John N. Baheall Jeremiah P. Ostriker, Editors UNSOLVED PROBLEMS IN ASTROPHYSICS e PRINCETON UNIVERSITY PRESS Princeton, New JerseyCopyright © 1997 by Princeton University Pest Published by Princeton University Press. 4 William Stet Princeton, New Jersey 08540 In the United Kingdom: Princeton University Press, (Chichester, West Sussex A igh Reserved ISBN 0-691.01607-0(€1) ISBN 0-691-01606-2 (pa) ‘he publisher would ike to acknowledge the editors ofthis volume for providing the eamer-eady copy from which this book was pinted ceton Univerity Pres books are printed on acide paper and meet the ideins for perrsanence and dumbiity ofthe Committe on Production (Guidelines for Book Longevity ofthe Couneil on Library Resouces btpurpup princeton Printed inthe United States of America 3579108642CosMOLOGY AND LARGE SCALE STRUCTURE 1, The Cosmological Parameters -- PILE, Peebles, 2, In The Beginning. .+...s.essecsee+ «Paul J. Steinhardt 3. Understanding Data Better with Bayesian and Global Statistical Methods ....+..c.csscsscessesseeesees e+» Williarn H, Press 4, Large Scale Structure inthe Universe....... »Neta A. Bahcall 5. Unsolved Problems in Gravitational Lensing sosR.D, Blandford, ‘6. What Can be Learned from Numerical Simulations Of Cosmology? .....esscsseceeseeseesesvseos Jeremiah P, Ostiker GALAXIES AND QUASARS 7. The Centers of Elliptical Galaxies coeescees Scott Tremaine 8. The Morphological Evolution of Galaxies. Richard . Ellis 9. Quasars . ASTROPHYSICAL LABORATORIES. 10. Solar Neutrinos: Solved and Unsolved Problems. ... 11. Particle Dark Matter 12, Stars in the Milky Way and Other Galaxies ....... 13, Searching for MACHOs with Microlensing .............Chasles Alcock 14, Globally Asymmetric Supernova «Peter Goldreich, eta 15, In and around Neutron Stars. eooces--Malvin Ruderman 16, Accretion Flows around Black Holes. ‘Ramesh Narayan 17. The Highest Energy Cosmic Rays... James W. Cronin 18, Toward Understanding Gamma-Ray Bursts. ‘ ‘Tsvi PiranCONTENTS fPeevace, viii 4 THE COSMOLOGICAL PaRAMereRs 1 1.1 INTRoDUCTION : 1 12. Wity MEASURE THE PARAMETERS? . . 2 12.1 Testing the Physics . 2 1.2.2 How Willlt All End? . ee 13. THE STATE OF THE MEASUREMENTS . 2 il 1A CosMoLoGy FOR THE NEXTGENERATION |... 0.6 17 2 INTHE BEGINNING ... 28 2.1 THE FUTURE FATE OF COSMOLOGY .............+. 25 22 TESTING INFLATION... . = aT 23. THE POWER OF THE CosMic MICROWAVE BACKGROUND |. . 28 24 Cosmic CONCORDANCE 2s 25 ANEWAGE?......... 40 3 UNDERSTANDING DATA BETTER WITH BAYESIAN AND GLOBAL STATISTICAL METHODS o 3.1 InTRopucTION : vee 49 32. CoMBINING EXPERIMENTAL MEASUREMENTS |. 50 33. BAYESIAN COMBINATION OF INCOMPATIBLE MEASUREMENTS 51 34° ANOTHER VARIANT OF THEMETHOD .......-...... 56 3.5 RESULTS FOR THE HUBBLE CONSTANT ......-...... 56 56 ConcwuNON: se 39 4) LARGE-SCALE STRUCTURE IN THE UNIVERSE a 4.1 InTRODUCTION : sete OL 42° CLUSTERING ANDLARGE-SCALE STRUCTURE =... 65 421 Galaxies and Large-Scale Structure... +. 0+ +++. 65 422 Clusters and Large-Scale Structure... 44... +... 68 43° PECULIAR MOTIONS ON LARGE SCALES "4 44. DARK MATTER AND BARYONS IN CLUSTERS OF GALAXIES |. 79 45 18% <1? cette eee 46 THESDSS AND LARGE-SCALE STRUCTURE... 0... 824.6.1 The Sloan Digital Sky Survey... . 82 46.2 Clusters of Galaxies . . : 3 47 SUMMARY «2... ee . UNSOLVED PROBLEMS IN GRAVITATIONAL LENSING % 5.1 INTRODUCTION ..... ta 94 52 GRAVITATIONAL LENS OPTS |... : os Ot 53. THE PROBLEMS : cece 8B 53.1 How Old Isthe Universe? os 98 532 Whats the Shape ofthe Universe? 101 53.3 What ls the Large Scale Distribuionoy Matter? 101 534 How Are Rich Clusters of Galaxies Formed? ..... . . 192 535 When id Galas Form and ow Di hey Boe? 536 How Big Are Galaxies? . . a 53.7 Of What Are Galasies Made? 107 538 How Big Are AGN Ultraviolet Emission Regions? 107 54 HOW MANY MORE SURPRISES WILL GRAVITATIONAL LENSES PROVIDE... 20... 108 ‘WHAT CAN BE LEARNED FROM NUMERICAL SIMULATIONS OF COSMOLOGY 1s Gi WWODCHON 1G (62 SIMULATION METHODS 2c: ug 62.1 Specification of Models... . ne 62.2. Physical Processes and Numerical Methods se 12 63. RESULTS: COMPARISON WiTH OBSERVATIONS... . . 126 63.1 Hot Components . . centre eee ee 126 632 Warm Components . . oF oe iat 633 Cold Condensed Components : 8 64 CONCLUSIONS, PROSPECTS, AND MORE QUESTIONS... . - 129 ‘THE CENTERS OF ELLIPTICAL GALAXIES 17 7.1 INTRODUCTION 13 TAA Black Holes and Quasars is +. 138 TA The Sphere of Influence... vo vee eevee e139 713. Cores and Cusps . ... 0 72 PHOTOMETRY ..... : 142 721 The Peebles Young Model... 14s 73. KINEMATIC EVIDENCE FOR CENTRAL BLACK HOLES 47 74 PHYSICAL PROCESSES ae 75 SUMMARY ee 8 12convents: THE MORPHOLOGICAL. EVOLUTION OF GALAXIES. 81 ITRoDUCTION : 8.2 EARLY FORMATION OF MASSIVE ELLIPTICALS 83 SLOW EVOLUTION oF MASSIVE Disk GALAXIES 84 REDSHIPT SURVEYS AND THE DWARP-DOMINATED UNIVERSI 85 FAINT GALAXY MORPHOLOGIES FROM HST 8.6 CoNcLusions 9 QUASARS 9.1 QUASARS AND THE END OF THE ‘DARK AGB’ 92 THE RELATION OF AGNS To THE CENTRAL BU OF GALAXIES 9.3 QUASARS AND THEIR REMNANTS: PROBES OF GENERAL RELATIVITY? 93.1 Dead Quasars in Nearby Galaxies. 93.2 Do These Holes Have a Kerr Metric? . 10 SOLAR NEUTRINOS: SOLVED AND UNSOLVED PROBLEMS 10.1 Wity STUDY SOLAR NEUTRINOS? 159 159 161 164 166 170 174 181 181 183 186 186 187 195 195 10.2 WHat Dos THE COMBINED STANDARD MobEI, TELL US ABOUT SOLAR NEUTRINOS? . 102.1 The Combined Standard Model 10.22 The Solar Neutrino Spectrum : 103. WitY ARE THE PREDICTED NEUTRINO FLUXES ROBUST? 104 WHAT ARE THE THREE SOLAR NEUTRINO PROBLEMS? 104.1 Calculated versus Observed Chlorine Rate 1042 incompatibility of Chlorine and Water (Kamiokande) Es- eriments . . 104.3 Gallium Experiments: No Room for Be Neutrinos 105 Witat HAVE We: LEARNED? 10.1. About Astronomy . 10.5.2 About Physics 10.6 Waar Nexi? 106.1 Solvable Problems in Physics 10.62. Solvable Problems in Astronomy 10.7 Summary 11 PARTICLE DARK MATTER 11,1 INTRODUCTION; THREE ARGUMENTS FOR NON-BARYONIC DARK MATTER 11.2 THE CASE FoR NON-BARYONIC MATTER 198 198 200 201 202 203 = 208 205 206 7206 208 209 209 212 27 224 2 22x Convenes 112.1 We've Looked for Baryonic Dark Matter and Failed... 222 11.22 We Cant Seem Te Mate the ObvervedLargeSonte Structure with Baryons 23 11.23 Dynamical Mass Is Much Larger than Big Bang Nucleosynthesis Allows : 24 11.3 NBUTRINOS AS DARK MATTER... 0-00-0022 +, 225 1131 Detecting Massive Neutrinos ...... - - 226 114 WIMPs Cee ay WA: Seng WIMP ae 11.42 Indirect WIMP Detection . . - 230 143 What Is To Be Done? . . 231 115 Axtons ..... . Co i 116 Conctustons : +233 12 STARS IN THE MILKY WAY AND OTHER GALAXIES 241 12.1 INTRODUCTION... . oe aeesoecce ml 122 RECENT STAR COUNT RESULTS... 2a 123 MICROLENSING AND STAR COUNTS... 06.0. oe ee ss 283 12,4 Disk DARK MATTER: STILL A QUESTION ae. 12.5 MYSTERY OF THE LONG EVENTS . . . . cence Dd 12,6 PROPER MOTIONS FROMEROS Il... .. - - wae 245; 127 PIXEL LENSING: STELLAR Mass FUNCTIONS IN OTHER GALAXIES. 26 128 STAR FORMATION HISTORY OF THE UNIVERSE |=. 248 12.9 ConcLUSIONS ........ ee = 249 13 SEARCHING FOR MACHOs WITH MICROLENSING 253 13,1 INTRODUCTION eee ee a 132 THE GRAVITATIONAL, MICROLENS tt 13.3 THE “MACHO FRACTION” IN THEGALACTIC HALO... . . . . 256 134 THE EXPERIMENTAL SITUATION... . . - : 258 13.5 NEXT GENERATION EXPERIMENTS . . . . . fe ee 260 135.1 What Can Be Achieved from the Ground? - = 260 13.5.2 Observing Macho Parallax. . 261 136 WORKING ON GRAVITATIONAL MICROLENSING 263, 13.7 SUMMARY 137.1 What We Know Now 13.72. What We Will Learn from Current Experiments 13.7.3. Next Generation Experiments 13.8 LATE BREAKING NEWS . . + 265Conrenrs 14 GLOBALLY ASYMMETRIC SUPERNOVA 14.1 INTRODUCTION 14.1.1 Preamble .. 14.12. Bvidence for Asymmetry 141.3. State of the Art 142 INSTABILITY DURING CoRE COLLAPSE - 142.1 Accomplishments... 1422 Future Directions... . 143 OVERSTABLE CORE G-MODES . 143.1 Accomplishments 143.2. Future Directions : 1433 Turbulent Excitation of Modes... 15 IN AND AROUND NEUTRON STARS 15.1 INTRODUCTION ......... 152 SUPERFLUID-SUPERCONDUCTOR INTERACTIONS IN A NEUTRON STAR CORE... .. 2. 153 THE STELLAR CRUST 15.4 SPUN-UP NEUTRON STARS . . 15.5. SPINNING-DOWN RADIOPULSARS 156 GLITCHES OF RADIOPULSAR SPIN PERIODS - 16 ACCRETION FLOWS AROUND BLACK HOLES 16.1 INTRODUCTION eee 162 K-RAYS AND 7-RAYS FROM ACCRETING BLACK HOLES 16.3 HoT ACCRETION FLow MODELS . 163.1 Corona Models... . 1632 SLE Two-Temperature Model 1633 Optcally-ThinAdvection-Dominated Model . 164 DIRECTIONS FOR FUTURE RESEARCH 164.1 Unresolved Theoretical Issues 164.2. Clues from Observations of Black Hole XRBS 1643 Black Holes versus Neutron Stars. . 165 ConcLusion 17 THe HiGuEsT ENERGY Cosmic Rays 17,1 INTRODUCTION... . . 172 REVIEW oF EXISTING DATA ON THE HiGHEST ENERGY Cosmic Rays 17.3 ACCELERATION AND TRANSPORT OF TIIE COSMIC Rays > 10°BV ... = 306 2B 274 2.205 - 276 27 281 281 - 284 =. 289 2.293 + 294 301 301 302 = 308 305 306 311 <3 313, 31s 316 325 325 327 32918 174 Tue BiG Events 175 THE AUGER PROJECT 176 WHAT CAN WE LEARN FROM TWO LARGE SURFACE ARRAYS? .. . 177 SHOULD A'STUDENT WoRk oN Tits PROBLEM? 178 FINAL REMARK ‘TowARD UNDERSTANDING GAMMA-RAY BURSTS 18.1 INTRODUCTION, bee eeee 182 OnseRvaTions 182.1 Observational Open Questions 18.3 A BRIEF SUMMARY bette eee 184 WiERE? 18.5 How? . 18.5.1 The ‘Compactness Problem 18.5.2 Relativistic Motion . Fi 18.5.3 Slowing Down of Relativistic Particles oe 185A. The Acceleration Mechanism? 18.6 WuaT? . 18.6.1 What Do We Need from the internal Engine? 18.6.2 Coincidences and Other Astronomical Hints . . 18.7 Way? . : 18.8 ConcLusioNs | | 189 SoME OPEN QuésTiONs Convants - 332 335 337 338 339 367 369 369PREFACE ‘The articles inthis volume were written in response to the following hypothetical situation. A second year graduate student walks into the author’s office and says ‘am thinking of doing a thesis in your area. Are there any good problems for me to work on?” [Leading astrophysical researchers answer the student's question in this collec- tion by providing their views as to what are the most important problems on which rajor progress may be expected in the next decade. The authors summarize the current state of knowledge, observational and theoretical, in their areas. They alS0 suggest the style of work thal i likely to be necessary in order to make progres. ‘The bibliographical notes al the end of each paper are answers othe parting ques- tion by the hypothetical graduate student: “Is there anything I should read to help sme make up mind about a thesis?” ‘As everyone knows who reals a newspaper or listens to the daily news, astto- physics isin the midst of a technologically driven renaissance; fundamental dis- coveries are being made with astonishing frequency. Measured by the number of ‘professional researchers, astrophysics is a small field. But, astronomical scientists have the entire universe outside planet earth as their exclusive laboratory. In the last decade, new detectors in space, on earth, and deep underground have, when coupled withthe computational power of modern computers, revolutionized our knowledge and understanding of the astronomical world, This isa great time for a student of any age to become acquainted with the remarkable universe in wich we live. In ofder to make the texts more useful to students, each ofthe papers was “ref- creed” by cooperative graduate students and colleagues. On average, each paper was refereed four times. We would like to express our gratitude to the referees; their work made the papers clearer, more accessible, and in some cases, more cor- rect, We are grateful to each ofthe authors for wonderful manuscripts and for their cooperation in what must have at times seemed like an endless series of iterations. ‘The excellent quality ofthe final texts justifies their hard work, In Jooking over the material as it now appears, we believe that these papers ‘may have a Wider readership than we originally anticipated. Most of the articles are accessible to junior or senior undergraduate students with a good science back- ground. The book can therefore be useful as an undergraduate introductionto some of the important topics in modern astrophysics. We hope that readers who arexiv PREFACE ‘graduate students now or in the future will solve many of the problems listed here a8 unsolved. Anyone, from an undergraduate science major to a senior science faculty member, who would like to know more about some of the active ateas of contemporary astrophysics can profit by reading about what these researchers think are the most important solvable problems. “The antics collected here were originally presented as invited talks at a con- ference entitled ‘Some Unsolved Problems in Astrophysics” that was held at the Institute for Advanced Study on April 27-29, 1995. This conference was spon- sored in part by the Sloan Foundation, to whom we express our gratitude, The dates forthe conference were related to the 60th birthday (and 25th year at the Tnstitue) of one of us (INB), but nearly every effort was made to focus the meet- ing on science, not anniversaries, However, large fraction of the attendees and speakers were alumni of the Institute's postdoctoral program in astronomy and astrophysics. ‘The manuscript for this book was expertly prepared by Margaret Best. All of us are grateful to Maggie for her exceptional editorial and TeX skills and for her constant good nature, John N. Bahcall, Jeremiah P, Ostriker Princeton, June 1996Unsolved Problems in AstrophysicsCHaPTER 1 ‘THE COSMOLOGICAL PARAMETERS PJ. E. PEEBLES Joseph Henry Laboratories, Princeton University, Princeton, NJ ABSTRACT ‘The tests ofthe relativistic cosmological model are well understood: what is new is the development of means of applying them in a broad variety of ways capable of producing a network of consistency checks. When the network is tight enough ‘we will have learned either that general relativity theory passes a highly nontrivial test or that something is wrong with classical physics. I see no reason to look for the later but we should keep it in mind. Inthe former casc we will have gained the ‘boundary conditions for a deeper cosmology, and a new set of puzzles to study. 11 Itropucrion The firs thing to know about the measurement of the parameters of the standard telativistic cosmological mode! is thatthe problem has been with us for @ long time. By the 1930s people understood the physics of the evolving relativistic cos- ‘mology and how astronomical observations might be used to test and constrain the values ofits parameters, The first large-scale application of the astronomical tests, the count of galaxies as a function of apparent magnitude (Hubble 1936),had Aready reached redshift z ~ 0.4 (inferred from the values of Hubble’s counts and ‘more recent measurements of the mean redshift-count relation). The application Of the cosmological tests was a “key project” for the 200 inch telescope when it ‘was under construction in the 1980s; now itis a key project for he Hubble Space ‘Telescope and the Keck Telescope. ‘There has been ample time for the development of strong opinions on what the results of the measurements arc likely tobe, and fora tendency to lose sight of2 ‘CHAPTER 1. THE COSMOLOGICAL PARAMETERS the reasons for measuring the parameters. To my mind there are two main goals: ‘extend the tests ofthe physics ofthe standard model, and seek clues to acosmology ‘based on a deeper level of physics. Despite the sobering record there is reason to believe we may actually be witnessing the end game in finding useful measure- ‘ments of the parameters. 1.2 WHY MEASURE THE PARAMETERS? 121 Testing the Physics In the standard cosmological model the universe is close to homogeneous and iso- tropic inthe large-scale average, and itis homogeneously expanding and cooling. ‘The evidence, which most cosmologists agree is quite strong, is summarized for example in Peebles, Schramm, Turner, and Kron (1991) and Peebles (1993). (I refer the reader to these references for details of the following comments.) The evolution ofthe standard model is described by general relativity theory. This part is not as closely probed, and one goal of the cosmological tests isto broaden the ‘constraints on the underlying gravity physics. Here isthe situation, “To begin, I assume the geometry of our spacetime is described by a single ‘metic tensor, that i, a single line clement which determines the relations among, measured distance or time intervals between events in spacetime. The evidence is, that our universe is close to homogeneous and isotropic in the large-scale average, and we know the lineelement of a homogeneous and isotropic spacetime is unique up to coordinate transformations; the Robertson-Walker form is! ds? ae pe 4 dP — ea, P= a dg? 4 r%sin?odg?. (LI) In this equation a comoving observer, who moves s0 the univers is seen to be isotropic in the large-scale mean, has fixed position r, 8, 6. The proper time kept by such comoving observers ist (ifthe observers’ clocks are synchronized so all see the same mean mass density pat a given time t). The parameter R-? in the expression di? for the spatial part ofthe line element measures the radius of curvature of the three-dimensional space sections of fixed world time t. Ifthe refactor of R? is negative space is closed, as in the surface of a balloon. The ‘expansion factor a(t) in the expression for ds? means the balloon in general may ‘be expanding oc contracting; the evidence is that our universe is expanding. If the prefactor of R~? is negative space sections are open, the circumference of ‘a circle of radius + being larger than er. In tis case the nearly homogeneous " Comecions forthe departs fram homogeastaeimporaat for some he comologicaltest, and will have tbe reconadered athe ets prove. Tho tues ae acisied ia Pebles (1993).1.2. WHY MBASURE THE PARAMETRRS? 3 space we observe may extend to indefinitely large distances, or space might be {odic, or conditions beyond the distance wecan observe mightbe very different. IR“? = O then dl? is the familiar Cartesian form for flat space, and spacetime is said to be cosmologically Nat. ‘The proper physical distance between comoving observers, measured at given ‘world time, is D = la(t), where the coordinate distance! is the result of integrat- ing the second part of equation (1.1) along the geodesic connecting the observers, at fixed t. The rate of change of the proper distance is aD dt do aa Ifvismuch less than the velocity of light this gives a good picture for the predicted linear relation between distance and relative velocity in a homogeneous expand- ing universe.? Thus a galaxy at distance D is moving moving away al recession 280, 8 2) velocity v = HD, and the resulting Doppler effect stretches the wavelength of the radiation received from te galaxy by he amount a =D = HDke a3) ‘Theconstant of proportionality #7 is Hubble's constant; itspresent value usually is ‘writenas H. The linear relation between redshift and distance, whichis Hubble's law, has been tested to reilshifts onthe order of unity; for an example see Figure 7 in McCay (1993). Hubble's Iaw was one of the fist pieces of evidence leading to the discovery Of the relativistic cosmology, but we see that the functional form follows more ‘Benerally from the observed large-scale homogeneity of the universe. When the redshift is comparable to or larger than unity equation (1.2) does ‘ot directly apply (because D is measured along a surface of fixed cosmic time 4, which is not how light from a distant galaxy travels to us). The easy way to ‘analyze the redshift in this case isto imagine that the electromagnetic field is de- ‘composed into normal modes of oscillation with fixed boundary conditions in the pace coordinates of equation (1.1) The boundary conditions mean the physical ‘wavelength A of a mode streiches as the universe expands: Acca(t). an 1 he interaction of the radiation with other matter and feds is weak then adi- batty tells us the number of photons in esch mode is conserved, which is to As in equation (1.1), this assurnes perfect homogeneity. I the teal world the galaxies are moving peciarelcoes~ 800 an lave othe el Hae How mequaon 0)4 (CHAPTER 1. ‘THE COSMOLOGICAL PARAMETERS ‘ay that equation (1.4) gives the time evolution ofthe wavelength of freely propa- ating radiation as measured by comoving observers placed along the path of the radiation. The cosmological redshift factor z is defined by the equation a(love) (tem) “The wavelength at emission from a source at epoch fem i Nem 88 Measured by 1 comoving observer at the source, and the radiation is devected at time tote at wavelength Aste Ifthe time between emission and detection is small we can ex- pand a(t) in a Taylor series o get equation (2.3), Ithas become customary to abel epochs in the early universe by the redshift (actor = considered as an expansion factor even when there is no chance of detection of radiation freely propagating to 1s from this epoch, ‘The thermal cosmic background radiation (the CBR) detected at wavelengths of millimeters to centimeters is very close to homogeneous — the surface bright- ness departs from isotropy by only about one past in 10° — and the specsrum is very close to blackbody at temperature T = 2.73 K. To analyze the behavior of this radiation in an expanding universe suppose the homogeneous space of equa- tion (1.1) at time ¢ contains a homogeneous sea of thermal radiation at tempera- ture. The photon occupation number of a mode with wavelength Ais given by Planck’s equation, ltrs as) 1 = aT a6) Ifthe radiationis freely propagating the occupation number NV is conserved, so we see from equation (1.4) thatthe mode temperature scales with time as Tc W/att). an Since this is independent of the mode wavelength am initially thermal sea of radi- ation remains thermal, even in the absence of te traditional thermalizing grain of dust ‘We know the universe now is transparent to the CBR, a east along some lines of sight, because distant galaxies are observed at CBR wavelengths. AL some ear- lier epoch the universe could have been dense and hol enough tohave been opaque and therefore capable of relaxing the radiation to the observed thermal spectrum. ‘When this was happening the interaction of matter and radiation cannot be ne- lected, of course, but since the heat capacity of the radiation is much larger than that of the matter equation (1.7) sil applies tothe coupled matter and radiation ‘when sources of sinks of the radiation may be neglected. N “ie sucey deny in the radiation bs aT, where ait Stan's constant. Ta plums of = per on volun and a Like numberof fice elecons the eneey density is Sak, ‘The rio (aT /{nk) ~ 10% early indoponden of redshift.ya, WHY MBASURS THE PARAMETERS? 3 ‘The conclusion is thatthe CBR is @ fossil of atime when our expanding uni- verse was hotter and denser than it is now. This argument does not require gen- ‘gal relativity theory, only conventional local physics, a nearly homogeneous and isotropic expansion described by a single line element (as in eq. (1.1)), and an ‘expansion ecto large enough tolead back oa time when the universe was opaque ‘enough tohave been capable of relaxing to equilibrium. I think all who have given the matter serious thought would agree wit this; the issue isthe minimum expan son factor needed to account fo the observations. Hoyle, Burbidge, and Narlikar (1993) proposea Quasi-Steady State scenario in which the present expansion phase traces back to redshift only slightly greater than the largest observed for galaxies ‘Others doubt that the properties of the postulated thermalizing dust grains can be hosen to relax the radiation blackbody at such low redshifts while sill allowing be observed visibility of high redshift galaxies at CBR wavelengths, though the {ssue certainly could be analyzed in more detail than has been done by ether side ‘so ar, Most cosmologists accept the evidence for the origin of the lightelements as, remnants of the rapid expansion and cooling ofthe universe through temperatures ‘on the order of 1 MeV, at redshift = ~ 10'°. This model for element formation depends on the rate of expansion through z ~ 10° and thus tests the gravity sheory, a5 follows. Tn general relativity the expansion factor a(¢) in equation (1.1) satisfies wa(E) hoes on ‘The mean mass density isp, A is Einstein's cosmological constant, and the con- ‘stant R? appears with the same algebraic sign as in equation (1.1). (I simplify ‘the equations by choosing units so the velocity of light is unity.) The equation ‘of Local energy conservation is j/p = —3(4/a)(p + P), where P is the pres- ‘sure. If the pressure is not negative the mass density varics with the expansion Parameter at least as rapidly as a~®, meaning it is the most rapidly varying term in ‘equation (1.8), and hence the dominant term at high redshift. Thus the predicted ‘expansion rate through the epoch of light element productions very well approx- ‘mated os : (39) = eo. as) For a reasonable value of the baryon number density, and assuming the beryon ‘istibution a high redshift is close to homogeneous and the lepton numbers are ‘umall, che predicted values ofthe lightelement abundances left over from the rapid ‘expansion and cooling ofthe early universe are close tothe observed abundances (with modest and reasonable corection forthe eects of nuclear burningin star).6 (CHAPTER 1. THE COSMOLOGICAL PARAMETERS ‘This isa non-trivial test of geneal relativity theory, through equation (1.9). Is, not difficult to arrive at equation (1.9) in a quasi-Newtonian picture, however, s0 the test arguably is not very specific, ‘Theprogram of cosmological tests may extend the checks ofthe gravity theory. (One part ofthe program uses galaxies or other markers to measure the curvature of spacetime. Inthe description of our universe by the homogeneous and isotopic line clement in equation (1.1) this amounts to measuring the value and sign ofthe parameter R? andthe expansion parameter a asa function of world time, These results with equation (1.8) predict he present value of the mean mas density 9, which we can hope to compare to what is deduced from studies ofthe dynam. ies of systems of galaxies, and the expansion time t(2) a6 a function of redshift, which we can hopeto compare to radioactive decay ages and stellar evolution ages applied atthe present epoch and, fom stellar evolution times derived from the specra of distanc galaxies, at earlier times. In equation (1.8) the parameter R=? Plays the role ofa consant of integration ora conserved energy forthe effective value /2 ofthe kinetic energy of expansion pe unit ass, whileineqution(I.1) itis a measure ofthe curvature of space sections. It willbe fascinating to see a check thatthe same parameter plays both roles, a predicted in general relativity. TW may be useful to note ina litle more deal the theory behind the use of astronomical observations to constrain a(t) and 2, We can define two distance functions, the proper rate of radial displacement of a light packet wit respect 0 redshi, a=, 19) and the angular size distance, vey rion 2 fA), aay ‘which isthe coordinate distance r in equation (1.1). This expression assumes an open model; ina closed model the hyperbolic sine is replaced by a trigonometric sine, Here are examples of the uses of these functions. ‘The integral of s(2) in equation 1.10) over redshiftis theexpansion time, tobe compared to other measures of time. The function s(z) also enters the analysis of the absorption lines in quasar spectra that give such a remarkably detailed picture of the distribution of gas along the line of sight. Ifa class of clouds has proper smumber density n(2) and cross section o(z) at redshift = then the density of ab- sorption lines produced by these clonds, measured asthe probability of finding a line ina quasar spectrum at redshift z inthe range dz. is AP = o(z)n(z) s(2) 42 1.12)1.2. Wav Measuas 7 ParaMarEns? 7 ‘A galaxy with angular diameter 0 appearing a redshift z has proper diameter d= alz)r(2)9, a3) which is why r is called the angular size distance. In a metric theory the surface brightness of a galaxy integrated over wave- length varies as (1 + z)~*. Iti an interesting exercise for the student to derive this from Liouville’s theorem, and to check that the angular size distance in equa- tion (1.13) thus predicts the detected energy flux density J from a galaxy with Iuminosity L isotropically radiating at redshint =: L Sabre 2 ‘The count of objects per steredian depends on the radial and angular size dis- tances: f aay an a ‘The first factor isthe proper area per steradian subtended at redshift z, the second factor isthe proper radial displacement of alight packet per increment of redshift, ‘and n(2) is the number of objects per unit proper volume, Similar expressions follow for other measurements that might be done at least {inpprinciple. A real measurement is quite another matter, of course, but there has been impressive progtess, a few examples of which will be considered in the next section. [a(z)r(2)]2a(2)n(2). ats) 1.2.2 How Will It Aul End? ‘kis fascinating to think we can discover how the world ends; the standard cos- ‘mological model offers a few definite possibilities. One is a violent collapse back toa “Big Crunch,” sort of time-seversod “Big Bang.” Another s expansion into the indefinitely emote future, a "Big Chill” marked by the eventual deaths ofall ‘the stars and by the eventual relativistic collapse of gravitationally bound islands ‘of matter and the slow evaporation and dissipation ofthe mass ofthe resulting lack holes, All this is romantic, but the science is debatable: why should we trust an extrapolation into the remote fute ofa theory we know can only be an ‘Approximation to reality? Surely the more likely outcome of a successful applica- tion ofthe cosmological tests would be the discovery ofthe boundary conditions for some deeper physical theory which most of us hope will not be blemished by the singular intial and final states ofthe universe or pars oft that follow within ‘Seneral relativity theory. According to this way of thinking an examination of the Bstory of eas on how the universe ought to cad is valuable forthe hints it ight8 (CHAPTER 1. THE COSMOLOGICAL PARAMETERS. offer to adeeper theory, and we hope aknowledge of how the world ends withinthe present model will focus our minds on the search for a deeper picture, While we ‘wait to soe whether anything comes ofthis we can contemplate some interesting examples in the sociology of science. If the curvature term does not stop the expansion then within the standard ‘model in equation (1.8) the long-term future ofthe universe depends on the value of A. A negative value, however small, eventually returns the universe to a Big Crunch. If Ais identically zero and the curvature term does not stop the expan sion then spacetime becomes arbitrarily close to Minkowski in the remote future. ‘Dyson (1979) points out that this allows interesting things to contimue happening for avery long time, albeit operating at ever increasing timescales, If Ais positive and the curvature term does not stop the expansion then the universe approaches the static de Sitter spacetime (where the cosmological constant term dominates the right-hand side of eq, [1.8)). In this Himit spacetime returns to inital conditions. like those of the cosmological inflation scenario described in these progeedings by Paul Steinhardt, who was one of the leaders in its development. Oxhe might ‘imagine that this signals another round of ination that willbe followed by aphase transition that converts the present A into entropy. The characteristic mass cor- responding to an astronomically interesting A ig only a milli-clectron volt, but it oes not seem inconceivable that a complex new world could develop out of what is to us an exceedingly small energy density (Misner 1992). If this does seem inconceivable one might consider the idea that the Big Chill ends in a burst of ‘creation of new material, asin the Quasi-Steady State scenario (Hoyle, Burbidge, & Narlikar 1993), the material being created at densities and temperatures large ‘enough Lo satisfy the observational constraints (assuming that what comes after resembles what we see) Lemaitre (1933) considered the idea that the collapse of a universe to a Big Crunch might be followed by a bounce and a new expanding phase, To Lemaitre this picture has in my loose translation) “an incontestable poetic charm, bringing to mind the Phocnix of the legend,” and others certainly have agreed. For exam- ple, Dicke was led tothe idea thatthe universe might contain an observable ther mal cosmic radiation background (the CBR) through the idea that starlight from the previous phase of an oscillasing universe would be thermalized during a deep ‘enough bounce and would be capable of evaporating the stars and heavy elements to provide fresh hydrogen for the next cycle (Dicke, Peebles, Rol, & Wilkinson 1965), There are no generally accepted ideas onthe physics of @ bounce, but evi- dlence tat there willbe a Big Crunch mightbe expected toconcentrate our thoughts, Einstein's original staticcosmological mode! requires nonzero values for space curvature and the cosmological constant, and it was natural therefore thatthe frst1.2. WHY MBASURE THE PARAMETERS? 9 siscussions of the evolving case, by Friedmann and by Lemaftre, included these terms. Einstein and de Sitter (1932) noted that neither stequired 1 account forthe ‘expansion of the universe, and nether was needed fit the available constraints on the values ofthe expansion rate, the mean mass density, and space curvature. They therefore that one might pay particular atention tothe case in which the sass density dominates the right-hand side ofthe expansion equation (1.8), as in equation (1.9). This has come to be called the Einstein-de Sitter model, Einstein ‘nd de Sitter concluded by remarking that the “curvature is, however, essentially determinable and an increase in the precision ofthe data derived from observations will enable us inthe future to fx ts sign and to determine its valu.” 1 do not get the impression from this thatthe authors felt the only reasonable possibility isthe Binsicin-de Sitter model. Prior to the discovery ofthe inflation scenario others were similarly cautions. Robertson (1955) stated that the Einstein-de Sitter case is “of some passing in- teres.” In the addendum to the second edition of his book, Cosmology. Bondi (1960) was a litle more enthusiastic: he wrote that, “in addition to its outstanding simplicity, this model has the remarkable property (unique amongst relativistic models) thet 7p R?/ isthe density parameter defined in equation (1.16) ‘up to a numerical factor (and 7 = G is Newion’s constant), “is constant. It may well be argued that as important a simple pure number as 7pR?/R? should be constant during the evolution of the universe so as to provide in some sense a ‘constant background to the application of the theory.” Bondi’sremark is close tothe coincidences argument I believe originated with Dicke. He noted that if the universe did expand froma dense statcthen there would Ihive been an earliest time inthe past when the physics of the expanding universe ‘an be well approximated by classical ied theory. Whatever had happened earlier ‘would have st the intial conditions for the subsequent classical evolution, andin Particular would have set the values ofthe time ty at which the mass density ceases tobe the dominant term in the right-hand side ofthe expansion equation (1.8) and the time at which we came on the scene ané measured the parameters in this ‘equation. We know ta is not much larger than ¢;, because the known masses of the galaxies contribute atleast 10% tothe present value of H? in equation (1.8). It Would be a curious coincidence it; and tz had similar values. The ikely possiil- ity therefore is that 3 i mach larger than ¢2, meaning we sil are in the Einstcin- e Sitter phase. 1 think F first learned this argument from Dicke in about 1960, ‘when he led me into research in gravity physics and cosmology; he got around to Publishing ita decade later (Dicke 1970), If others knew the Dicke coincidences argument before the discovery of the ‘inflation scenario it did not prevent discussions of alternatives. For example, Pet-10 (CHATTER L. THE COSMOLOGICAL PARAMETERS rosian, Salpeter, and Szerkes (1967) and Shklovsky (1967) discussed the possible role of a nonzero cosmological constant A in interpreting the quasar redshift dis- tribution, and Gott etal. (1974) argued that the estimates ofthe mean mass density and the expansion rate Ho, and the theory of light element production i the early ‘universe, seemed to be most readily understandable if the mean mass density is less than the Binstein-de Sitter value. During the mid-1980s there was a sharp swing toward the opinion that the instein-de Sitter model is the only reasonable possibility. This was driven by an idea, cosmological inflation (Guth 1981). ‘The main point of inflation for our purpose is the concept that the universe ‘might have evolved through an epoch of rapid expansion that would have stretched the length scales ofthe wrinkles of the primeval ooze into values much larger than those we can explore: the universe may be disorganized and without form inthe large yet very close to homogeneous and isotropicin he bit we can see. The stretch ing that ironed out the wrinkles would have ironed out any mean space curvature 400, o in this picture the curvature term in equation (1.8) forthe present value of #7 is negligibly smal. ‘This leaves the cosmological constant A, aterm for which the particle physics ‘community demonstrates a love-hate attivude, In the modern and successful the- cries for the weak, electromagnetic, and strong interactions there very naturally spear a set of contributions to the stress-energy tensor that act like a time-variable ‘A. This was one ofthe element that led Guth to inflation: he postulated that a large effective A from particle physics drives the rapid expansion during inflation. The ‘decay of this .erm into ordinary matter and radiation would end inflation and begin evolution according to the standard cosmological model. A residual nonzero A could be leftover afer inflation, and present now, but the expected value of a A term coming out of standard particle theory is ridiculously large compared to what is acceptable for cosmotogy. The naturat presumption is that the cosmological ‘constant has settled down to the only reasonable and observationally acceptable value, A = 0, It would follow from allthis that inflation predicts the Einstein- de Sitter model. During the past decade many people, including respected and thoughtful ob- servational astronomers, were led to conclude that the observable part of the uni- verse likely is well described by the Einstein-de Sitter model. Thave mentioned the Unree main drivers: the Dicke coincidences argument, the belief that an astronom- ically interesting value for a cosmological cosmological is not likely to come out ‘of pantcle physics, and the inflation scenario forthe early universe, All three are ‘worth serious consideration, and inflation in particular has been very influential in the development of ideas of what our universe might have been like before it was,‘pg. THE STATE OF THE MEASUREMENTS n ying. Theory was the driver, however: there was litle observational evi- ence forthe Finstein-de Sitter case and a nontrivial case against it.I can say this Gerbout excessive resort to hindsight because a nftation was becoming popular iny reading ofthe evidence was leading me to abandon my earlier enthusiasm for the Einstein-de Siter model: it seemed difficult to see where the large mass re- _uired by this model might be located (for the reasons discussed in Peebles 1986) ‘As described in the next section, where some of the observational arguments pro snd con © = 1 are discussed, I have seen no reason to change my mind, but the case certainly isnot yet closed. ‘The observational pressures on the Einstein-de Sitter model have led people swexplore alternatives. Inflation’s original explanation for the near homogeneity ofthe observable space requires that the curvature term R-? be negligibly small, ‘utitcertainly would allow a nonzero A. N. Bahcall and Ostriker describe in these ‘proceedings the benefits this parameter offers in imterpretingthe observations. An- ather possibility is inflation in an open universe (Ratra and Peebles 1994; Bucher, Goldhaber, & Turok 1995), Here we need another explanation for homogeneity; the best bet seems to be Got’s (1982) picture for the growth of an open universe out of an event in a de Siter spacetime, ‘My impression is that research om inflation is ina healthy state: the science {sbeing driven by advances in the observational evidence that ae leading people to consider new ideas. It is too soon to decide whether the rch flow of ideas on how the world begins and ends within inflation and other scenarios is leading us, toward a deeper understanding of physical reality: perhaps that will depend on the ‘outcome of the measurements of the cosmological parameters. 1.3. THE STATE OF THE MEASUREMENTS ‘he fist thng to understands that we have no measurement presently capable of ‘warbiguousl distinguishing a Big Crunch from a Big Chill. We do have some Promising lines of evidence, however, and the reasonable hope that research in Progress wil show ushow thisevidence canbe tied together ina concordance tight enough tobe believable. The situations ilustated in Table 1.1. The density parameter is ue fac- tional contribution ofthe mass density tothe present value of the expansion rate ‘in equation (1.8): axGp eee (1.16) Tae casein the second column ofthe abe, labeled @ = 1. is meant 0 be the Binstcin-e iter model. Its posible thatthe curvature and terms cancel each Other athe presen epoch, leaving = 1 with nonzero values of R=? and A, but2 (CHAPTER 1. THE COSMOLOGICAL PARAMETERS ‘mil we are driven from tits sensible to operate under the assumption tbe Nature ‘would not have been so unkind. By the same hopeful reasoning te third column assumes there are just two significant terms inthe expansion rate equation, the ‘dominant one being space curvature or else a cosmological constant, with almost all the rest in the mass density. ‘The mark x/ means thee is significant evidence in favor ofthe case, the mark LX significantevidence against, Lhave atempte to give some indcationof the de- ae of significance by assigning question marks to the more debatable evidence. ‘Table 1.1: Scorecard 1995 Observation Q=1 AvO1 ‘Dynamics and biasing on cales SSMpe XV Dynamics on scales 2 10 Mpe vy Expansion time Hoto 1 Radial and angular size distances xX? oy? Plasma mass fraction in clusters x oy Models for structure formation van ‘The first entry in the table is based on the dynamical studies of groups and the contral parts of clusters of galaxies, on scales less than a few megaparsecs. ‘The derived mass per galaxy multiplied by the mesn galaxy number density yields the contribution to the mean mass density from the material that is concentrated around galaxies; the result is equivalent to the density parameter o1gesor ain ‘Another way to put i is that the small-scale dynamical measurements of the mass per galaxy agree with the mass found in the dark halos of spiral galaxies within radi Taato ~ 300 kpe. 1g) ‘This is discussed further in these proceedings by N. Bahcall “The measurement in equation (1.17) seems to be secure; the open issue is ‘whether there might be a good deal more mass in a more broadly distributed com- pponest. The galaxy masses derived from the relative gravitational accelerations in samples of close pairs of galaxies are measures of the mass Concentrated around13. THE STATE OF THE MEASUREMENTS 3 the pairs on scales less than or comparable to the separation, and miss mass in ‘common more extended envelopes. This isthe mass biasing picture, meaning the aries are not fair measures of where the mass is. T think iis prety clearly established that mass biasing is required if © = 1, ‘The biasing concept is most commonly discussed in terms of the cold dark amatier (CDM) cosmogony. The key idea here is that galaxies form alte peaks of primeval initially Gaussian mass density fluctuations, and that the familiar giant spiral and elliptical galaxies form at unusually high peaks that tend to appear in groups and clusters, leaving most ofthe mass on the outskzts ofthese concen- trations. Branchini and CarIberg (1994) present an instructive example of the ef- fect within a CDM cosmogony with © = 1; they find thatthe mass concentrated around pairs of galaxies increases markedly between separations of 1 Mpc and ‘3 Mpe. We have a tes, from samples ofthe radial velocities of the galaxies and ‘groups of galaxies around the Local Group, at distances les than about 4 Mpc. The (CDM prediction is not observed. Rather, a reasonable fit the radial velocities, follows f the mass per galaxy is consistent with equation (1.17) inbalos with sizes given by equation (1.18) (Zartsky etal. 1989; Pecbles 1995). ‘We have another more qualitative but I think very significant test of the bias- ing concept within the CDM model. The cosmogony assumes galaxy formation {is seeded by Gaussian primeval density uctuations, which means that all matter hs beea seeded for galaxy formation, The mass biasing assumption here is that seeds in the mater outside the halos ofthe bright galaxies failed to germinate or lee produced unobtrusive galaxies. Since the failure of germination could hardly be compete it would follow that space outside the concentrations of the bright alates, tr > rpalo, contains dwatt irregular galaxies marked by the low ferlty tthe ground on which their seeds were cast. The effec is seen in numerical N- body studies, of course (an example is given by Brainerd and Villumsen 1992), batitis not observed: known dwarf and irregular galaxies avoid the voids defined by the giant galaxies. offer one example, Absorption lines in quasar spectra reveal gas clouds tht happen to intersect {he lines of sight. For our purpose a useful sample is the clouds with column den- sites © & 1017 neutral hydrogen atoms cm, tree or four orders of magnitude below that typical of che bright pats a spiral galaxy, and at redshifts in the range 05 S z £ 1.5, close enough thet ordinary large galaxies arc observable and deep ‘Bough to give good statistics, The remarkable Bergeron eect is hat with high Probability he clouds at this column density are within about 100 kpe of a humi- ‘Rous galaxy (Steidl, Dickinson, & Persson 1994; Lanzetia etal, 1995). Thais, at “Was tn fo tthe covery atone cas en the gly whone geo bao ia oe | Bieta Sper 160) . =4 CHAPTER 1, THE COSMOLOGICAL PARAMBTERS redshift z ~ 1 this diffuse and generally unobtrusive gas occupies only aboot one part in 10 of space: itis within the dark halo radii rhalo ~ 300 kpe (€9 [1.18)) of the large galaxies. Under the biasing picture most of the mass ofthe universe ‘has to be Outside mato, is seeded for galaxy formation, yet contains insignificant concentrations of neutral gas at column densities 2, 10'7 cm, Is it reasonable 1 think germination could have been so efficiently suppressed? ‘My conclusion is that the biasing concept of the © = 1 CDM model is un- ‘promising. The story is different if @ ~ 0.1, of course, for then we can imagine that at z $1.5 there is litte matter of any kind further than aio from a giant galaxy. ‘The story is different also if = 1 and the dominant mass component at F > Mralo WaS not seeded for galaxies, or the dominant mass component is not capable of clustering on the scale of individual galaxies, We have a tes from the ‘elative peculiar velocities on larger scales, and here there are conflicting indica- tions, Analyses of large-scale peculiar velocities by the method of Bertschinger and Dekel (1989) are consistent with @ = 1 (Dekel etal. 1993; Dekel & Rees 1994), Another approach to the interpretation of the large-scale peculiar velocities using independent data and a difterent method of analysis fails to reproduce the evidence for ahigh mass density however; it indicates? ~ 0.1 (Shaya, Peebles, & Tully 1995). A galaxy-by-galaxy comparison of predictions of peculiar velocities by thetwo camps is under study, and we may hope a resotution ofthe discrepancy is within reach, But meanwhile each side can claim observational support, as in- dicated in the second entry inthe table. ‘The relative contributions of space curvature and the cosmological constant to the expansioa rate equation (1.8) determine the predicted value of the dimen- sonless product Hofo of Hubble's constant and the expansion time from high red- shift to the present, In the Einstein-de Siter model Ht, = 2/3. Lowering @ lowers the predicted gravitational deceleration of the expansion rate and so in- creases the time ack tohighrodshift. A positive A has a larger effect than postive ‘R-? at the same (2, because the former tends to accelerate the expansion rate. If fo = 50 km s“? Mpe“!, close tothe smallest value currently under discussion, the Einstein-de Sitter model predicts ¢. = 13 Gyr, about the smallest value under discussion for stellar evolution ages. If, a8 many but not all expert ae arguing, HE isnot likely tobe less than about 60 in the shove units c.g. Freedman 1994), the expansion time in an Einstein-de Sitter mode! would seem to be too small. I hhave accordingly entered the result ofthe timescale test as tentative bad news for ‘this model, good news for the low density cases. ‘Most of the current work on the extragalactic distance scale follows Hubble's path: use the inverse square law to estimate distances by the comperison of the1.3. THE STATE OF THE MrasuREMENTS 15 \ brightnesses and intrinsic luminosities of stars that are luminous enough tobe observationally interesting at interesting distances, and maybe also periodic “rexploding. New directions have been found. One involves the measurement of| fine delays inthe gravitational lensing events Blandford discusses in these pro- ceedings. Another uses the physics of plasmas in clusters of palaxes, as discussed recently by Herbig etal. (1995), Yet another uses the remarkably precise measure- seats ofthe velocities and accelerations of massing interstellar clouds close to the nuclei of galaxies (Miyoshict al. 1995). Perhaps out ofall ofthis we may hope to know H, to 10% by the turn of the millennium, ‘The angular size distance (eq. 1.11) enters the analysis of other cosmological tests, including galaxy counts, the rate of lensing of quasar images by the mass concentrations in foreground galaxies (c.g., Fukugita & Turner 1991), and the magnitude-redshiftrelation applied to distant supernovae (e.g, Goober & Perl- mutter 1995). ‘Arrecent example of the fist of these tests is the deep K-band (2.2 micron) counts of Djorgovski et al. (1995). Many experts advise against interpretation ‘until the counts a6 a function of apparent magnisude are better established and ‘we have a better understanding of the time evolution of galaxy luminosities and colors, but we can take note of the following indications. If the giant galaxies that tend to dominate a sample selected by apparent magnitude are evolving only through te stella evolution of populations formed at igh redshifts, the predicted ‘counts a the faint end of the Djorgovski etal. sample differ by nearly an order of ‘magnitude inthe high and low density cases with A = 0. Thats, the observations seem to be reaching redshifts where the effects ofthe cosmology are large. The effets of galaxy evolution canbe large too, but we do ave checks. For example, the galaxies seleced by the Bergeron effect at redshift: ~ | show litle evolution ‘from the present (Steidel, Dickinson, & Persson 1994). If evolution in galaxies selected in the K-band is modest then the counts favor low 2, with A = 0, and Thave accordingly entered this result as tentative good news for the low density case, ‘The gravitational lensing of quasar images by foreground galaxies also sam- les large redshifts, and again the effect of the cosmological parameters on the Predicted lensing rate is large (Fukugita & Turner 1991). The analysis by Maoz And Rix (1993) indicates that, if space curvature vanishes, 2 is no less than about 3. An interesting probiem inthis analysis is that it essumes elliptical galaxies ‘have massive dark halos, in analogy with the dark halos needed in Newtonian me- ‘hans to account forthe rotation curves in spiral galaxies. Massive dark halos are eeded 1 gravitationally contain the pools of plasia observed around giant ell Aicals in clusters of galaxies. Atissue here are massive halos around the more nu-16 (CHAPTER 1, THR COSMOLOGICAL PARAMETERS ‘merous less luminous ellipticals that would be responsible for lensing cases where the angular separations of the lensed images is less than about one second of arc, ‘Tests of the mass distributions in such ellipticals based on the distribution and mo- tons ofthe stars are difficult to apply, and the constraints on massive halos in these galaxies are still subject 10 debate. There is one case, the galaxy M 105, where a detailed study ofthe motions of planetary nebulae (Ciardullo, Jacoby, & Dejonghe 1993) and of a conveniently in placed ring of atomic hydrogen (Schneider 1991) yield a well-motivated mass model that requires little dark matter. If this were a ‘common situation among the less luminous ellipticals, it would reduce the bound ond. ‘The fifth entry inthe table isan elegant new test based on the standard model {or the origin of the light elements and the measurements of the baryonic mass fraction in clusters of galaxies (White et al. 1993; White & Fabian 1995). A rea- sonable fit to the observed light element abundances follows out ofthe computed production of elements as the young universe expands and cools, if the density Parameter in baryons isp ~ 0.01h7%, where Ho = 100h kms! Mpc~! (Walker etal. 1991). If the Hubble parameter is no less than about & = 0.5 and Q = 1 this would say the cosmic baryonic mass fraction is no more than about 0.05, the rest being some kind of nonbaryonic dark matter. In the central parts ‘of rch clusters the mass fraction in observable baryons — stars and intracluster plasma —exceeds this number if 2 0.5. Numerical studies of cluster formation indicate thatthe baryons are tot likely to have been able to settle relative to the dark matter. The straightforward interpretation is thatthe cosmic baryonic mass fraction is significantly Larger than 0.05 because the mean mass density i ess than the Einstein-e Sitter value ‘The last entry inthe table refers to theories of the origin of structure: inter- galactic gas clouds, galaxies, and the Large-scale galaxy distribution, The theories 4 they would have tended to gravitationally accrete the intergalactic matter, removing the Lya forest. Because the forest clearly is, resent at z = 4 this line of argument suggests galaxies were not assembled much ‘arlier. A counter argument uses te abundant evidence of mass exchange between ‘resent-day galaxies and the intergalactic medium (e.g., Irwin 1995). Couldit be ‘that galactic winds in gas-rich young galaxies born at 2 > 4 fed the intergalactic ‘medium, accounting forthe presence of the Lymana forest without overly perturb- {ng the CBR or overly polluting the forest clouds with heavy elements? ‘Which formed first, galaxies or clusters of galaxies? Hubble (1936) argued forthe lauer Lemaitre (1934) argued that galaxies and clusters of galaxies may ‘have been produced by the gravitational instability of the expanding universe. The em radhcnd by the ravi SA good sample ofthe debates Fenk tal (1989) 8 argumcnt is manly of historical Jaceest. Following Jean, Hubble considered the Hea tha galaiesevlveist spas. Since sisal prefer less dense sutroundiogs ta lpia, ht ‘aponed the field gulaies might have evelved out of elipical tht excaped20 ‘CHAPTER 1. THE COSMOLOGICAL PARAMETERS density j of a newly virialized mass concentration would be some characteristic multiple ofthe cosmic mean (p) atthe time of virialization. Since galaxies are a {ood deal denser than clusters it would follow that galaxies formed a good desl catlie. I consider the Local Group of galaxies a good example of Lemaitre's pic- ture: the evidence is that this system is forming now by the rmitual gravitational attraction of old galaxies” ‘An important number in Lemaftre’s pictur is the ratio f/(p) of the mean den- sity withina newly virilized system and the cosmic mean density at this time. The spherical collapse model indicates 6/{p) ~ 200 (eg., Efstathiou & Rees 1988) Tntich clusters of galaxies the ratio ofthe galaxy number density within the Abell radius ra = 1.5h1 Mpc to the present cosmic mean is i/(n) ~ 200, and the mass contrast is similar. Does this mean the great clusters are being assembled ‘now at the Abell radius? T have trouble believing it, because this seems to be an ‘extreme application of spherical symmetry, and because the Local Group seems {o be a counterexample, having been assembled and viralized at a much smaller density contrast. We may hope to see a key test in the reasonably near future: the analysis of the typical mass distribution within clusters at redshift z ~ 1. The spherical model predicts the mean density at r4 = 1.5"! Mpc and = ~ 1 is well below the value observed in low redshift clusters. Ifthe density collapse factor st formation isan order of magnitude less than the spherical prediction then the densities at r4 = 1.54" Mpe will about the same at =~ 1 and z ~ 0. ‘The collapse factor issue applies also tothe epoch of gravitational assembly of the galaxies, The dark halo in a spiral galaxy such as the Milky Way dominates the mass at radius r ~ 10 kpc, where the circular velocity of rotation is v ~ 200 am s~*, The enclosed mass is M = v?r/G, and the ratio of the enclosed mean ‘mass density p tothe large-scale mean (p) evaluated at redshift = is BD 2 vy 10° Grae Ge) ~ oa a» ‘If the mass concentration were assembled at z = 2 the density contrast at assembly ‘would have been j/(p} ~ 3000/0, which requires a collapse factor even larger than that of the spherical model. This is why the very old discussions emphasized galaxy asscmbly at z 2 10 (Partridge & Pecbles 1967). In the past decade many fave avored moc recut formation bute end of een iacusons at Sng ‘back to higher redshifts. We must expect still more swings of opinion, but obser- vation bak os ~ 8 show usa lou sbout whet glist were ding a Perhaps ou oti wl eome evidence tha wil ake believes Or al "ala and ferences Peas (195.‘pwus00eArnic NOTES a AckNOWLEDaMENTS ‘This paper was improved by helpful comments (rom Jonn Bahcall, Karl Fishes, ‘Wayne Hu, and David Spergel, The work was supported in part at Princeton Uni- versity by the National Scicnce Foundation. BIBLIOGRAPHIC NOTES « Dicke, R. H. 1970, Gravitation and the Universe (Philadelphia: Ameri- ‘can Philosophical Society). Three lectures on gravitational physics by one ‘of the master physicists and astrophysicists. The third lecture isan illami- nating account of the discovery of the cosmic microwave background raui- ation. ‘¢ Hubble, F. 1936, The Realm of the Nebulae (New Haven: Yale Univer- ‘sity Press). A thillingand charming account of early cosmological research, ‘¢ Kolb, E, W.and Turner, M.S, 1990, The Farly Universe (Redwood City: Addison-Wesley). A standard compendium ofthe physics of the bebaviour of the universe at very high redshift ‘¢ Ohanian, H. C, and Ruffini, R, 1994, Gravitation and Spacetime (New York: Norton). There are many textbooks on general relativity theory; in my experience this presents the subject al about the right level, ‘ Peebles, P, J. F. 1980, The Large-Scale Structure of the Universe; 1993, Principles of Physical Cosmology (Princeton: Princeton University Press). Those monographs provide introductions and standard treatments of the subjects discussed in this lecture, BrptiocRapiy 10] Banca, J.N,, & Spitzer, L. 1969, Apl, 156, L63. 2} Bergeron, J. 1986, A&A, 155, L8. ) Bertschinger, E, & Dekel, A. 1989, Apl, 336, LS. [6] Bondi, H. 1960, Cosmology, 2nd edition (Cambridge: Cambridge Univ, Press), 166, (5) Brainerd, T. G., & Villumsen, J. V. 1992, Apl, 394, 409, $6) Branchini, E., & Cariberg, R. G. 1994, Apl, 434, 37,2 CHAPTER 1. THB CosMoLOGICAl. PARAMETERS {7] Butcher, M., Goldhaber, A. S., & Turok, N. 1995, Phys. Rev. B, in press. [8] Ciardullo,R., Jacoby, G, H, & Dejonghe, H, B. 1993, Apt, 414, 454, [9] Dekel, A., et al. 1993, ApJ, 412, 1. [10] Dekel, A., & Rees, M. J. 1994, Apl, 422, L1. [11] Dicke, R. H. 1970, Gravitation and the Universe (Philadelphia: American Philosophical Society). [12] Dicke, R. H., Peebles, P.J. E., Roll, P.G., & Wilkinson, D. T. 1965, ApJ, 142,414, 13} Djorgovski,S., etal, 1995, ApI, 438, L13. [14} Dyson, F. 1979, Rev, Mod, Phys., 51, 447, [15] Efstathiou, G., & Rees, M. J. 1988, MNRAS, 230, SP. [16] Einstein, A., & de Sitter, W, 1932, Proc. NAS, 18, 213. (17) Freedman, W.L,, et al. 1994, Nature, 371, 757, (18} Frenk, C. S., et al. 1989, The Epoch of Galaxy Formation (Dordrecht: Kluwer). (19] Fukugita, M,, & Turner, E. L. 1991, MNRAS, 253, 99. [20] Goobar, A., & Perlmutter, S. 1995, ApJ, 450, 14 [21] Gout, J. R. 1982, Nature, 295, 304, [22] Gow, J. R, Gunn, J. E, Schramm, D.N., & Tinsley, B.M, 1974, Apl, 194, 543. {23] Guth, A. H. 1981, Phys. Rev, D, 23, 347, (24] Herbig, T., Lawrence, C. R., Readhead, A.C. S., 449,L5, Gulkis, S. 1995, ApJ. [25] Hoyle, F, Burbidge, G., & Narlikar, J. V 1993, ApJ, 410, 437, [26] Hubble, E, 1936, Realm of the Nebulae (New Haven: Yale Univ. Press). [271 Trewin, J, 1. 1995, PASE, 107, 715.[pILsOORAPHY 2B 28) Laszett, K.M, Bowen, D. V, Tyler, D, & Webb, J. K. 1995, Apt, 442, 538. {29 Lemaive, G. 1933, Ann, Soc. Sci, Brussels, AS3, 51 {30} Lemafie, G. 1934, Proc. NAS, 20, 12, [B1) Maoz, D., & Rix, HW. 1993, Apl, 416,425, G2} McCarthy, P J. 1993, ARARA, 31, 639, [33] Misner, C. W. 1992, in the 13th International Conference on General Relativity and Gravitation, Univ. Nacional de Cordoba, Argentina, ed, P. W. Lamberti, & O. E, Ortiz, p17. [34] Miyoshi, M., et al. 1995, Nature, 373, 127. [5] Mo, H. J, McGaugh, S. S., & Bothun, G. D. 1994, MNRAS, 167, 129. (36) Oke, J. B. 1984, in Clusters and Groups of Galaxies, ed. F. Mardirossian, et al, (Dordrecht: Reidel), 99, [7] Partridge, R. B., & Peebles, P.J.E. 1967, Apl, 147, 868. (38) Peebles, P. J. E. 1986, Nature, 321, 27. (39) Peebles, P.J. E. 1993, Principles of Physical Cosmology (Princeton: Prince- ton Univ. Press) [40] Peebies, P.J. E, 1995, Apl, 449, 52. [41] Peebies, P,J.E,, Schramm, D.N., Turner, E.L., & Kron, R.G. 1991, Nature, 352, 769, 142] Petrosian, V, Salpeter, EE., & Szckeres, , 1967, Apl, 147, 1222. 143) Ratra, B, & Peebles, P.J.E, 1994, Apy, 432, LS. (44) Robertson, H. P. 1955, PASP, 67, 82, 4S) Satzer, J. 1989, Apl, 347, 152. (46) Schneider, $. E. 1991, in Warped Discs and Inclined Rings Around Galaxies, 4. S. Casertano, PD. Sackett, & F.H. Briggs (Cambridge: Cambridge Univ. Press), 25. (47) Shaya, B. 1, Peebles, P.J.E., & Tully, R. B. 1995, Ap), 454, 15.mw (CHAPTER 1, THE COSMOLOGICAL PARAMETERS [48] Shklovsky, 1. S. 1967, Aste. Tsk, No. 429. [49] Steidel, C. C., Dickinson, M., & Persson, S. E. 1994, ApJ, 437, L75, (50) Walker, TP, et al. 1991, ApJ, 376, 51 [51] White, D. A., & Fabian, A. C. 1995, MNRAS, 273, 73. [52] White, S. D.M., 366,429, lavarro, J. F, Evrard, A. E, & Frenk, C. S. 1993, Nature, [53] Zaritsky, D., Olszewski, E. W, Schommer, R. A, Peterson, RC, & ‘Aaronson, M, A. 1989, Apl, 345, 759.CHAPTER 2 IN THE BEGINNING ... PAUL J. STEINHARDT ‘Department of Physics and Astronomy, Univessity of Pennsylvania, Philadelphia, PA ABSTRACT ‘New observational technologies are transforming cosmology from speculative in- bir to rigorous science. In the next decade, observations will be made which will provide redundant, quantitative tesis of theoretical models for the origin and ‘evolution of large-scale structure in the universe. Using the inflationary theory as am example, this paper discusses some of the crucial tests and how they can be ‘combined to discriminate among different models and measure essential parame- tees, 2.1 ‘THe FUTURE FATE oF CosMoLoGY In the beginning long before there was science, thero was cosmology. AS soon as the first Amelligent eyes were capable of pecring up at the heavens, questions began to arise concerning the origin and evolution of the Universe: How big is the Universe? How old isthe Universe? How did it begin and how did it develop into what we 86 today? Many answers have been proposed over the centuries, yet these remain ‘the fundamental, unsolved problems: of cosmology today (sec the contribution by PILE. Peebles to these proceedings). ‘What has changed is that there is now the teshnology needed 0 est our pro- Posed answers to these questions. The transformation began near the turn ofthis ‘Seatury with the construction of the first, giant, optical telescopes. It was first ‘iscovered thatthe Universo is composed of galaxies and thatthe galaxies are re- 25