophole 537 Susana F. Huelga, Miguel Ferrero and Emilio Santos Atomic Cascade Experiments with Two-Channel Polarizers and Quantum Mechanical Nonlocality ......................................................... 545 Mohammad Ardehali New Tests on Locality and Empty Waves 555 Ramon Risco-Delgado Wave-Particle Duality 561 Marius Borneas Quantum Correlations from a Logical Point of View 565 Nikos A. Tambakis Local Realism and the Crucial Experiment 571 Yoav Ben-Dov The Space of Local Hidden Variables Can Limit Non-Locality And What Next? ...................................................................................... 575 Milan Vinduska How the Quantum of Action Cannot Be a Metric one 583 Constantin Antonopoulos The Ghostly Solution of the QuantumParadoxes and Its Experimental Verification .................................................................................................. 591 Raoul Nakhmanson Index .......................................................................................................................... 597 xvii EMPIRICAL EVIDENCE ON THE CREATION OF GALAXIES AND QUASARS Halton Arp Max-Planck-Institut fur Astrophysik Garching bei Munchen, Germany Simply the arrangement on the sky of extragalactic objects has long shown that the youngest, smallest quasars and compact galaxies have been created recently in the vicinity of older progenitor galaxies. Now high energy observations in X-rays and -y-rays confirm these connections and require the creation of matter as an ongoing process marked by an initially high intrinsic redshift. The nearest superclusters of galaxies show creation along lines in space originating from the central, ejecting galaxy. String theory may be pertinent. The existence of preferred values of redshift (periodicity) rule out, again, an expanding universe. They also imply quantum mechanical effects at the m = 0 creation points of particulate matter. No theory has been advanced, however, which numerically predicts the quantization values. Introduction The Big Bang theory of the universe precludes any scientific observation of creation because the event is so remote in time and space. But even if we could observe this singular event at a distance of 15 bilion light years this age zero universe would supposedly surround us in every direction. That leads to the rather bizarre conclusion that we are, at this moment, "inside" a point that is so small it is dimensionless (the point from which the universe is supposed to have suddenly expanded). Perhaps the conclusion is illogical enough to send us back to what we should have been doing all along - looking at the actual observations. If we do, we find that they all point to the incorrectness of one key assumption in the current theory. That assumption is that extragalactic redshifts measure velocities of expansion. If redshifts are not due to recessional velocities the expansion of the universe and Big FronJiers ofFuruiamenJai Physics. Edited by M. Barone and F. Selleri, Plenum Press, New York, 1994 Bang is wrong and consequently creation must take place throughout the universe in events which can be observationally (hence scientifically) studied. Alignment of Quasars and High Redshift Galaxies Across Low Redshift Galaxies The clear observational pattern that emerges from systematic study of the actual sky is that galaxies occur in groups. Large, dominant galaxies tend to be surrounded by smaller younger galaxies of somewhat higher redshift (c6z~100kms-l). Even younger, more active companions tend to have higher excess redshifts. The youngest, most compact galaxies and quasars tend to be associated with active galaxies in these groups and have the largest excess redshifts (from c6z~100kms-l to c6z~2). Statistically these associations are overwhelmingly significant (see for review Arp 1987). In addition there are numerous instances of interactions or connections between individual low redshift galaxies and high redshift compact galaxies and quasars (see for update Arp 1993). The obvious validity of these observations has not been accepted by influential astronomers because the evidence falsifies the assumption that redshift equals velocity and hence the expanding universe to which most scientists are committed. As would be expected of any valid conclusion, more evidence is continually being discovered which confirms these empirical relationships between objects of widely varying redshifts. Judging from past behavior, the latest evidence will not sway the opinion of those whose interest lies with the status quo. But since the latest evidence deals with high energy X-rays and very high energy ,-rays it is of prime usefulness to those interested in real processes of matter creation in the universe. NGC 4258 One of the most striking new observations is shown in Fig.I. The galaxy is an unusually active one, known to be ejecting hydrogen emission, proto spiral arms and radio material from an excited (Seyfert) nucleus. (van der Kruit, Oort and Mathewson 1972; Arp 1986a; Courtes et al. 1993). A spectacular result emerges from recent observations in X-ray wavelengths. (Pietsch et al. 1994). As Fig.1 shows, the two most conspicuous, point X-ray sources in the field pair exactly across the nucleus of this galaxy which is so well known for ejecting excited material. Any two X-ray sources in this field would only have about one chance in a thousand of accidentally pairing this exactly across the galaxy. But we have to multiply this by the small chance that two such strong X-ray sources would fall so close to the galaxy plus the extraordinary coincidence that the pairing would occur across one of the most striking examples known of an ejecting spiral galaxy. Altogether there is clearly negligible chance that the pair of X-ray sources is not associated with the galaxy. The authors of the new X-ray paper suggest they may be bipolar ejecta from the nucleus of NGC4258. The crowning result, however, is that both components of the X-ray pair are identified with blue stellar objects. One of these has been confirmed as a quasar of z rv .4 (W. Pietsch, private communication) and the other is almost certainly a quasar, probably of comparable redshift. 2 NGC 4258 • 0 01 2', ROSAT-PSP 0 ., • C!. 0 • • 0 "& . 00 ·0 o • ~~ -~ Go 0 . . @• 0 , .. 0 0 . • • • '.5 arcmin t ' I Figure 1. X-ray observations by W. Pietsch et al. (1974) of the active, ejecting, galaxy NGC4258. Conspicuous X-ray sources paired across the minor axis are identified with blue stellar objects, one of which has been confirmed as a quasar with the other being investigated. The upshot of this one observation, by itself, is to confirm unequivocally that high redshift quasars are physically associated with and presumably ejected, from active, low redshift galaxies. This is far from the first example of this kind of association. The first one was discovered among the initial surveys of the brightest radio quasars. 3C273 and M87 Fig.2 shows that the brightest apparent magnitude quasars in the sky, 3C273, and the most active, bright radio galaxy (M87 = 3C274) - these two are aligned almost perfectly across the brightest galaxy in the Virgo Cluster (Arp 1967). The chance of such a configuration being accidental was calculated to be about one in a million. Many observational arguments point to the ejection origin of these two famous active objects from the central galaxy in the Virgo Cluster and, in fact, the origin of the whole cluster from this central point (Arp 1978). The Virgo Cluster is central to the Local Supercluster which is the largest aggregation of galaxies known in our sector of the universe. 3 DEC. (1950) 10" 3C274 3C273 O"~~----~--~----~----~--~~ 4S m 12h30m 14m R.A.{i950) Figure 2. The brightest apparent magnitude quasar in the sky, 3C273 and the brightest jet radio galaxy 3C274 (M87) are aligned exactly across the brightest galaxy in the center of the Virgo Cluster, M49 (from Arp 1967; 1990). (# 134 from Atlas of Peculiar Galaxies = M49) So just the original geometrical configuration on the sky showed 27 years ago that the quasar was a member of the relatively nearby Virgo Cluster despite its much higher redshift (cz = c x 0.16 = 48, OOOkms-1 versus cz ~ 1000kms-1 for the Virgo Cluster). Of course, during the following years all sorts of evidence accumulated to confirm that the quasars actually inhabited the Virgo Cluster. A brief summary of this evidence is as follows: 1) A class of relatively radio bright quasars was shown in 1970 to be strongly associated with bright galaxies in the Local Supercluster of which Virgo is the center (Arp 1970). 2) The brightest quasars in an objective prism survey by X. T. He et al. in 1984 were shown to be associated with the M87 region of the Virgo Cluster (Arp 1986b ). 3) In the Palomar Survey of ultraviolet selected quasars brighter than V", 16.2 mag., J. Sulentic showed in 1988 that these bright quasars were concentrated in the region of the Local Supercluster (Sulentic 1988). 4) Quasars with measured Faraday rotation show effects in the direction of the Virgo Cluster which require some to be in front of cluster (Arp 1988). 4 5) An extremely unusual, low density hydrogen cloud was discovered in the Virgo cluster by R. Giovanelli and M. Haynes in 1989. It lay only 45' distant from 3C273 and was elongated accurately back toward the position of 3C273. As a clinching property the famous nonthermal jet in 3C273 pointed down the length of this extended feature (Arp and Burbidge 1990). Since the cloud had redshift of z = 1248kms-1 it was clearly a member of the Virgo cluster and its association with 3C273 therefore marked the latter as also a member. 6) When Hubble Space Telescope obtained spectra in the far ultraviolet of 3C273 it was found that lower redshift absorption lines were about an order of magnitude more numerous than expected from high redshift quasars in other directions (Weymann 1991). Although the conclusion was avoided, it was obvious that the extra absorption systems were most simply explained as objects in the Virgo cluster with a range of redshifts between that of the large galaxies in the cluster and the redshift of 3C273. 7) Most recent, high resolution images with Hubble Space Telescope (Nature 9 Sept. 1993) lead to an interpretation that the famous jet of 3C273 "... must be viewed in the plane of the sky nearly perpendicular to the line of sight" (Thomson et al. 1993). It is well known that, when placed at its redshift distance the quasar exhibits superluminal motion. The customary model invoked to avoid this difficulty is to have the jet aimed almost exactly at the observer. If this geometry is no longer possible then the only way to escape faster than light motion is to significantly decrease the conventional distance of 3C273. Virgo Cluster ROSAT PSPC OSQ lCl7Q ---'___ --; lC 27) Figure 3. Low surface brightness X-rays connect M49 to M87 in the north and 3C273 to the south. Upper integration from ROSAT Sky Survey by Bohringer et al. 1994, lower integration by Arp from same survey. 5 But now the German X-ray satellite, ROSAT, has been observing famous objects in very high energy bands and startling results have appeared. One result is the pair of quasars across NGC4258 as just described. Another result, partially in press, is shown below in Fig.3 (Bohringer, et al. 1994). A glance at the figure shows that the previously known pairing of active objects across the central galaxy in the Virgo Cluster is now confirmed by the new observation of high energy X-rays. An actual continuous path of X-rays now connects M49 northward to M87 and southward to 3C273. Southward, in the direction of 3C273 the trail of X-rays leads to another quasar of z = .334 and then to an active galaxy of cz = 2075kms- 1 (3C270) and finally, in a special analysis of an area extended further south by H. Arp (paper to be submitted), the X-ray trail leads into 3C273 where it appears to end. The appears to be the "smoking gun" where the smoke leads all the way from the active gun to the bullet which it has ejected. It is difficult to imagine what further proof one should hold out for. The Bright Apparent Magnitude, Active Quasar 3C279. Further south from 3C273 is a quasar which, although moderately faint in appearance now, was much brighter only about 40 years ago. At that time it was comparable with the brightest quasar in the sky, 3C273. Since for a long time it has been clear that 3C273 was a member of the Virgo Cluster, it was highly probable that the violently variable 3C279, falling very close in the sky to 3C273, was also a member. Now confirmation of this has recently been obtained from observations at even higher energy wavelengths, namely ,-rays. (The X-rays we have discussed are in the range of photon energy from 0.1 to 2.0 keV whereas the ,-ray observations shown below in Figs. 4 and 5 are in the range 0.7 to 20,000 Mev!) Fig.4 shows observations published by a team of observers in the 0.7-30 Mev range of ,-rays (Hermsen et al. 1993). These COMTEL observations in Fig.4 were then later confirmed by the entirely independent EGRET observations in the higher Mev range shown in Fig.5. The startling aspect of the publication of these results was that despite the huge team of scientists reporting the results none ventured to mention the extraordinarily important fact that the two quasars, 3C273 and 3C279 were linked together by a connection of ,-rays. The highest energy EGRET results were published in the form of a color picture in Sky and Telescope (Dec. 1992 p. 634). The strong emission from 3C279 was clearly extended to the northwest and it must have been known that it terminated on the position of 3C273. Yet the position of 3C273 was not plotted on the picture nor any mention made of it in the text. The intensity isophotes Fig.5 shown here were simply estimated and traced from that color Sky and Telescope picture by the present author and the position of 3C273 and 3C279 indicated here by + signs. Although this situation has been discussed in meetings and privately in 1993, to date the further ,-ray observations of this crucial pair have not been released. In Fig.6 the X-ray observations of the Virgo Cluster have been plotted to the same scale as the ,-ray observations of 3C273 and 3C279. The extraordinary result is that the major active galaxies in the Virgo Cluster lie along an X-ray delineated extension which passes through the largest galaxy, M49 and extends southward to 6 Figure 4 Observations of 3C273 and 3C279 in low energy ,-rays (Hermsen et al. 1992) from COMPTEL instrument aboard the Gamma Ray Observatory (GRO) 3C 279 z=.54 Figure 5 Observations of 3C279 by high energy, EGRET instrument aboard GRO (I-rays 10 to 104 Mev). Isophotes of picture in Sky & Telescope (Dec. 1992) have been copied by present author who has added positions of 3C273 and 3C279 as crosses to show that these independent observations confirm the connection in ,-rays found in the observations by COMPTEL shown in Fig.4. 7 3C273. The X-rays are high energy ('" 1 - 2keV) and as they approach 3C273 the photons become harder until 3C273 is conspicuous in lower energy {'-rays. The final part of the connection to 3C279 is only in high energy {'-rays and 3C279 itself is most conspicuous in the highest of all observed energy {'-rays. VIrVD Cluft.. ROSA' PSf'C Figure 6. Plot of X-ray emitting material in Virgo Cluster which ends on the quasar 3C273. The higher energy {'-rays continue on to 3C279 and are shown by approximate isophotes. The connecting material appears to rise in energy toward the highest energy quasar, 3C279 (z = .538). Trying to Understand the Observations The first order result is to confirm decisively all the previous evidence that objects of widely disparate redshifts are physically grouped together in the same assocations. Further it is confirmed that the most compact, and hence youngest, objects such as quasars have the highest intrinsic (non-velocity) redshifts. Empirically this requires the younger age to be related to the cause of the intrinsic redshift. Fortunately now there is known a solution to the field equations of general relativity which is more general than the traditional, Friedmann, expanding universe solutions. The more general solution allows creation of matter at any epoch in the universe and since the matter is created with initially low particle masses, the newly born matter has initially high intrinsic redshift which declines as it ages (Narlikar and Arp 1993). This theoretical interpretation accounts for the numerous discrepant redshift 8 observations that have accumulated over the past 28 years (83 years if one wishes to count the unexplained systematic redshifts of young stars, the so called K effect). In particular, the discoveries reported here of intense emission of very high energy x-rays and ,-rays from quasars linked to nearby galaxies shows especially clearly that the higher redshift of these objects is connected with their extreme youth. The point is that the emission is supposed to result from acceleration processes arising from travel of charged particles through magnetic fields (synchrotron radiation). But even for the X-ray wavelengths the decay time is of the order of only 50 yrs. (Arp 1994) and for ,-rays, correspondingly shorter. This marks the higher redshift objects as characteristically in a young, active stage where they are intermittently injecting high energy particles. But the shocking surprise is that the low density connections between these young objects are emitting such short lived radiation. Until now the working hypothesis has been that the creation process takes place in the active nuclei of older galaxies. The new matter in compact form is then ejected in opposite directions in the form of high redshift quasars which evolve, as they age, into only moderate excess redshift companion galaxies. The optical connections that are occasionally observed between the older galaxies and the higher redshift companions have naturally been supposed to consist of gas, dust and or stars from the older galaxy that have been entrained during the ejection process. But now we see many more connections consisting of very high energy, short lived radiation. The only possible suggestion would seem to be that very small "retarded cores" were also thrown out with the quasar and that the quasar has left a sparkling trail of rapidly decaying high energy radiation. There are some difficulties with this model, however, which suggest the consideration of some fascinating alternative possibilities. The difficulties are: 1) The lifetime of the high energy radiation is so short that it would seem difficult to sustain the emission of the connection even for the relatively short lifetime of the ejected quasar. This radiation would have to live for at least 106 - 107 years in what appears to be a low density environment. 2) Even with low ejection speeds some of the lower intrinsic redshift ejects should show observable blue shift and red shift differences as they are ejected toward and away from us. This situation has not been ruled out by the observations but for a long time it has been estimated to be an uncomfortable restraint. 3) Although there is abundant evidence for secondary and even tertiary ejection coming off at arbitrary angles to the original ejection lines, the development of great clusters like Virgo and Fornax seems to be in an appreciably broad filament stretching great distances and drifting somewhat irregularly from a straight line. All this suggests a modification of the ejection hypothesis based on reconsideration of the assumption that creation of matter takes place only in point locations in space. Dislocations in spacetime along lines in space which enable the emergence of new matter would not seem to be forbidden and could possibly explain better the newer observations. (This immediately suggests string theory although that theory has not been developed to the extent that it could make predictions of actual events in the extragalactic realm.) The possible amelioration of the aforementioned difficulties which such a "white line" theory could offer are enumerated below: 9 1a) If matter wells up at one point in such a "fault line" in space this could represent the original galaxy. Later emergences, perhaps due to a creation signal from the original galaxy, could produce secondary creation along this same line. The useful point would be that smaller upwelling over an extended period all along the line could possibly account for the currently observed high energy connections between high and low redshift objects. 2a) Since the creations take place from preexisting locations in space there need be no velocities of transport from the original galaxy nuclei and the blue and redshifts from ejection velocities could be avoided (This latter is particularly important in the matter of quantization of redshifts which would place limits of lsim20kms-1 on true translational velocities of galaxies in space). 3a) If secondary creation lines, in analogy with strings, move through space - where ever they intersect the original creation line may promote creation nodes. If later nodes produced younger quasars and compact galaxies, the ejection lines from these secondary objects would be situated at arbitrary angles to the original creation line as observed. This interpretation suggests that jets represent material under pressure guided out of active nuclei by creation lines. Quantization of Redshifts. The one problem that seems to present unresolvable contradictions at this time are the observed quantization of redshifts. Evidence has been available for a long time which establishes that the whole redshift plane is quantized - the quasars in large steps, the galaxies in smaller (Arp 1993). Recently the smallest quantization steps of 37.5 kms- 1 seen by William Napier in the most accurate HI redshifts have become overwhelmingly statistically significant (Napier 1993). It is tempting to connect this quantization with periodicity in the creation process. Since the matter is created with zero mass it transitions from a a quantum mechanical realm where discretization is expected. But if they are not all at the same distance, any intrinsic galaxy redshift would be smeared out by continuously changing lookback times if the distribution of galaxies were continuously spread throughout space. This is presently what I would consider the most difficult unsolved problem in the subject of galaxies and galaxy creation. Summary Comments The observations push us irresistably toward a certain empirical picture of the creation of galaxies and quasars. This in turn opens exciting opportunities for theory. The creation processes of matter are no longer some kind of obscure miracle but we can actually measure the state of the matter from its radiation property at various stages in its evolution. In order to make progress, however, researchers must give up the arbitrary assumption that particle masses are constant in time. When the general case, m = m(x, t) is taken as a starting point the general solution of the Einstein field equation corresponds very well to the observed phenomena. The general connection between age and redshift becomes natural and we can hope to trace the materialization of matter from the quantum mechanical field (or material vacuum) 10 to its better known state in the form of large galaxies. The problem of ultimate destiny of matter in these old galaxies lies untouched. The prediction of observed quantization of redshifts as a function of fundamental cosmic parameters forms a formidable challenge. But as a first step, before the vast majority of observers and researchers can undertake anything meaningful, they must admit the zero order result that extragalactic redshifts are not due to velocities. The empirical evidence on this point was already overwhelming and the new observation in high energy x-rays and ,),-rays now render the evidence completely inescapable. The vast observational facilities, exponentiating publication and well funded theoretical schools will continue to produce misinformation until the crucial issue of the empirical disproof of the redshift assumption is faced. References Arp, H. 1967, Ap.J. 148: 32l. Arp, H. 1970, A.J. 75: l. Arp, H. 1978, Problems of Physics & Evolution ofthe Universe, Acad. Sci., Armenian SSSR, Yerevan 1978. p.65. Arp, H. 1986a, IEEE Transactions on Plasma Science 14: 748. Arp, H. 1986b, Astrophys. Astr. (India) 7: 77. Arp, H. 1988, Astrophys. Lett. A 129: 135. Arp, H. 1994, "ROSAT Survey of an Area 10 Degrees Square around the Active Radio Galaxy Cen A", A&A, in press. Arp, H. and Burbidge, G. 1990, Ap.J. 353: Ll Arp, H. 1987, "Quasars, Redshifts and Controversies" Interstellar Media, Berkeley. Arp, H. 1990, Astronomy Now 4: 43. Arp, H. 1993, "Progress in New Cosmologies: Beyond the Big Bang" ed. H. Arp, C.R. Keys, K. Rudnicki, Plenum Press, New York p. 1-28. Courtes, G. Petit H., Hua, C.T. et al., 1993 A&A 268: 419. Hermsen, W. and 25 collaborators, 1992, A&A Suppl. in press. Napier, W. 1993, Progress in New Cosmologies: Beyond the Big Bang, ed. H. Arp, C.R Keys, K. Rudnicki, Plenum Press, New York. Pietsch, W., Vogler, A., Khabaka, P., Jain, A. and Klein, K. 1994, A&A , in press. Thomson, RC., Mackay, C.D. and Wright, A.E. 1993, Nature 365: 133. van der Kruit, P.C., Oort, J.H., Mathewson, D.S. 1972, Astron. Astrophys. 21, 169. Weymann, RJ. 1991 "The First Year of HST Observations STCI", Baltimore May 1991, p.58. 11 PERIODICITY IN EXTRAGALACTIC REDSHIFTS W.M. Napier Royal Observatory Blackford Hill Edinburgh EH9 3HJ Scotland, U.K. ABSTRACT. Claims that the redshifts of galaxies are quantized at intervals of ~24, ~36 or ~72 km s-1 are being subjected to rigorous statistical scrutiny using new, accurate redshift data. The results of this enquiry to date are reviewed. The presence of a global galactocentric periodicity ~ 37.5 ± 0.2 kms- 1 is confirmed at a high confidence level. A strong redshift periodicity of ~ 71.1 ± 1.3 kms- 1 also exists amongst the galaxies of the outer regions of the Virgo cluster. INTRODUCTION The expression 'fool's experiment' appears to have been coined by Charles Darwin to describe the investigation of a hypothesis which no sensible individual would regard as worth testing; Darwin himself often undertook such enquiries. One imagines that, for most astronomers, the testing of 'redshift quantization' belongs firmly to this category. In essence, the hypothesis developed originally by Dr. Tifft and colleagues (e.g. Tifft 1977, 1980, 1993; Tifft & Cocke 1984) is that the redshifts of galaxies tend to occur in multiples of "'24, ",36 or ",72 km S-1 , the latter periodicity being local (applying to the redshifts of binaries or clusters of galaxies), the former two being global (depending on morphological type, and applying to the galactic redshifts after subtraction of the component due to the solar motion around the centre of the Galaxy). A clear verification of redshift quantization would have far-reaching consequences. In cosmology, the derivation of virial masses, and even the existence of dark matter, would be thrown in doubt; and in astrophysics, there would be a distinct shift of balance in the debate over the discordant redshifts claimed by Arp (this volume). In fact, is not clear that current cosmological and astrophysical paradigms are capable of accommodating the phenomenon. Evidently, only clearly derived, unambiguous and strong results will suffice if the phenomenon is to be taken seriously. However, the Tifft quantization studies, which have relied heavily on histogram binning techniques, have raised questions about the a posteriori selection of binning intervals (cf. Cocke & Tifft 1991, Schneider & Salpeter 1992) and the criteria for selection of binary galaxies; while in the case of global periodicity, a signal is in essence maximized through va,rying three parameters (the three components of solar motion), and before statements can be made about the statistical significance of the claimed periodicities, the effects of this freedom have to be assessed. Further, it is not always clear to the reader why one sample of galaxies rather than another has been chosen, Frontiers ofFundamental Physics, Edited by M. Barone and F. Selleri, Plenwn Press. New York, 1994 13 and the sceptical reader may be left wondering whether negative results have gone unreported. Finally, while it is reasonable to expect an initial hypothesis to be modified with the accumulation of new or better data, several such modifications have occurred, raising the question of which version one is trying to test. For example, in its most recent manifestation it is claimed that periodicities occur over a wide range, from 2.66 km s-l upwards (but with peak power at ",36.5 km S-l: Tifft, peTS. comm.). These issues suggest that there is scope for a fresh approach to the quantization issue. In recent years there has been a great increase in the number of accurately measured HI profiles of galaxies; as a result, there is now a sufficient body of new data for the existence of the claimed periodicities to be settled one way or another. A colleague (Dr B.N.G. Guthrie) and I therefore embarked on a study of the quantization issue a few years ago. The philosophy adopted was to apply an objective, rigorous scrutiny to new and unbiased data, with the intention of publishing the results whether they turned out to be positive or negative. The current state of this project is summarised in the present paper: it is already clear that extragalactic redshifts are indeed strongly quantized along the lines claimed by Tifft and others. Methodology A hypothesis, once set up, may be tested against new, independent and unbiased data by asking whether they confirm a prediction unique to it. The need for lack of bias requires that any selection of the new data from a larger dataset should be carried out with prescribed, simple rules which will not affect the outcome of the enquiry. If it turns out that some modification of the original hypothesis gives a better fit to the new data, then the 'improved' hypothesis should be put to the test against further data, and so on: the protocol thus requires a clear alternation between 'playing hunches' and verifying them. Technique The statistic we have generally used in the study is I=2R2 /N, N the number of galaxies and R the length of the resultant vector in the Argand diagram when the data are wrapped round a drum of circumference P and each assigned a unit vector. A power spectrum is a plot of I against frequency l/P. For a uniform, random distribution of independent redshifts, and neglecting edge effects, the I-distribution has a mean value I = 2, and the probability of exceeding some value 10 by chance is p(I2: 10 ) = exp( -10 /2). This formula becomes inaccurate for extreme values of I, and the statistic is also biased and inconsistent. These problems may be circumvented by comparing the signal strength obtained for the real data with those obtained from large numbers of trials in which suitably constructed synthetic data are analyzed in identical fashion. A full discussion of the technicalities is given elsewhere (Guthrie & Napier, in preparation). Hypothesis Tifft & Cocke (1984) - TC hereinafter -claimed to observe periodicities of ",24.2 and ",36.3 km S-l in the redshift distributions of spiral galaxies with narrow and wide HI profiles respecti vely. These periodicities were global, applying to galaxies distributed over the celestial sphere, and galactocentric, emerging only when the redshift component due to the Sun's motion around the centre of the Galaxy was subtracted from each heliocentric redshift. The differential redshifts in binary galaxies (Tifft 1980) and the Coma Cluster (Tifft 1977) were said to be quantized at intervals of ",72 km S-l . 14 The subsequent modifications of the above basic hypothesis have added to, rather than replaced, the above claims. A new dataset should therefore still show the above periodicities and we do not require to discuss the refinements in testing the above. Samples In our study so far, two spiral galaxy samples have been examined and are discussed here. The first comprises the nearby galaxies with the most accurately measured redshifts out to roughly the edge of the Local Supercluster; the second comprises the galaxies in the Virgo Cluster, which happens to be the nearest rich cluster of galaxies. We can see no bias in these choices of sample. Evidence for redshift periodicity is found in both of them. THE LOCAL SUPERCLUSTER Nearby Galaxies Guthrie & Napier (1991) first tested the global periodicity hypothesis by examining galaxies within 1000 km S-1 of the Galactic centre. The database employed was a recent catalogue of 6439 extragalactic redshifts compiled by Bottinelli et al. (1990). In the spirit of keeping the selection criteria simple and unbiased, spiral galaxies were culled from the dataset according to the following rules: (i) the quoted accuracies were (J" ::;4 km s-1; (ii) galaxies used by TC in formulating the hypothesis under test were excluded; and (iii) galaxies within 12° of M87 were excluded. The latter restriction applied because such galaxies might belong to the Virgo Cluster, which was the subject of a separate enquiry. By hypothesis, the global periodicity is galactocentric; thus from each observed (heliocentric) redshift one must subtract Ve cos X corresponding to the motion Ve of the Sun around the Galactic centre, X the elongation of the galaxy from the solar apex. According to TC, the solar vector yielding the periodicities was (Ve = 233.6 km S-1 Ie = 98.6°, be = 0.2°), and we first carried out a power spectrum analysis on the redshifts corrected for this vector. A prominent peak was found at a period P=37.1 km S-1 , within one of the ranges 35-37.5 km S-1 then under test. No evidence was found for the 24.2 km s-1 peak claimed by TC for narrow-line galaxies, but then only two galaxies in the list had narrow line profiles. The peak had a value 1=18.1 which, for a white noise distribution, has a singletrial probability,...., 10-4 according to the exponential formula. About two independent trials were involved in searching within 35-37.5 km S-1, while a signal in the range ,....,24 km s-1 (and perhaps ,....,72 km s-1 , although not strictly part of the hypothesis), would no doubt have been regarded as significant. The signal was therefore real at a confidence level C,....,0.999 according to the formula. An independent assessment of C was made by generating sets of 89 synthetic redshift data and determining the Imax-distribution in the range 35-37.5 km s-1. For the synthetic redshifts, the positions of the gaJaxies over the sky were preserved. The redshift data were generated by adding, to each measured redshift, the correction for the TC solar vector and then a random displacement in the range 0-8V km S-l , where 8V was small compared with the range of the redshifts and the likely dispersion within any groups and associations within the dataset. Thus the synthetic data were created by applying a 'haze' of width::; 8V to the real data, sufficient to obliterate any periodicity in the range under test but too small to have any other effect. Any significant difference between the real and the synthetic data, could thus only be due to periodicity in the former. 15 For each of 8V =80, 60, 40 and 20 1011 S-1 , 3000 sets of 89 data were constructed and their power spectra obtained. The distribution did not change appreciably until 8V =20 km s-1 , corresponding to synthetic redshifts within ±10 km S-1 of the real ones. Typically, one in a thousand trials yielded I-values of 18.1 or higher. Allowing for the two or three ranges under test, the periodicity hypothesis was thus again preferred over the null one (random distribution) at a. confidence level C~0.997 or 0.998, a result of high statistical significance. However, the motion of the Sun around the centre of the Galaxy is known with limited accuracy, and the (Vr,),P) found by TC was based on only 40 broad-lined galaxies. It therefore remained possible that the periodicity would emerge more strongly for some other solar vector in the neighbourhood of the TC solution. Guthrie & Napier (1991) therefore used published estimates of the motion of the solar neighbourhood around the Galactic centre, taking account of the solar motion relative to the neighbourhood, a probable expansion, and the ullcertainty introduced by warping of the Galactic disc, to obtain a solar vector (V0 = 2:33 ± 7 km S-1 10 = 93 ± 10°, b0 = 2 ± 10°) - the errors cannot be taken too literally. Power spectra were obtained by varying the solar vector over a wide volume of V (oj-space, 60° by 60° in longitude and latitude, and 130 km s-1 in Vr,), which adequately encompassed the error box of the solar motion. For each V 0 a set of corrected redshifts was obtained and analyzed. Several high peaks were found, the two highest being Imax =29.2 for a periodicity P=37.2 km S-1 and Imax=28.0 for a periodicity P=37.5 kms-1. The corresponding vectors were (228 kms-1 , 99°, _3°) and (212 kms- 1, 94°, -13°). Within the errors, these are reasonably close to the solar motion around the Galactic centre, but the speeds are significantly lower than estimates of the solar speed with respect to the Local Group, which lie in the range 250:S V;v :S :310 km S-1 . The significance of the peaks was again assessed by comparison with closely similar synthetic data treated identically to the real data, and for each peak the periodicity hypothesis was preferred over the null one at a confidence level ~0.999. One further test was applied to this dataset: if the apparent periodicity was a statistical artefact, it would not in general vary with the accuracy of the data. Trials on synthetic data, on the other hand, revealed that the measured strength of the signal is highly sensitive to the redshift dispersion (Fig. 7). The resulting 89 galaxies conveniently divided up into 40 with 0'=2 or 3 km S-1 , and 49 with a =4 km S-I. For each of the two solar vectors above, trials were carried out in which 40 redshifts were randomly extracted from the 89 and I-values computed. Twenty thousand such trials were conducted, and the periodic signal was found to be significantly concentrated in the more accurat.e data, this conclusion having a confidence level 0.93 for the 228 km S-1 peak and 0.984 for the 212 km S-1 peak. Combining these factors, we concluded (Guthrie & Napier 1991) that the field galaxies within 1000 km S-1 of the Sun have a redshift periodicity of ~ 37.5±0.3 km S-I. The probability that the periodicity occurred by chance was found to be 3 x 10-6 :S p:S 3 X 10-4 . Extension of the Sample An unexpected consequence of the study out to 1000 km S-1 was that the periodicity emerged with respect to the Sun's local galactocentric motion: the Galaxy's motion within the Local Group, or its iufall towards the Virgo Cluster, did not appear to be relevant. The phenomenon was thus nucleus to nucleus between galaxies, irrespective of large scale motions in the field. If this continued to hold out to greater distances, then the signal should appear with increasing strength as the search volume around 16 b0 00 I-- • • '\ 292 • 258 W 223 206 217 • • I I 240 0 1200 10 Figure 1. The ten highest peaks out of rv 106 in a whole-sky search (140::; V(0) ::;360 km S-1 ), over :W::; P ::;200 kIll S-I . • = 24 ± 3 km 8-1 , * = :37.5 ± 0.2 kIll S-I. The formal error box of the solar galactocentric motion is shown. the Galaxy is increased. We have therefore extended our analysis out to 2600 km S-1 , the edge of the Local Supercluster. There is no immediate reason to suppose that the periodicity should be confined to the LSC but the cut-off was convenient as one is running out of sufficiently accurate data beyond there. The criteria for selection of galaxies from the Bottinelli et aJ. dataset were as before (essentia.lly, all accurate redshifts excluding those which TC had used). Two Virgo-like clusters (UMA and Fornax) are now incorporated in this enhanced volume, but only one eligible galaxy was found in them: for the sake of consistency, the Virgo Cluster having been excluded from the study, it too was excluded. The sample was now extended from 89 to 247 galaxies, and those with measured redshifts of extreme accuracy (0"=2 or 3 km S-1 ) increased from 40 to 97. For practical reasons we concentrated on analyzing the 97 highly accurate redshifts. In this extended analysis we varied the solar vector over the whole celestial sphere and over 140::; '/;'1 ::;:360 kIll S-1 in speed, stepping in 2 or 3° intervals and in units of .5 km s-l. For each pixel in this box, a power spectrum was constructed over the range 20--200 km S-1 and the highest peak recorded, irrespective of its period. About a million power spectra were generated in this search. The ten highest peaks are shown in Fig. 1. Five of the ten have P=:37.5 km S-1 (Fig. 2), and three of them lie within or very close to the error box of the solar motion. Spearman ranking of the departures of individual reclshifts from the periodicity show that all ten peaks are correlated, the three highest strongly so: thus a single underlying phenomenon is being detected. This whole-sky search confirmed that V ('.l-space is not filled with all sorts of 'periodicities' in all sorts of directions, and that indeed the only outstanding phenomenon is the TC one. The significa.nce of these extremely high peaks (Irv39) was assessed by constructing synthetic data as before. ThuOi each set of 97 artificial data was analyzed by varying the solar motion within a box of side :300 by :wo by 60 km s-1 centred on the galactocentric solar vector, and searching over :lO-:~9 km 8-1 . The highest peak out of tJw rv 104 power spectra so constructed was recorded, and the procedure repeated for ten thousand sets of artificial data. The distribution of high peaks in the 104 search volumes was then 17 30 200 km/s 20 Figure 2. Power spectrum (1 vs frequency) associated with the solar vector (V0 ,10, b0 ) = 217 km/s,9So,--12"). The high peak occurs at 37.S kms- l . compared with that obtained for the true data over the same search volume. Note that this procedure, involving as it does the overall power distribution, avoids extreme value statistics of any sort. None of the 104 artificial datasets had statistical behaviour at all close to that of the real data. The real data, for example, threw up 2S peaks with 1 > 20 (n2o=2S) and 12 with 1 > 2S within the search volume, whereas none of the artificial data did. Extrapolation shows that, roughly, the real redshifts differ from the random ones at about. the million to one level (Fig. 3). Since periodicity is the only phenomenon which may be obliterated by the randomization procedure, it follows that the 97 galaxies possess a redshift periodicity of 37.S km S-1 at about this confidence level. A close examination of individual HI profiles revealed that a few of them had slightly asymmetric profiles. as might arise from foreground contamination. An objective criterion was used to reject 16 possibly contaminated galaxies from the list, and 24 TC galaxies which satisfied the other criteria were added. The resulting list of 103 galaxy redshifts constitutes the most accurate sample we currently have for the Local Supercluster. Its optimized power spectrum is shown in Fig. 4: I",S2 for a periodicity 37.S km S-I. A histogram of the red shift differences for this sample is shown in Fig. S: the periodicity is strong and coherent, with no sign of drift, from centre to edge of the LSC. We have not used this extraordinarily high peak to attempt a probability assessment; rather, it serves to confirm that the phenomenon in question is indeed a redshift periodicity. Statistical Behaviour The robustness of the result was tested in various ways. (a) If the periodicity arose from some obscure statistical artefact, then it might be expected to behave erratically with respect to sample size, accuracy of data and magnitude of the optimum solar vector. Fig. 6 reveals that, for a fixed solar vector (217 kIns- l , 9So, -12°), the signal strength increases linearly with N, as expected theoretically for a real signal. The observed slope is consistent with a true dispersion (J" ",8 kms- l . As can be seen from Fig. 7, this dispersion is rather critical for the detection of periodicity when the sample size is N",100, and this may account for the 18 -2 log P -8 L -________~_________L_ _ _ _ _ _ _ _~--------~~------~ o 10 15 20 2S n 20 Figure ;~. Probability that a set of randomized redshifts, constructed and analyzed as described in the text, would yield more than n20 spectral peaks. difference in behaviour between the galaxies with a formal a ~3 kms- 1 and those with a =4 km S-l, since unknown systemic errors of order several km S-l probably exist in these measurements. In this larger sample too, the signal strength was found to concentrate strongly in the best data, at a significance level ",0.998. (b) If, instead of holding the solar vector fixed, the optimum solar vector is derived as a function of sample size, the result shown in Table 1 is obtained. The period holds steady to within ±0.15 kms- 1 for VI') varying by only ±2 kms- 1 in speed and ±1° in direction, as the sample doubles in size from "-'50 to ",100 redshifts. This is a remarkable degree of stability, difficult to reconcile with a statistical fluke or artefact. Trials on sets of random data with in--bnilt periodicity revealed that, for a = 8 km s-1 , the r.m.s. dispersions expected are 0.2 km 8- 1 in derived period, 3 kIn 8-1 in V(.) and 1.2') in direction. (c) The inclusion of the Virgo galaxies, or the arbitrary exclusion of 15 redshift calibrators from the list (Baiesi-Pillastrini & Palumbo 1986) made little difference to the result. Thus the signal strength is robust to modest changes in the dataset and cannot be attributed to a particularly favorable choice of sample. Table 1, Optimized parameters as a function of sample size, The solution holds steady to .6.V := ±2 kms- 1 , .6.0:= ±lo atld .6.P:= ±O,15 kll1s- 1 , throughout the LSC, \'ma..r N V" I,;, b{.) P Imax 1000 ,51 215 9:3 -13 37.8 :30 1400 72 213 94 -1:3 37.7 :31 1800 86 215 94 -13 37.7 ;36 2600 97 217 95 -12 37.5 38 19 Period 200 100 70 50 40 30 25 20 40 20 0.02 0.04 Frequency Figure 4. Power spectrum of the 103 most accurate, uncontaminated redshifts corrected for the optimum solar vector shown. 20 VI '- IU 500 CJ.. "0 '- (lJ ..0 E ::J :z ':::MAJ::::: 500 1000 ~~y.i1\,1~\f.\)W\ 1'\~II \~ r. l: V: S: 2: . 1000 1500 ,,~MMlliMLM~o lIT~06n ~ An\~ : I'0 2000 2500 .6.(Z (orr. (km S-l) Figure 5. Two-point correlation function corresponding to the redshifts and optimum solar vector employed in the previous figure. Vertical dashed lines represent the best-fit periodicity, which seems to hold over the whole of the Local Supercluster. 21 40 30 20 40 60 80 100 N Figure 6. Signal strength I as a function of numbers N of galaxies. The dots are for galaxies out to ,500, 1000, 1500, 2000 and 2500 km S-1. The stra.ight line represents the mean behaviour of I(N) for an assumed true dispersion of 0'=8 km S-1 about P=37.,5 km S-1 . 22 10 n(I) o 40 80 120 Figure 7. Power distributions n(l) obtained for synthetic datasets each con- taining 97 redshifts distributed with periodicity 37.5 kms-1 and dispersions (left to right) a = 32, 12, 8 and 6 kms- 1 respectively. 23 (d) The data were also divided by morphological type, radio telescope employed and celestial position; no correlation was found with any of these. (e) Finally, a whole-sky search using synthetic data with 1'=37. 5 kms- l and 0"=8 km S-1 was conducted. The behaviour was found to have the same general characteristics as shown in Fig. l: a few peaks with the inbuilt periodicity clustered around the galactocentric solar vector, and a scattering of peaks with fractional periods over the celestial sphere and with various V",. Groups and Associations About half the galaxies in the sample of 97 belonged to loose groups or associations (Fouque et al. 1992) containing a few bright galaxies (these groupings are preserved in the Monte Carlo simulations). The data were divided into two appropriate sets to explore whether the periodicity concentrated in either the field or group galaxies. The full sample of 247 spirals was used, enhanced to 261 by the addition of a few galaxies previously used by TC. Correlation analysis revealed a strong tendency for those galaxies which belonged to groups or clusters to possess the most accurately determined redshifts. The question arises whether the periodicity truly exists in clusters, or is simply detected preferentially there because cluster galaxies have been more accurately measured. A correlation analysis supports the latter at a confidence level ~0.96. There are 9 doubles, 6 triplets, 3 quadruplets and a quintuplet in the sample of 97, yielding 5.5 local differential redshifts within these small groups, of which 34 are independent. The differential heliocentric redshifts are plotted in Fig. 8a, and the galactocentric ones in Fig. 8h. Becansp of the small angular extent of the groups the galactocentric correction is now second order. In effect the large number of trials involved in varying V (') are replaced by a single trial, and so in effect the differential redshifts yield a parameter-free test of periodicity. It is clearly present, power spectrum analysis and comparisOll with Monte Carlo trials yielding a confidence level C 2: 0.9999. a b o dV km/s 310 I I I I I II I Figure 8. Histograms of differential redshifts dV for the 53 galaxies linked by group membership. (a) heliocentric redshifts; (b) redshifts corrected for V(') = (216 kms- l , 93°, -13°). I3inwidth is 10 kIns- l . Vertical arrows mark a periodicity of 37.6 km s-1 and zero pha.se. Further trials involved scatkring the groups around the LSC and confirmed that the coherence in phase of the periodicity, from one group to another, is real, and not an artefact induced by the optimization procedure. Thus the ~37.5 km S-1 periodicity 24 is a truly global phenomenon, the galaxies being in effect test particles whose group membership is incidental THE VIRGO CLUSTER REVISITED The Virgo cluster is the nearest rich cluster of galaxies and, at the outset of the enquiry by Guthrie & Napier (1990), it had not been used in the formulation of the quantization hypothesis. It therefore constituted an unbiased and independent sample, suitable for the purposes of testing. In their study, Guthrie & Napier (1990) first tested for quantization in a sample of 112 Virgo spirals with relatively well-determined redshifts, initially applying a correction using a solar apex (252 km S-I, 1000 , 00 ) to obtain the galactocentric redshifts. A signal was found in the range 70-75 km s-1 then under test, but at a confidence level only 0.96::; C ::;0.99. However it was found that this signal (P=71.1 km S-1 ) was strongly concentrated in 48 galaxies situated within the less dense parts of the Virgo cluster, galaxies in the core itself showing little sign of redshift periodicity. Allowing for the a posteriori nature of the finding, and the arbitrariness involved in defining 'core' and 'less dense' regions, the periodicity was confirmed at a confidence level 0.996::; C ::;0.999. A similar result was obtained when the solar vector was allowed to vary over the whole sky, the speed being maintained at V(3 = 252 km s-1 . 30 200 km/s 20 Figure 9. Power spectrum (I vs frequency) of 48 Virgo Cluster galaxies avoiding the core of the cluster. The high peak (1",26.5) is at 71.1 kms- l . The solar apex adopted in the above study was taken with respect to the Local Group. However the subsequent studies, described above, reveal that the global peri- odicity emerges strictly with respect to the Galactic nucleus: the motion of the Galaxy with respect to Local Group, Virgo Cluster or whatever seems not to be relevant. Thus in testing for periodicity within the Virgo Cluster, the Sun's galactocentric vector should have been subtracted. I have therefore repeated the analysis by conducting a box search within ±30° and ±20 kms- 1 of (220 kms- l , 960 , 00 ). A signal of strength I"'26.5 appears at P=71.1 km S-1 (Fig. 9), and varies little within the error box of the solar apex: thus there are no degrees of freedom within the error box of the latter and a single-trial probability exp-13.S '" 1.7 x 10-6 is obtained (the exponentia.i for- mula is valid to this height, against a white noise background: loco cit.). Assuming (J" ",10 kms- l , the periodicity which emerges is 71.1±1.3 kms- l which is, within the errors, the periodicity under test for a galaxy cluster. . 25 The ahove significance level should be reduced by a factor of order five to allow for the a posteriori selection of low-density regions, and a further factor of perhaps two or three to allow for the possibility of redshift interdependence through the presence of binaries (loc. cit.). Thus the periodicity hypothesis is confirmed for the Virgo cluster at a confidence level C ~ 1 - 2 X 10-5 . The earlier result yielding the same periodicity for a different vector arose because of the small angular extent of the Virgo Cluster; thus the differential solaT apex correction is small and the signal is seen over a wide range of V(0). CONCLUSION We have tested the prediction that redshifts show global periodicities ~24.2 or ~36.3 km S-1 after correction for the solar ga.lactocentric vector. We find a strong periodicity of 37.5±O.2 km S-1 to be present in accurate, independent redshift data. We have also used the Virgo Cluster to test the further prediction that a periodicity of ~72 km S-1 occurs in clusters of galaxies; we find a periodicity 71.1±1.3 km s-1 in the outer regions. The confidence levels of both these results are extremely high, and we conclude that extragalactic reclshifts are quantized. Clearly there must be a transition regime between field galaxies, loose groups and rich clusters, but this matter has still to be explored. The astronomer who wishes to build a cosmology based on quantized redshifts cannot be faulted on observational grounds. Thus the periodicities shown in Figs. 5, 8(b) and 9 are not 'statistica.l results': rather, they are the observed outcome of a single correction applied to the best heliocentric redshifts. Within the uncertainties, this solar correction is one which transforms our redshift catalogues to those which would be obtained by a civilization at the nucleus of our Galaxy. The phenomenon appears at about the expected strength for a given sample, it behaves as expected in respect of such matters as sample size, and it is robust to the choice of redshift data. If it is due to a gremlin in radio telescopes, then the gremlin concerned knows the galactocentric solar velocity. Statistical analysis enters the issue when ascertaining whether there are sufficient degrees of freedom within the error box of the solar motion for a similar result to be derived from a random redshift distribution. This question can be settled by trials, and the answer is strongly in the negative. Thus the astronomer who wishes to maintain existing cosmological paradigms must first face the challenge set by the periodicities. ACKNOWLEDGEMENTS I am indebted to Bruce Guthrie for allowing him to describe some results in advance of publication, to Franco Selleri for the invitation to address the conference, and to numerous colleagues for stimulating discussions. REFERENCES Bottinelli, L., Gougenheim, L., Fouque, p, & Paturel, G., 1990. Astr. Astrophys. Supp!. 74,391. Cocke, W.J. & Tifft, W.G., 1991. Astrophys. J. 368, 383. Fouque, P., Gourgoulhon, E., Charmaraux, P. & Paturel, G., 1992. Astr. Astrophys. Supp!. 93,211. Guthrie, B.N.G. & Napier, W.M., 1990. MNRAS 243,431. Guthrie, B.N.G. & Napier, W.M., 1991. MNRAS 253, 533. Schneider, S.E. & Salpeter, E.E., 1992. Astrophys. J. 385, 32. Tifft, W.G., 1977. Astrophys. J. 211,31. Tifft, W.G., 1980. Astrophys. J. 236, 70. Tifft, W.G., & Cocke, W.J., 1984. Astrophys. J. 287, 492. 26 QUASAR SPECTRA: BLACK HOLES OR NONSTANDARD MODELS? Jack W. Sulentic Dept. of Physics & Astronomy University of Alabama Tuscaloosa, USA 35487 INTRODUCTION The "Big Bang" model has ascended to a powerful position in modern cosmology over the past few decades. This position has become so strong that investigation of alternate ideas has almost ceased. Observational counter-evidence certainly exists (for reviews see e.g. Arp 1987; Sulentic 1987; Tifft 1987). The general belief is that this counter-evidence consists of misinterpreted data and false clues. One could easily get the impression that all of the observations fit easily into the accepted model. In fact, at least three new concepts have assumed great importance in preserving the Big Bang against observational and theoretical challenges. In temporal order of acclamation they are: 1) black holes; 2) merger phenomena and 3) dark matter. Gravitational accretion onto supermassive black holes was required as soon as it became generally accepted that the quasar redshifts were cosmological (i.e. proportional to their distance). It provides a mechanism for producing the enormous energies implied by the assumption that quasars are at their redshift distances. Mergers came upon the scene in order to account for nearby peculiar galaxies and for the increasing luminosity and size of many objects at higher redshift. Dark matter helps to bind the groups and clusters of galaxies as well as to explain the flat rotation curves in spirals. In reality, it can help to fit almost any observation into the conventional picture. What is the observational evidence for the above three phenomena? Dark matter poses a very real threat to the observational astronomy profession. The fraction of matter in the universe thought to be invisible exceeds 90% in some recent estimates (e.g. Mulchaey et al. 1993). As this number converges toward 100%, astronomy may well cease to be an observational science. At the same time alternative explanations for the dynamical peculiarities are few with the most discussed being MOND (Milgrom 1983; see also Sanders 1987). Dark matter is sufficiently unconstrained at this time to handle most challenges to Big Bang theory that might arise in the forseeable future. An example of its tremendous versatility can be illustrated by the recent "rediscovery" of an association between higher redshift quasars and lower redshift galaxies (Rodrigues-Williams and Hogan 1993; see Sky and Telescope, November 1993, p.12). The reviews of counter-evidence cited above give extensive discussion to the past evidence for these associations. This evidence was Frontiers ofFundamental Physics. Edited by M. Barone and F. Selleri. Plenum Press. New York. 1994 27 long disputed as much because of its implications as for doubts about the observations. The excess reported in the latest study was so great that it was apparently necessary to invoke a closure density of dark matter in the galaxy clusters in order to forestall a crisis for conventional ideas. In recent years, almost anything that looks peculiar or that shows a luminosity excess at some wavelength has been attributed to merger activity. In fact the observational definition for a merger is quite vague. We recently studied the observational properties of compact galaxy groups (Sulentic and Raba~a 1994ab). We found that these systems show very few of the accepted observational properties of mergers. At the same time very few potential merger remnants of past compact groups are observed. This is in spite of the fact that dynamical theory predicts that compact groups should be very unstable to collapse and coalescence (Barnes 1989). If the systems most susceptible to merging show such a low level of merger activity, how can the phenomenon be common in less dense environments? The observational evidence suggests that while mergers occur, they are relatively uncommon and cannot be used to explain most peculiar extragalactic objects. There is no direct evidence for black holes. Indirect support comes from observation of rapid X and, ray variations in active galactic nuclei (AGN:= quasars, QSO's, Seyfert galaxies, broad line radio galaxies and BLLAC objects). Evidence for massive cores, believed to be inactive black holes, in the nuclei of nearby normal galaxies is also regarded as a form of indirect evidence. We report here on a study of the emission line properties of quasars. The motivations for this study were a) to study the frequency of occurence and magnitude of the internal redshift discrepancies observed in quasars and b) to use this data to critically test predictions of physical models for the central structure of AGN. Recently this work has become relevant to the first of the above three phenomena. The observation of double peaked emission lines in the quasar Arp 102b was interpreted as line emission arising from a radiating accretion disk (Chen, Halpern and Filippenko 1989). Direct observation of emission from an accretion disk would be tantamount to proof for the existence of black holes. Our unpopular conclusion was that the bulk of the data do not support this interpretation. We consider first the basic properties of quasar line spectra followed by the results of our study. This is followed by the results of our comparison with the predictions of line emitting accretion disk models. Finally we discuss evidence that the line shifts observed in AGN might arise from a non-Doppler cause. EMISSION LINES IN AGN Normal galaxies show spectra dominated by absorption lines that arise from the composite spectra of stars that represent their principal visual constituent. The advent of sensitive detectors in the past two decades has revealed the signature of emission from hot gas in most galaxies as well. There are two kinds of emission lines that are observed: 1) permitted lines arising from transitions following photoionization (the recombination spectrum characterized in the visual by the Balmer lines and 2) the "so-called" forbidden lines arising from transitions following collisional excitation. These lines are somewhat unique to astronomy because they arise in such low density emitting regions. The principal optical features are due to [011], [01II], [NIl] and [SII]. The critical density for [01II] )'5007A is about 106 . AGN provided an introduction to UV spectroscopy long before the first satellite telescopes opened this domain to our direct study. The higher redshift objects opened up the spectral region with rest wavelenths between 1000 and 3000 A to our view. Higher ionization broad lines such as [CIV] and [CIII as well as Lyman a are redshifted into the visible region of the spectrum (e.g. we find Lyman a at about 4800A in a quasar with redshift z=3.0). The UV lines are often referred to as high ionization broad lines (HIL's) to distinguish them from the lower ionization lines observed at lower redshift 28 (LIL's). In summary, optical spectra of low redshift quasars show primarily NLR and LIL-BLR lines while high redshift quasars show primarily HIL-BLR lines. Emission line widths in velocity units are typically less that 200 km s-1 for normal galaxies. It is the singular and unifying feature of AGN that the permitted lines are very broad (full width half maxima from 103 to 2x 104 km S-l). The forbidden lines show FWHM similar to or a little broader than the corresponding lines in normal galaxies. The Balmer lines also often show a narrow component. Figure 1 shows an example of a low redshift quasar spectrum in the region near 5000A where we find both broad line (BLR) Balmer features (H,B and 1') and narrow line (NLR) [0III] features. Broad lines in AGN show peculiarities that for a long time were only discussed privately in hushed tones. They were not even mentioned in the most recent textbook written on the subject (Weedman 1986). Different lines often show different redshifts in the same object. The range of redshifts in a single object can exceed 2000 km S-l. At the same time the lines can show striking deviations from symmetry. The 5 Mark 1320 4 ~3 0 x r...'" 2 C') to C') 6'" ",::: ::r::2. 0 0 «) -< ~ ~ '"«) E''"" to :O:r:l:. ~ a:J to :':"s :W:r:: 4600 4800 5000 5200 5400 >-- (Angstroms) 5600 Figure 1. Spectrum of Markarian 1320 in the region between HI' and [0111].\5007A. Principal lines are identified. Note that H,B shows both NLR and BLR components. 1A= 10-8 cm. keen-eyed reader will have noted a redshift discrepancy in the spectrum illustrated in Figure 1. The NLR component of H,B is centered at a higher wavelength than the BLR component in that object (Markarian 1320). Figure 2 shows an even more striking example (OQ208) where the H,B component redshifts differ by 2700 km s-l. Two kinds of internal line shifts are recognized: 1) red and blue shifting of the LIL BLR with respect to the NLR in low redshift AGN and 2) an apparent systematic blueshift (700-1000 km/s) ofthe HIL with respect to the LIL. The latter shift has to be inferred from observations of two different sets of AGN since both sets of lines (in the same object) have not, until recently, become accessible to study and measurement. Our study of the LIL vs. NLR shifts was the first of a recent flurry of activity in this area. The line shifts were originally noted by Gaskell (1982) and Wilkes (1984). It is clear that the shift and asymmetry properties of the AGN emission lines are important. 29 Are they giving us an invaluable clue into the internal geometry and kinematic of the line emitting region or is it possible that the shifts are evidence for a non-Doppler phenomenon? If the former is true then they pose an important challenge for current theories while the latter possibility is regarded as unthinkable. We consider both possibilities here. OQ 208 5 o x 4 4800 5000 5200 A (Angstroms) 5400 5600 Figure 2. Spectrum of OQ208 in same wavelength region as Figure 1. Note th" large NLR vs. BLR velocity shift (",2500 km s-1). SYSTEMATIC STUDY OF LINE SHIFTS AND ASYMMETRIES Our first study (Sulentic 1989) focused on a comparison of the redshift of HI' and [OIII]. Line asymmetry properties of HI' were also measured. We chose these lines for two reasons. 1) Most published data involve optical spectra of low redshift AGN (z~0.5) where HI' is one of the most prominent features. 2) We wanted to compare the (1IL) BLR vs. NLR redshift between lines that were close together but not too close to be severely blended with one another. Ha as a LIL line was ruled out because it is redshifted out of the visible at much lower redshift and is severely blended with NLR lines of [NIl). [OIII)A4959, 5007A are close to HI' without excessive overlap in most cases. There are two other important considerations for a study of this kind: 1) the standard of rest and 2) contamination by multiplets of Fell emission. There has been considerable confusion over which, if any, lines provide a measure of the "rest frame" for the quasar. We studied all available data that might be relevant which includes 21cm emission from neutral hydrogen (HI) surrounding the AGN and absorption line redshifts from "fuzz" surrounding some low redhift quasars. The latter is thought to be starlight from a galaxy believed to be "hosting" the AGN. This data showed agreement in redshift .6V ~ 200 km s-l of the NLR redshift (Wilson and Heckman 1985). Usually the agreement was even better, leading us to adopt the NLR redshift from the lines [OIlI)A 5007 and NLR HI' (recombination emission is often observed from the forbidden line region) as our zero point in the line shift study. Some 30 recent studies have continued to use H,8 as a reference despite the fact that it shows large velocity excursions. Care should be exercized in using published data and in comparing their results with our work. The region of H,8 is infested with emission from myriads of lines arising from various multiplets of Fell. These lines inhibit accurate measurements of H,8 and [0111] as well as preclude reliable determination of the continuum level in this region. They are also a problem in other regions of AGN spectra (notably the region near rest wavelength 2800A). There are ways to model and correct for this contamination but we will not discuss them here. We avoid this discussion because our study focused on AGN where the presence of Fell emission was weak or absent. Figure 3 shows an example of an AGN with serious Fell contamination in this region. 15 Q 1126-04 12.5 00 10 0 ..::.. ~-< 7.5 Fell blends 4600 4800 5000 5200 A (Angstroms) 5400 Figure 3. Spectrum of Q1126-04 in same wavelength region as Figure 1. Note the strong Fen emission blueward of H,8. Data for Figures 1,2 and 3 were obtained at Kitt Peak. In the end we quantified the H,8 emission line properties with the following measures: a) centroid redshift at 1/4, 1/3, 1/2 and 3/4 height; b) full width half maximum (FWHM). The centroid and width at half maximum were also measured for [0111]. The LIL line shift was then measured with respect to [0111] at each of the four heights in the line profile. In addition, an asymmetry index was defined as R= [C(3/4)-C(1/4)]/FWHM. Our first study involving 61 AGN was the largest for a sample of high resolution and SIN spectra. It revealed that LIL H,8 shows almost equal numbers of red and blue shifts (and asymmetries). This was a surprise because both were thought to be predominately red. There may still be an excess of red or blue shifts and/or asymmetries, but a larger sample will be needed to establish it. Figure 4 summarizes the kinds of line profiles that we were able to identify. Most other recent studies (e.g. Corbin 1989; Francis et al. 1992) have focused on the HIL line properties derived from samples of high redshift quasars observed on the ground. They suggest that HIL are generally more symmetric than 1IL and that HIL are blueshifted with respect to LI1. The magnitude of the latter shift (700-1000 km s-l) could only be inferred indirectly since LIL (and NLR for that matter) features 31 in the same objects were unobtainable. The most recent survey of the bright low redshift PG quasar sample (Boroson and Green 1992) confirm the basic results of our study. Despite a more sophisticated PCA analysis of their sample, these authors could not uncover any significant correlations between line properties that were not known (or suspected) previously. The next step is obvious; comparison of UV and optical spectra for HIL, LIL and NLR in the same objects. This has become possible because the UV sensitivity of the Hubble Space Telescope permits us to observe the HIL lines in the same sample that we have studied on the ground. Several groups, including our own, are engaged in such comparisons at this time. IMPLICATIONS OF LINE STUDIES What do we really know or think we know about quasars? We believe that they are powered by gravitational accretion onto a black hole. The only conventional alternative (the starburst model; Terlevich et al. 1992) argues that chain reactions of supernovae are responsible for the energy output. Anyone who has seen the spectrum of a supernova has been struck by the similarities with AGN emission spectra. The BLR emission line spectra in AGN are consistent with a gas at T", 1-2x104K and the size of the BLR line emitting region is constrained by variablity studies to be in the range of 10-100 light days (1 light day= 2.6 x 1010 km.). The absence of broad forbidden emission lines like [OIII] indicates that n;::: 109 in the LIL-BLR, while the presence of weak [CIII A 1909 indicates that n::=; 1011 in the HIL-BLR. There is growing evidence for stratification in the BLR zone with HIL emitting region closer to the ionizing continuum source. Estimates of the covering factor yield small numbers, suggesting that the BLR is made up of a collection of clouds rather than diffuse gas. The predominance of forbidden lines in the LIL argues that it lies outside the BLR. Finally there is growing evidence, already implied by radio structures, that anisotropic emission may arise from jets or cones of radiation. A symmetric I~-- blue A I\ asymmetricJ I~-- A red ) I~asymmetric --~I I NO SHIFT I blue I I shift I red I shift _ _ _ I-_~/\,-+: symmetri_c_ _--+--:/ \......._ _ I ~ _ _.1\"", asy;::tric 0I ,--- 4 I _~J/\ I red asymmetric I !\ : ) L=_ I SHIFTED I I Figure 4. Combinations of emission line shift and asymmetry observed in our survey. Double peaked profiles are not included and were not found in our study. 32 Figure 4 shows that almost every possible combination of line shift and asymmetry is observed in AGN spectra. Both symmetric and asymmetric line profiles show (red and blue) line shifts. The only possible region of avoidance involves ones that are blue shifted and blue asymmetric at the same time (only one example was found).This stochastic property provides a very strong constraint on possible models for the kinematics and geometry of the emitting region. Favored models have invoked either a dominance of gravitational forces (infall or rotation) or radiative pressure driven outflow (ballistic models have also been considered). The stochastic nature of the line profile properties rules out such "single-force" models. It requires some hybrid explanation that can account for both red and blue shifts. We have considered the possibility that the BLR line radiation originates anisotropically (Zheng, Binette and Sulentic 1990) in a double stream model. Gaskell (1983) proposed a binary black hole model and Mathews (1993) recently proposed a model involving bouncing clouds. Any conventional model is a long way from adequately explaining the observations. It is usually possible to adequately model any single AGN but an attempt at generalization always leads to difficulties. 4 p::CIl ...:IlCl Arp 102B Zc-o< :r:g~o;:::' .-< ('J c/'o0 3 co to .-< .'m.>..:., ...