The history of modern cosmology has been divided into two periods. The primary discovery phase extended from 1920 to ∼1950 when most of the classical tests were developed. The subsequent consolidation period from 1950 to ∼1980 saw a deeper understanding of the tests and the introduction of concepts of stellar populations that led to estimates of evolutionary trends with look-back time. The only test not affected by evolutionary effects is the comparison of the Hubble time, Ho−1, with independent estimates of the age of the Universe. The N(m), m(z), θ(z), θ(m), and HoTo (qo, Λ) tests are reviewed as they have been discussed in the archive literature. The problem of observational bias is emphasized, as is the difficult recent search for secular evolutionary effects. Current problems which have prospects for solutions in the next decades are set out in the final section.

This review summarizes recent observational work on the cosmic microwave background, or 3 K radiation. Recent measurements of its spectrum and large-scale angular distribution are described, as well as searches for small angular scale fluctuations on arcsecond to degree scales. A few of the consequences of these measurements and upper limits for cosmology, astrophysics, and theories of galaxy formation are touched on here.

Angular fluctuations in the cosmic microwave background radiation are indicators of the seed structure at recombination (z ≅ 1000) which subsequently evolves into the present-day Universe (z = 0). Such measurements provide stringent tests of the cosmological theories of universal evolution and have implications for allowed values of the density and composition of matter in the Universe.

Observations of the interstellar lines of CN near 3875Å toward ζ Oph provide precise measurements of the Microwave Background Radiation (MBR) temperature and determine the intrinsic linewidth necessary for the saturation correction. We report TMBR=2.74±0.05 K at λ=2.64 mm and TMBR=2.75 (+0.24, −0.29) K at λ=1.32 mm.

Sensitive measurements of a dipole anisotropy of the Cosmic Background Radiation at wavelengths of 1.7 and 0.8 mm are presented. Under the assumption that the dipole moment is caused by a doppler shift, these measurements are of sufficient precision to constrain spectral distortions of the CBR at short wavelengths. Measurements of the brightness and shape of diffuse galactic emission are also presented.

A rocket experiment was carried out to search for the extra-galactic background light at 1–5 μm. After subtracting the foreground radiation, there still remains an appreciable amount of isotropic diffuse radiation with a complex spectral feature which is possibly attributed to extragalactic origin.

The problem of determining the fundamental constants of the Universe - Ho, qo, Ω, Λ, hidden matter, and the primordial perturbation parameters is considered. The likelihood of finding new evidence for large-scale Δ T/T-fluctuations is discussed.

Observational data on the Sunyaev-Zel'dovich effect for three clusters are used to obtain limits to the value of the Hubble constant. Only a wide bound on the value of H0, 10 km s−1 Mpc−1 < H0 < 160 km s−1 Mpc−1, can be derived, mostly because of the large uncertainty in the measured temperature of the intracluster medium in each cluster. Better information on this temperature and on the distribution of the intracluster medium are needed for this method to yield a good measurement of H0.

The abundances of the very light elements (D, 3He, 4He and 7Li) constitute indeed one of the most powerful constraints in cosmology: they are known to fix very interesting limits on the baryonic density of the Universe and on the maximum number of neutrino (lepton) families in the frame of the simplest canonical models. Given the importance of these predictions, these models should be analysed very cautiously at the light of recent developments in the observations of these elements. In order to make the simplest models consistent with the observations, it is argued that a thorough destruction of D should occur during the galactic evolution. Moreover this review deals also with some models invoking the possible existence of massive unstable neutrinos, gravitinos or photinos which would decay into high energy photons or of quark nuggets which could be created during the quark-hadron phase transitions. Such models have been designed in an attempt to overcome the limitation on the Universe density coming from these abundance determinations. Although the simple canonical models are especially attractive such models cannot be disregarded.

We have observed 23 halo stars with space velocities and metallicities [Fe/H] ≤ −0.6. Twelve of these 23 show the more extreme properties and [Fe/H] ≤ −1.4 and should therefore constitute an especially old, homogeneous subgroup. The principal results for these 12 extreme halo stars and 5 similar stars observed in previous studies are that (1) a single, well defined relation, previously discovered and discussed by Spite and Spite, exists without exception between the atmospheric Li/H ratio and Te, and (2) at Te ≥ 5600 K the average lithium abundance is <Li/H> = 1.2 ± 0.3 × 10−10. The latter value constitutes a lower limit on the 7Li fraction produced in primordial nucleosynthesis and thereby significantly constrains the cosmic ratio of baryons to photons.

We present visual and near-infrared Hubble diagrams for first-ranked cluster galaxies, and moderate-flux and high-flux radio galaxies. The photometric improvements and the extension of the diagrams to large redshifts (up to z ≃ 1.8) for both 3CR and “1 Jy” class radio galaxies are highlighted. The absolute luminosities of these three types of “standard candles”, where their redshifts overlap, agree adequately. The near-IR (2μm) Hubble diagrams may be used to determine the global deceleration parameter, q0, and the current data favor q0 ∼ +0.2 to +0.3. The sensitivity to evolutionary changes in the giant galaxies is quite modest at these long wavelengths. On the other hand, the visual regime (BV R) shows a dramatic dependence on differential evolution effects, which dominate over the cosmological model differences. Most radio galaxies apparently have had active star formation, with a continually declining rate, but a few are quite faint and red; they may have had a relatively passive evolution.

The infrared photometric study of a sample of 90 3CR radio galaxies by Lilly & Longair (1984, hereafter LL) has demonstrated that the high redshift objects are brighter in the infrared than their low redshift counterparts; this has been interpreted as being entirely due to the evolution of their constituent stellar populations. There is however a great difference between the radio luminosities of the high and low redshift objects in this flux limited sample and we have therefore examined statistically the possibility of a correlation between the infrared and radio luminosities of these galaxies, the presence of which could bias our interpretation of the infrared Hubble diagram. We find that the radio and infrared luminosities do indeed correlate for the most powerful radio galaxies.

It is found that the Hubble diagrams for quasars and giant radio galaxies are very similar. Either the evolution of giant branch stars in the galaxies closely matches the evolution of the quasar nonthermal emission or the cosmological deceleration parameter qo is close to 3.

The zero-point of the extragalactic distance scale, defined by about two dozens of nearby, late-type galaxies, has remained nearly unchanged for the last decade, in spite of the advent of new techniques and great efforts. The distances are essentially tied to trigonometric parallax stars and hence independent of the Hyades modulus; they are consistent with RR Lyr stars. The mean zero-point is therefore probably secure to better than 10%.

All known secondary distance indicators are still affected by zero-point errors, by problems in the definition of their relation between distance indicator and absolute magnitude (or linear size), and/or by selection bias. The effect of the very important selection bias (Malmquist effect), which causes a seemingly non-linear expansion field, is illustrated by two examples. To test for any true deviations from a linear expansion the Hubble diagram of nearly bias-free first-ranked cluster galaxies and supernovae Ia is shown; this imposes stringent limits on any non-linearity of the Hubble flow within v<5000 km s−1.

After freeing the available distances of field galaxies from selection bias and after reducing them to a common zero-point, one finds HO=55–65. Several distance indicators require a best Virgo cluster modulus of (m-M)=31.60, which implies for the Coma cluster (m-M)=35.38 and, with v(Coma)=7217 km s−1, HO=60. Supernovae Ia and first-ranked cluster galaxies out to large distances give HO (global)=53. Thus the evidence from clusters and field galaxies is best satisfied by HO=55; the assigned mean error of ±7 is to indicate a 3σ range of 35<HO<75.

Purely physical methods to determine extragalactic distances have modest weight yet; they will contribute eventually much to the determination of HO.

If HO were as large as 100, several paradoxa would arise. The Milky Way would have a very high supernova frequency, our Galaxy and M31 would be oversized, the baryon density would fall short to bind clusters, and Friedman universes were excluded.

Because all systematic errors have conspired and probably still conspire to measure HO too high, the true value could well be 40. Until new, decisive evidence becomes available, it is suggested for all practical purposes to use HO=50.

A summary is given of recent work using the IR/H I method, arguably the best global distance indicator presently available. Reflex motion toward the microwave dipole anisotropy has been seen relative to a sample of nearby galaxy clusters; this result is contrasted with the somewhat divergent conclusions obtained by Burstein et al. from a similar study of elliptical galaxies. A best guess calibration of the IR/H I zero point continues to lead to a high value for the expansion rate. This finding cannot be explained by appealing to Malmquist bias, as demonstrated by a straightforward linearity test of the velocity-distance relation. However, all current estimates of the Hubble constant are plagued by the large uncertainties to the distances of nearby calibrating galaxies, a problem whose full solution probably must await the launch of the Hubble Space Telescope.

We have used the CFH 3.6 m telescope in excellent seeing to observe RR Lyrae variables in a halo field in M31. From 32 variables found, we obtain (m - M)0 = 24.34 ± 0.15 for M31.

When the Bottinelli et al. version of the Tully-Fisher relation is applied to derive distances, and when the observed parameters are used to predict the Malmquist bias at a given distance, the observed variation of the Hubble ratio as a function of kinematic distance is about 2.5 times the predicted variation.