Introduction

The title not withstanding, this is not a history of dark matter. Until we know what the dark matter is, we cannot know its history. Instead, this is a brief history of how astronomers converged to the view that most of the matter in the universe is dark. This paper deals principally with the early studies which helped to answer the questions “Are rotation curves flat? If so, why?” It also includes some early history in deciphering the signature of clusters of galaxies as gravitational lenses, which seems to have been little investigated. This account covers the years up to 1980; achievements since 1980 are science, not history. Several excellent, informative brief histories exist, and interested readers should see Trimble (1987, 1995) and van den Bergh (1999). We can all thank Sidney van den Bergh for correctly translating Zwicky's “dunkle (kalte) materie” as “dark (cold) matter” and finally putting to rest the myth that Zwicky called it “missing matter.”

The notion that there are stars that are dark was a common one in the 18th and 19th Century. Walt Whitman's (1855) lines in Leaves of Grass, “The bright suns I see and the dark suns I cannot see are in their place” and Bessel's “Foundation of an Astronomy of the Invisible” (Clerke 1885 and reference therein) are early manifestations of this belief.

The primordial abundances of deuterium, helium, and lithium probe the baryon density of the universe only a few minutes after the Big Bang. Of these relics from the early universe, deuterium is the baryometer of choice. After reviewing the current observational status (a moving target!), the BBN baryon density is derived and compared to independent estimates of the baryon density several hundred thousand years after the Big Bang (as inferred from CMB observations) and at present, more than 10 billion years later. The excellent agreement among these values represents an impressive confirmation of the standard model of cosmology, justifying—indeed, demanding—more detailed quantitative scrutiny. To this end, the corresponding BBN-predicted abundances of helium and lithium are compared with observations to further test and constrain the standard, hot, big bang cosmological model.

Introduction

As progress is made towards a new, precision era of cosmology, redundancy will play an increasingly important role. As cosmology is an observational science, it will be crucial to avail ourselves of multiple, independent tests of, and constraints on, competing cosmological models and their parameters. Furthermore, such redundancy may provide the only window on systematic errors which can impede our progress or send us off in unprofitable directions.

We present photometric observations of an apparent Type Ia supernova (SN Ia) at a redshift of ∼1.7, the farthest SN observed to date. The supernova, SN1997ff, was discovered in a repeat observation by the Hubble Space Telescope (HST) of the Hubble Deep Field-North (HDF-N), and serendipitously monitored with NICMOS on HST throughout the Thompson et al. GTO campaign. The SN type can be determined from the host galaxy type: an evolved, red elliptical lacking enough recent star formation to provide a significant population of core-collapse supernovae. The classification is further supported by diagnostics available from the observed colors and temporal behavior of the SN, both of which match a typical SN Ia. The photometric record of the SN includes a dozen flux measurements in the I, J, and H bands spanning 35 days in the observed frame. The redshift derived from the SN photometry, z = 1.7±0.1, is in excellent agreement with the redshift estimate of z = 1.65 ± 0.15 derived from the U300B450V606I814J110J125H160H165Ks photometry of the galaxy. Optical and near-infrared spectra of the host provide a very tentative spectroscopic redshift of 1.755. Fits to observations of the SN provide constraints for the redshift-distance relation of SNe Ia and a powerful test of the current accelerating Universe hypothesis. The apparent SN brightness is consistent with that expected in the decelerating phase of the preferred cosmological model, ΩM ≈ 1/3, ΩΛ ≈ 2/3.

The planet Uranus was discovered in 1781 by the British astronomer William Herschel. Not long after its discovery, astronomers charting the orbit of Uranus found small discrepancies between the predicted and observed positions of the planet. In September 1845, British astronomer John Adams proved mathematically that the deviations in Uranus' orbit could not result merely from the gravitational pull of the other known planets and he predicted the existence of another, previously undetected planet in the solar system. The eventual discovery of the planet Neptune in September 1846 by the German astronomer Johann Galle thus marked the first detection of astronomical “dark matter” whose presence was first deduced by its gravitational effects. However, in the history of physics, we also find a case in which the assumption about the existence of an unseen medium was later proven to be totally wrong. Until 1887, physicists assumed that aether—a substance that pervades all space—was a necessary medium for the propagation of light. A famous experiment by American researchers Albert Michelson and Howard Morley not only showed unambiguously that this medium does not exist, but the experimental results also set Einstein on the road to a new theory of space and time—special relativity.

Astrophysicists today are faced with a similar “Neptune vs. aether” dilemma.

A straightforward interpretation of the MACHO microlensing results in the direction of the Magellanic Clouds suggests that an important fraction of the baryonic dark matter component of our Galaxy is in the form of old white dwarfs. If correct, this has serious implications for the early generations of stars that formed in the Universe and also on the manner in which galaxies formed and enriched themselves in heavy elements. I examine this scenario in some detail and in particular explore whether the searches currently being carried out to locate local examples of these MACHOs can shed any light at all on this scenario.

Introduction

A conservative estimate of the mass of the Galaxy out to a distance of about 2/3 of that of the Large Magellanic Cloud is MG = 4 × 1011 M⊙ (Fich & Tremaine 1991). With a total luminosity in the V-band of 1.4 × 1010 L⊙ (Binney & Tremaine 1987) the Galactic mass to light ratio in V (M/Lv) out to 35 kpc is ∼ 30. Since normal stellar populations do not generally produce M/Lv ratios higher than about 3, this is usually taken as evidence for an important component of dark matter within an extended halo surrounding the Galaxy.

FRIIb radio galaxies provide a tool to determine the coordinate distance to sources at redshifts from zero to two. The coordinate distance depends on the present values of global cosmological parameters, quintessence, and the equation of state of quintessence. The coordinate distance provides one of the cleanest determinations of global cosmological parameters because it does not depend on the clustering properties of any of the mass-energy components present in the universe.

Two complementary methods that provide direct determinations of the coordinate distance to sources with redshifts out to one or two are the modified standard yardstick method utilizing FRIIb radio galaxies, and the modified standard candle method utilizing type Ia supernovae. These two methods are compared here, and are found to be complementary in many ways. The two methods do differ in some regards; perhaps the most significant difference is that the radio galaxy method is completely independent of the local distance scale and independent of the properties of local sources, while the supernovae method is very closely tied to the local distance scale and the properties of local sources.

FRIIb radio galaxies provide one of the very few reliable probes of the coordinate distance to sources with redshifts out to two. This method indicates that the current value of the density parameter in non-relativistic matter, Ωm, must be low, irrespective of whether the universe is spatially flat, and of whether a significant cosmological constant or quintessence pervades the universe at the present epoch.

X-ray clusters provide excellent constraints on cosmological parameters such as ΩM. I will describe measurements of cluster masses and of cluster evolution. The cluster baryon fraction and the evolution of the cluster temperature function strongly constrain the mean density of matter in the universe (ΩM). The constraints are consistent with ΩM = 0.2–0.5, with best fit values of ΩM = 0.3–0.4. The systematic uncertainties are of the same size as the statistical uncertainties, even with the small number of clusters in our current temperature surveys (ΔΩM ∼ 0.1.) Thus, reduction of the uncertainties in these methods requires not only an increased number of hot massive clusters in a given sample but much better quantification of the systematics, a goal which demands not only more clusters but clusters with a range of properties and redshifts. The current constraints are not particularly sensitive to the particular form or value of the acceleration parameter Λ and therefore these constraints provide an limit on cosmological parameters complementary to the limits imposed by the cosmic microwave background studies and by the Type Ia supernovae at cosmological distances.

Introduction

I seek to make the following three points in this review:

1. (a) Clusters of galaxies are excellent targets for cosmological studies.

2. (b) Existing studies have already placed very strong constraints on the mean density of matter in the universe.

3. (c) These constraints are nearly orthogonal to constraints from the cosmic microwave background and type Ia supernovae.

4. […]

One of the most fundamental questions in cosmology is: How much matter is there in the Universe and how is it distributed? Here I review several independent measures—including those utilizing clusters of galaxies—that show that the mass-density of the Universe is only ∼ 20% of the critical density. Recent measurements of the mass-to-light function—from galaxies, to groups, clusters, and superclusters—provide a powerful new measure of the universal density. The results reveal a low density of 0.16 ± 0.05 the critical density. The observations suggest that, on average, the mass distribution follows the light distribution on large scales. The results, combined with recent observations of high redshift supernovae and the spectrum of the CMB anisotropy, suggest a Universe that has low density (Ωm ≃ 0.2), is flat, and is dominated by dark energy.

Introduction

Theoretical arguments based on standard models of inflation, as well as on the demand of no “fine tuning” of cosmological parameters, predict a flat universe with the critical density needed to just halt its expansion (1.9 × 10–29 h2 g cm–3). Observations, however, reveal only a small fraction of the critical density, even when all the unseen dark matter in galaxy halos and clusters of galaxies is included. There is no reliable indication that the matter needed to close the universe does in fact exist. Here I review several independent observations of clusters of galaxies which indicate, independently, that the mass density of the universe is sub-critical.

Milgrom has proposed that the appearance of discrepancies between the Newtonian dynamical mass and the directly observable mass in astronomical systems could be due to a breakdown of Newtonian dynamics in the limit of low accelerations rather than the presence of unseen matter. Milgrom's hypothesis, modified Newtonian dynamics or MOND, has been remarkably successful in explaining systematic properties of spiral and elliptical galaxies and predicting in detail the observed rotation curves of spiral galaxies with only one additional parameter—a critical acceleration which is on the order of the cosmologically interesting value of CH〪. Here I review the empirical successes of this idea and discuss its possible extention to cosmology and structure formation.

Introduction

Modified Newtonian dynamics (MOND) is an ad hoc modification of Newton's law of gravity or inertia proposed by Milgrom (1983) as an alternative to cosmic dark matter. The motivation for this and other such proposals is obvious: So long as the only evidence for dark matter is its global gravitational effect, then its presumed exitance is not independent of the assumed form of the law of gravity or inertia on astronomical scales. In other words, either the universe contains large quantities of unseen matter, or gravity (or the response of particles to gravity) is not generally the same as it appears to be in the solar system.

A close scrutiny of the microlensing results towards the Magellanic clouds reveals that the stars within the Magellanic clouds are major contributors as lenses, and the contribution of MACHOs to dark matter is 0 to 5%. The principal results which lead to this conclusion are the following:

1. (i) Out of the ∼17 events detected so far towards the Magellanic Clouds, the lens location has been securely determined for one binary-lens event through its caustic-crossing timescale. In this case, the lens was found to be within the Magellanic Clouds. Although less certain, lens locations have been determined for three other events and in each of these three events, the lens is most likely within the Magellanic clouds.

2. (ii) If most of the lenses are MACHOs in the Galactic halo, the timescales would imply that the MACHOs in the direction of the LMC have masses of the order of 0.5 M⊙, and the MACHOs in the direction of the SMC have masses of the order of 2 to 3 M⊙. This is inconsistent with even the most flattened model of the Galaxy. If, on the other hand, they are caused by stars within the Magellanic Clouds, the masses of the stars are of the order of 0.2 M⊙ for both the LMC as well as the SMC.

3. (iii) If 50% of the lenses are in binary systems similar to the stars in the solar neighborhood, ∼10% of the events are expected to show binary characteristics.

4. […]

There are now two cosmological constant problems: (i) why the vacuum energy is so small and (ii) why it comes to dominate at about the epoch of galaxy formation. Anthropic selection appears to be the only approach that can naturally resolve both problems. This approach presents some challenges to particle physics models.

The problems

Until recently, there was only one cosmological constant problem and hardly any solutions. Now, within the scope of a few years, we have made progress on both accounts. We now have two cosmological constant problems (CCPs) and a number of proposed solutions. In this talk I am going to review the situation, focusing mainly on the anthropic approach and on its implications for particle physics models. I realize that the anthropic approach has a low approval rating among physicists. But I think its bad reputation is largely undeserved. When properly used, this approach is quantitative and has no mystical overtones that are often attributed to it. Moreover, at present this appears to be the only approach that can solve both CCPs. I will also comment on other approaches to the problems.

The cosmological constant is (up to a factor) the vacuum energy density, ρv.

For physicists, recent developments in astrophysics and cosmology present exciting challenges. We are conducting “experiments” in energy regimes some of which will be probed by accelerators in the near future, and others which are inevitably the subject of more speculative theoretical investigations. Dark matter is an area where we have hope of making discoveries both with accelerator experiments and dedicated searches. Inflation and dark energy lie in regimes where presently our only hope for a fundamental understanding lies in string theory.

Introduction

It is a truism that the development of astronomy, astrophysics, cosmology relies on our understanding of the relevant laws of physics. It is thus no surprise that my astronomy colleagues tend to know more classical mechanics, electricity and magnetism, atomic and nuclear physics than my colleagues in particle theory.

As we consider many of the questions which we now face in cosmology, we must confront the fact that we simply do not know the relevant laws of nature. The public often asks us “What came before the Big Bang?” We usually think of this as requiring understanding of physics at the Planck scale. But at present we can't even come close. Ignorance sets in slightly above nucleosynthesis, and becomes severe by the time we reach the weak scale. Some of the questions which trouble us will be settled by experiment over the next decades; some require new theoretical developments. Needless to say, it is possible that much will remain obscure for a long time.

Recent measurements of temperature fluctuations in the cosmic microwave background (CMB) indicate that the Universe is flat and that large-scale structure grew via gravitational infall from primordial adiabatic perturbations. Both of these observations seem to indicate that we are on the right track with inflation. But what is the new physics responsible for inflation? This question can be answered with observations of the polarization of the CMB. Inflation predicts robustly the existence of a stochastic background of cosmological gravitational waves with an amplitude proportional to the square of the energy scale of inflation. This gravitational-wave background induces a unique signature in the polarization of the CMB. If inflation took place at an energy scale much smaller than that of grand unification, then the signal will be too small to be detectable. However, if inflation had something to do with grand unification or Planckscale physics, then the signal is conceivably detectable in the optimistic case by the Planck satellite, or if not, then by a dedicated post-Planck CMB polarization experiment. Realistic developments in detector technology as well as a proper scan strategy could produce such a post-Planck experiment that would improve on Planck's sensitivity to the gravitational-wave background by several orders of magnitude in a decade timescale.