The Universe is not completely homogeneous. Even if it is sufficiently so on large scales, it is very inhomogeneous at small scales, and this has an effect on light propagation, so that the distance as a function of redshift, which in many cases is defined via light propagation, can differ from the homogeneous case. Simple models can take this into account. I review the history of this idea, its generalization to a wide variety of cosmological models, analytic solutions of simple models, comparison of such solutions with exact solutions and numerical simulations, applications, simpler analytic approximations to the distance equations, and (for all of these aspects) the related concept of a "Swiss-cheese" universe.

The $m$-$z$ relation for type Ia supernovae is one of the key pieces of evidence supporting the cosmological `concordance model' with $\lambda_0 \approx 0.7$ and $\Omega_0 \approx 0.3$. However, it is well known that the $m$-$z$ relation depends not only on $\lambda_0$ and $\Omega_0$ (with $H_0$ as a scale factor) but also on the density of matter along the line of sight, which is not necessarily the same as the large-scale density. I investigate to what extent the measurement of $\lambda_0$ and $\Omega_0$ depends on this density when it is characterized by the parameter $\eta$ ($0 \le \eta \le 1$), which describes the ratio of density along the line of sight to the overall density. I also discuss what constraints can be placed on $\eta$, both with and without constraints on $\lambda_0$ and $\Omega_0$ in addition to those from the $m$-$z$ relation for type~Ia supernovae.

The $m$-$z$ relation for Type Ia supernovae is compatible with the cosmological concordance model if one assumes that the Universe is homogeneous, at least with respect to light propagation. This could be due to the density along each line of sight being equal to the overall cosmological density, or to `safety in numbers', with variation in the density along all lines of sight averaging out if the sample is large enough. Statistical correlations (or lack thereof) between redshifts, residuals (differences between the observed distance moduli and those calculated from the best-fitting cosmological model), and observational uncertainties suggest that the former scenario is the better description, so that one can use the traditional formula for the luminosity distance safely without worry.

I briefly review the flatness problem within the context of classical cosmology and examine some of the debate in the literature with regard to its definition and even the question whether it exists. I then present some new calculations for cosmological models which will collapse in the future; together with previous work by others for models which will expand forever, this allows one to examine the flatness problem quantitatively for all cosmological models. This leads to the conclusion that the flatness problem does not exist, not only for the cosmological models corresponding to the currently popular values of lambda_0 and Omega_0 but indeed for all Friedmann-Lema\^itre models.

We present the selection of the Jodrell Bank Flat-spectrum (JBF) radio source sample, which is designed to reduce the uncertainties in the Cosmic Lens All-Sky Survey (CLASS) gravitational lensing statistics arising from the lack of knowledge about the parent population luminosity function. From observations at 4.86 GHz with the Very Large Array, we have selected a sample of 117 flat-spectrum radio sources with flux densities greater than 5 mJy. These sources were selected in a similar manner to the CLASS complete sample and are therefore representative of the parent population at low flux densities. The vast majority (~90 per cent) of the JBF sample are found to be compact on the arcsecond scales probed here and show little evidence of any extended radio jet emission. Using the JBF and CLASS complete samples we find the differential number counts slope of the parent population above and below the CLASS 30 mJy flux density limit to be -2.07+/-0.02 and -1.96+/-0.12, respectively.

Although controversial, the scenario of microlensing as the dominant mechanism for the long-term optical variability of quasars does provide a natural explanation for both the statistical symmetry, achromaticity and lack of cosmological time dilation in quasar light curves. Here, we investigate to what extent dark matter populations of compact objects allowed in the currently favored Omega_M=0.3, Omega_Lambda=0.7 cosmology really can explain the quantitative statistical features of the observed variability. We find that microlensing reasonably well reproduces the average structure function of quasars, but fails to explain both the high fraction of objects with amplitudes higher than 0.35 magnitudes and the mean amplitudes observed at redshifts below one. Even though microlensing may still contribute to the long-term optical variability at some level, another significant mechanism must also be involved. This severely complicates the task of using light-curve statistics from quasars which are not multiply imaged to isolate properties of any cosmologically significant population of compact objects which may in fact be present.

We report the final results of the search for gravitationally lensed flat-spectrum radio sources found in the combination of CLASS (Cosmic Lens All Sky Survey) and JVAS (Jodrell-Bank VLA Astrometric Survey). VLA observations of 16,503 sources have been made, resulting in the largest sample of arcsec-scale lens systems available. Contained within the 16,503 sources is a complete sample of 11,685 sources having two-point spectral indices between 1.4 and 5 GHz flatter than -0.5 and 5 GHz flux densities $\geq$30 mJy. A subset of 8,958 sources form a well-defined statistical sample suitable for analysis of the lens statistics. We describe the systematic process by which 149 candidate lensed sources were picked from the statistical sample on the basis of possessing multiple compact components in the 0.2 arcsec-resolution VLA maps. Candidates were followed up with 0.05 arcsec resolution MERLIN and 0.003 arcsec VLBA observations at 5 GHz and rejected as lens systems if they failed well-defined surface brightness and/or morphological tests. Maps for all the candidates can be found on the World Wide Web at http://www.jb.man.ac.uk/research/gravlens/index.html We summarize the properties of each of the 22 gravitational lens systems in JVAS/CLASS. Twelve are double-image systems, nine are four-image systems and one is a six-image system. Thirteen constitute a statistically well-defined sample giving a point-source lensing rate of 1:690$\pm$190. The interpretation of the results in terms of the properties of the lensing galaxy population and cosmological parameters will be published elsewhere. (Abridged)

We derive constraints on cosmological parameters and the properties of the lensing galaxies from gravitational lens statistics based on the final Cosmic Lens All Sky Survey (CLASS) data. For a flat universe with a classical cosmological constant, we find that the present matter fraction of the critical density is $\Omega_{\rm m}=0.31^{+0.27}_{-0.14}$ (68%) $^{+0.12}_{-0.10}$ (systematic). For a flat universe with a constant equation of state for dark energy $w = p_x({pressure})/\rho_x({energy density})$, we find $w < -0.55^{+0.18}_{-0.11}$ (68%).

By comparing the results of numerical microlensing simulations to the observed long-term variability of quasars, strong upper limits on the cosmological density of compact objects in the mass range 0.01 to 0.0001 solar masses may be imposed. Using recently developed methods to better approximate the amplification of large sources, we investigate in what way the constraints are affected by assumptions concerning the size of the optical continuum-emitting region of quasars in the currently favored (Omega_M=0.3, Omega_Lambda=0.7) cosmology.

I review simultaneous constraints on the cosmological parameters lambda_0 and Omega_0 from gravitational lensing. The emphasis is on systematic extragalactic surveys for strong gravitational lenses, mainly the largest and best-defined such survey, JVAS/CLASS.

High resolution MERLIN observations of a newly-discovered four-image gravitational lens system, B0128+437, are presented. The system was found after a careful re-analysis of the entire CLASS dataset. The MERLIN observations resolve four components in a characteristic quadruple-image configuration; the maximum image separation is 542 mas and the total flux density is 48 mJy at 5 GHz. A best-fit lens model with a singular isothermal ellipsoid results in large errors in the image positions. A significantly improved fit is obtained after the addition of a shear component, suggesting that the lensing system is more complex and may consist of multiple deflectors. The integrated radio spectrum of the background source indicates that it is a GigaHertz-Peaked Spectrum (GPS) source. It may therefore be possible to resolve structure within the radio images with deep VLBI observations and thus better constrain the lensing mass distribution.

Kochanek (1992) suggested that the redshifts of gravitational lens galaxies rule out a large cosmological constant. This result was questioned by Helbig & Kayser (1996), who pointed out that selection effects related to the brightness of the lens can bias the results of this test against a high lambda value; however, we did not claim that the observations favoured a high lambda value, merely that current observational data were not sufficient to say either way, using the test as proposed by Kochanek (1992) but corrected for selection effects. Kochanek (1996) pointed out that additional information (fraction of measured lens redshifts) provides additional information which restores the sensitivity of the test to the cosmological model, at least somewhat. Here, I consider three aspects. First, I examine the accuracy of the correction to the test proposed by Kochanek (1996). Second, I compare the slightly different statistical methods which have been used in connection with this test. Third, I discuss what results can be obtained today now that more and better-defined observations are available.

I give a brief overview of the current status of some aspects of the Cosmic Lens All-Sky Survey (CLASS): description of the survey, current list of lens systems, cosmological parameters from lensing statistics, H_0 from time delays, dark lenses, wide-separation lenses.

We present radio, optical, near-infrared and spectroscopic observations of the source B0827+525. We consider this source as the best candidate from the Cosmic Lens All-Sky Survey (CLASS) for a `dark lens' system or binary radio-loud quasar. The system consists of two radio components with somewhat different spectral indices, separated by 2.815 arcsec. VLBA observations show that each component has substructure on a scale of a few mas. A deep K-band exposure with the W.M.Keck-II Telescope reveals emission near both radio components. The K-band emission of the weaker radio component appears extended, whereas the emission from the brighter radio component is consistent with a point source. Hubble Space Telescope F160W-band observations with the NICMOS instrument confirms this. A redshift of 2.064 is found for the brighter component, using the LRIS instrument on the W.M.Keck-II Telescope. The probability that B0827+525 consists of two unrelated compact flat-spectrum radio sources is ~3%, although the presence of similar substructure in both component might reduce this. We discuss two scenarios to explain this system: (i) CLASS B0827+525 is a `dark lens' system or (ii) B0827+525 is a binary radio-loud quasar. B0827+525 has met all criteria that thus far have in 100% of the cases confirmed a source as an indisputable gravitational lens system. Despite this, no lens galaxy has been detected with m_F160W<=23 mag. Hence, we might have found the first binary radio-loud quasar. At this moment, however, we feel that the `dark lens' hypothesis cannot yet be fully excluded.

We present constraints on the cosmological constant lambda_0 and the density parameter Omega_0 from joint constraints from the analyses of gravitational lensing statistics of the Jodrell Bank-VLA Astrometric Survey (JVAS), optical gravitational lens surveys from the literature and CMB anisotropies. This is the first time that quantitative joint constraints involving lensing statistics and CMB anisotropies have been presented. Within the assumptions made, we achieve very tight constraints on both lambda_0 and Omega_0. These assumptions are cold dark matter models, no tensor components, no reionisation, CMB temperature T_CMB=2.728, number of neutrinos n_nu=3, helium abundance Y_He=0.246, spectral index n_s=1.0, Hubble constant H_0=68km/s/Mpc, baryonic density Omega_b=0.05. All models were normalised to the COBE data and no closed models (k=+1) were computed. Using the CMB data alone, the best-fit model has lambda_0=0.60 and Omega_0=0.34 and at 99% confidence the lower limit on lambda_0+Omega_0 is 0.8. Including constraints from gravitational lensing statistics doesn't change this significantly, although it does change the allowed region of parameter space. A universe with lambda_0=0 is ruled out for any value of Omega_0 at better than 99% confidence using the CMB alone. Combined with constraints from lensing statistics, lambda_0=0 is also ruled out at better than 99% confidence. As the region of parameter space allowed by the CMB is, within our assumptions, much smaller than that allowed by lensing statistics, the main result of combining the two is to change the range of parameter space allowed by the CMB along its axis of degeneracy.

We discuss the consistency of constraints in the lambda_0-Omega_0 plane from gravitational lensing statistics, the m-z relation for type Ia supernovae and CMB anisotropies, based on our own (published or unpublished) work and results from the literature.

This poster presents a new 5 GHz combined MERLIN/VLA map of B0218+357 which shows for the first time believable substructure in the Einstein ring. This will now be exploited for further constraints on the model which presently dominates the error on the estimate of H_0 derived from the time delay (10.5+/-0.4 days) measured for this system.

I present constraints on cosmological parameters in the lambda_0-Omega_0 plane from a joint analysis of gravitational lensing statistics (astro-ph/9904175) and the magnitude-redshift relation for Type Ia supernovae (astro-ph/9812133 and astro-ph/9805201). I discuss reasons why this particular combination of tests is important and how the constraints can be improved in the future. The lensing statistics and supernova results are not inconsistent, thus it is meaningful to determine joint constraints on lambda_0 and Omega_0 by combining the results from both tests. The quantity measured by the lens statistics and the m-z relation for type Ia supernovae discussed here is approximately lambda_0 - Omega_0. At 95% confidence, the upper limit on lambda_0 - Omega_0 from lensing statistics alone is 0.45 and from supernovae alone is in the range 0.65--0.81 (depending on the data set). For joint constraints, the upper limit on lambda_0 - Omega_0 is in the range 0.55--0.60 (again depending on the data set). For a flat universe with lambda_0 + Omega_0 = 1, this corresponds to upper limits on lambda_0, taking the top of the range from different data sets, of 0.72, 0.90 and 0.80 for lensing statistics alone, supernovae alone and the joint analysis, respectively. This is perfectly consistent with the current `standard cosmological model' with lambda_0 approximately 0.7 and Omega_0 approximately 0.3 and is consistent with a flat universe but, neglecting other cosmological tests, does not require it.

We present constraints on the cosmological constant $\lambda_{0}$ from gravitational lensing statistics of the Jodrell Bank-VLA Astrometric Survey (JVAS). Although this is the largest gravitational lens survey which has been analysed, cosmological constraints are only comparable to those from optical surveys. This is due to the fact that the median source redshifts of JVAS are lower, which leads to both relatively fewer lenses in the survey and a weaker dependence on the cosmological parameters. Although more approximations have to be made than is the case for optical surveys, the consistency of the results with those from optical gravitational lens surveys and other cosmological tests indicate that this is not a major source of uncertainty in the results. However, joint constraints from a combination of radio and optical data are much tighter. Thus, a similar analysis of the much larger Cosmic Lens All-Sky Survey should provide even tighter constraints on the cosmological constant, especially when combined with data from optical lens surveys. At 95% confidence, our lower and upper limits on $\lambda_{0}-\Omega_{0}$, using the JVAS lensing statistics information alone, are respectively -2.69 and 0.68. For a flat universe, these correspond to lower and upper limits on \lambda_{0} of respectively -0.85 and 0.84. Using the combination of JVAS lensing statistics and lensing statistics from the literature as discussed in Quast & Helbig (Paper I) the corresponding $\lambda_{0}-\Omega_{0}$ values are -1.78 and 0.27. For a flat universe, these correspond to lower and upper limits on $\lambda_{0}$ of respectively -0.39 and 0.64.

We reanalyse optical gravitational lens surveys from the literature in order to determine relative probabilities in the $\lambda_{0}$-$\Omega_{0}$ plane, using a softened singular isothermal sphere lens model. In addition, we examine a portion of the $\lambda_{0}$-$\Omega_{0}$ plane which includes all viable cosmological models; this is vital for comparison with other cosmological tests. The results are, within the errors, consistent with those of more specialised analyses, such as those concerning upper limits on $\lambda_{0}$ in a flat universe. We note that gravitational lensing statistics can provide a quite robust LOWER limit on the cosmological constant as well, which could prove important in confirming current claims of a positive cosmological constant. At 95% confidence, our lower and upper limits on $\lambda_{0}-\Omega_{0}$, using lens statistics information alone, are respectively -3.17 and 0.3. For a flat universe, these correspond to lower and upper limits on $\lambda_{0}$ of respectively -1.09 and 0.65.

The JVAS (Jodrell Bank-VLA Astrometric Survey) and CLASS (Cosmic Lens All-Sky Survey) are well-defined surveys containing about ten thousand flat-spectrum radio sources. For many reasons, flat-spectrum radio sources are particularly well-suited as a population from which one can obtain unbiased samples of gravitational lenses. These are by far the largest gravitational (macro)lens surveys, and particular attention was paid to constructing a cleanly-defined sample for the survey itself and for the underlying luminosity function. Here we present the constraints on cosmological parameters, particularly the cosmological constant, derived from JVAS and combine them with constraints from optical gravitational lens surveys, `direct' measurements of $\Omega_{0}$, $H_{0}$ and the age of the universe, and constraints derived from CMB anisotropies, before putting this final result into the context of the latest results from other, independent cosmological tests.

Measurement of the time delay between multiple images of a gravitational lens system is potentially an accurate method of determining the Hubble constant over cosmological distances. One of the most promising candidates for an application of this technique is the system B0218+357 which was found in the Jodrell Bank/VLA Astrometric Survey (JVAS). This system consists of two images of a compact radio source, separated by 335 milliarcsec, and an Einstein ring which can provide a strong constraint on the mass distribution in the lens. We present here the results of a three-month VLA monitoring campaign at two frequencies. The data are of high quality and both images show clear variations in total flux density, percentage polarization and polarization position angle at both frequencies. The time delay between the variations in the two images has been calculated using a chi-squared minimization to be 10.5+/-0.4 days at 95 per cent confidence, with the error being derived from Monte-Carlo simulations of the light curves. Although mass modelling of the system is at a preliminary stage, taking the lensing galaxy to be a singular isothermal ellipsoid and using the new value for the time delay gives a value for the Hubble constant of 69+13/-19 km/s/Mpc, again at 95 per cent confidence.

In a recent paper, Hawkins (1997) argues on the basis of statistical studies of double-image gravitational lenses and lens candidates that a large population of dark lenses exists and that these outnumber galaxies with more normal mass-to-light ratios by a factor of 3:1. If correct, this is a very important result for many areas of astronomy including galaxy formation and cosmology. In this paper we discuss our new radio-selected gravitational lens sample, JVAS/CLASS, in order to test and constrain this proposition. We have obtained ground-based and HST images of all multiple-image lens systems in our sample and in 12 cases out of 12 we find the lensing galaxies in the optical and/or near infrared. Our success in finding lensing galaxies creates problems for the dark lens hypothesis. If it is to survive, ad hoc modifications seem to be necessary: only very massive galaxies (more than about one trillion solar masses) can be dark, and the cutoff in mass must be sharp. Our finding of lens galaxies in all the JVAS/CLASS systems is complementary evidence which supports the conclusion of Kochanek et al. (1997) that many of the wide-separation optically-selected pairs are physically distinct quasars rather than gravitational lens systems.

Recently, Park & Gott claimed that there is a statistically significant, strong, negative correlation between the image separation and source redshift for gravitational lenses. This is somewhat puzzling if one believes in a flat (k = 0) universe, since in this case the typical image separation is expected to be independent of the source redshift, while one expects a negative correlation in a k = -1 universe and a positive one in a k = +1 universe. Park & Gott explored several effects which could cause the observed correlation, but no combination of these can explain the observations with a realistic scenario. Here, I explore this test further in three ways. First, I show that in an inhomogeneous universe a negative correlation is expected regardless of the value of k. Second, I test whether the image separation-source redshift relation can be used as a test to determine lambda and Omega, rather than just the sign of k. Third, I compare the results of the test from the Park & Gott sample to those using other samples of gravitational lenses, which can illuminate (unknown) selection effects and probe the usefulness of the image separation-source redshift relation as a cosmological test.

The calculation of distances is of fundamental importance in extragalactic astronomy and cosmology. However, no practical implementation for the general case has previously been available. We derive a second-order differential equation for the angular size distance valid not only in all em homogeneous Friedmann-Lemaitre cosmological models, parametrised by $\lambda_{0}$ and $\Omega_{0}$, but also in inhomogeneous `on-average' Friedmann-Lemaitre models, where the inhomogeneity is given by the (in the general case redshift-dependent) parameter $\eta$. Since most other cosmological distances can be obtained trivially from the angular size distance, and since the differential equation can be efficiently solved numerically, this offers for the first time a practical method for calculating distances in a large class of cosmological models. We also briefly discuss our numerical implementation, which is publicly available.

We reanalyse the spectra used by D.-E. Liebscher (no relation) et al with the same goal -- determining the cosmological parameters -- and basically the same assumptions but with a different statistical method, which does not rely on binning the data. Also, we correct for selection effects. We basically confirm their result, with somewhat larger but still very small (formal) errors. However, all world models within the 99% confidence region are ruled out because a firm lower limit on $\Omega_{0}$ is significantly larger. This directly demonstrates for the first time the existence of intrinsic evolution in the Lyman-$\alpha$ forest.

For known gravitational lens systems the redshift distribution of the lenses is compared with theoretical expectations for $10^{4}$~Friedmann-Lema\^\i tre~cosmological models, which more than cover the range of possible cases. The comparison is used for assigning a relative probability to each of the models. The entire procedure is repeated for different values of the inhomogeneity parameter~$\eta$ and the limiting spectroscopic magnitude~$m_{\rm lim}$, which is important for selection effects. The dependence on these two parameters is examined in more detail for the special cases~$\lambda_{0}=0$ and $k=0$. Previous results that this method is a better probe for~$\lambda_{0}$ than $\Omega_{0}$ are confirmed, but it appears that the low probability of models with large~$\lambda_{0}$~values found using similar methods is due to a selection effect. The power of this method to discriminate between cosmological models can of course be improved if more gravitational lens systems are found. However, our numerical simulations indicate that a reasonable number of observed systems cannot deliver interesting constraints on the cosmological parameters.

For suitable gravitational lens systems with unknown lens redshifts, the redshifts and brightnesses (in different colours) of the lenses are predicted for a variety of cosmological models, for both elliptical and spiral galaxy lenses. Besides providing hints as to which systems should be observed with a realistic chance of measuring the lens redshifts, which are needed for detailed lensing statistics and for modelling the lenses, these calculations give a visual impression of the influence of the cosmological model in gravitational lensing.

The general idea of determining cosmological parameters with gravitational lensing statistics is outlined, and then recent work---with an emphasis on applicability to all cosmological models, observational bias, better statistics and robustness testing through numerical simulations---is discussed with relevance to a scheme originally introduced by Kochanek to avoid some of the uncertainties which plague other methods. The main result is that the observations at present---and probably in the future as well---are more indicative of intrinsic scatter than a hint of the correct cosmological model.