In the present-day universe, supermassive black hole masses (MBH) appear to be strongly correlated with their galaxy's bulge luminosity, among other properties. In this study, we explore the analogous relationship between MBH, derived using the virial method, and the stellar R-band bulge luminosity (Lr) or stellar bulge mass (M*) at epochs of 1 < z < 4.5 using a sample of 31 gravitationally lensed AGNs and 20 non-lensed AGNs. At redshifts z > 1.7 (10--12 Gyrs ago), we find that the observed MBH--Lr relation is nearly the same (to within ~0.3 mag) as it is today. When the observed Lr are corrected for luminosity evolution, this means that the black holes grew in mass faster than their hosts, with the MBH/M* mass ratio being a factor of > 4(+2)(-1) times larger at z > 1.7 than it is today. By the redshift range 1<z<1.7 (8-10 Gyrs ago), the MBH/M* ratio is at most two times higher than today, but it may be consistent with no evolution. Combining the results, we conclude that the ratio MBH/M* rises with look-back time, although it may saturate at ~6 times the local value. Scenarios in which moderately luminous quasar hosts at z>1.7 were fully formed bulges that passively faded to the present epoch are ruled out.

Many strong gravitational lenses lie in complex environments, such as poor groups of galaxies, that significantly bias conclusions from lens analyses. We are undertaking a photometric survey of all known galaxy-mass strong lenses to characterize their environments and include them in careful lens modeling, and to build a large, uniform sample of galaxy groups at intermediate redshifts for evolutionary studies. In this paper we present wide-field photometry of the environments of twelve lens systems with 0.24 < z_lens < 0.5. Using a red-sequence identifying technique, we find that eight of the twelve lenses lie in groups, and that ten group-like structures are projected along the line of sight towards seven of these lenses. Follow-up spectroscopy of a subset of these fields confirms these results. For lenses in groups, the group centroid position is consistent with the direction of the external tidal shear required by lens models. Lens galaxies are not all super-L_* ellipticals; the median lens luminosity is < L_*, and the distribution of lens luminosities extends 3 magnitudes below L_* (in agreement with theoretical models). Only two of the lenses in groups are the brightest group galaxy, in qualitative agreement with theoretical predictions. As in the local Universe, the highest velocity-dispersion groups contain a brightest member spatially coincident with the group centroid, whereas lower-dispersion groups tend to have an offset brightest group galaxy. This suggests that higher-dispersion groups are more dynamically relaxed than lower-dispersion groups and that at least some evolved groups exist by z ~ 0.5.

Gravitational lensing assists in the detection of quasar hosts by amplifying and distorting the host light away from the unresolved quasar core images. We present the results of HST observations of 30 quasar hosts at redshifts 1 < z < 4.5. The hosts are small in size (r_e <~ 6 kpc), and span a range of morphologies consistent with early-types (though smaller in mass) to disky/late-type. The ratio of the black hole mass (MBH, from the virial technique) to the bulge mass (M_bulge, from the stellar luminosity) at 1<z<1.7 is broadly consistent with the local value; while MBH/M_bulge at z>1.7 is a factor of 3--6 higher than the local value. But, depending on the stellar content the ratio may decline at z>4 (if E/S0-like), flatten off to 6--10 times the local value (if Sbc-like), or continue to rise (if Im-like). We infer that galaxy bulge masses must have grown by a factor of 3--6 over the redshift range 3>z>1, and then changed little since z~1. This suggests that the peak epoch of galaxy formation for massive galaxies is above z~1. We also estimate the duty cycle of luminous AGNs at z>1 to be ~1%, or 10^7 yrs, with sizable scatter.

We present results from a survey for gravitationally lensed radio lobes. Lensed lobes are a potentially richer source of information about galaxy mass distributions than lensed point sources, which have been the exclusive focus of other recent surveys. Our approach is to identify radio lobes in the FIRST catalog and then search optical catalogs for coincident foreground galaxies, which are candidate lensing galaxies. We then obtain higher-resolution images of these targets at both optical and radio wavelengths, and obtain optical spectra for the most promising candidates. We present maps of several radio lobes that are nearly coincident with galaxies. We have not found any new and unambiguous cases of gravitational lensing. One radio lobe in particular, FOV J0743+1553, has two hot spots that could be multiple images produced by a z=0.19 spiral galaxy, but the lensing interpretation is problematic.

The majority of quasars are weak in their radio emission, with flux densities comparable to those in the optical, and energies far lower. A small fraction, about 10%, are hundreds to thousands of times stronger in the radio. Conventional wisdom holds that there are two classes of quasars, the radio quiets and radio louds, with a deficit of sources having intermediate power. Are there really two separate populations, and if so, is the physics of the radio emission fundamentally different between them? This paper addresses the second question, through a study of radio variability across the full range of radio power, from quiet to loud. The basic findings are that the root mean square amplitude of variability is independent of radio luminosity or radio-to-optical flux density ratio, and that fractionally large variations can occur on timescales of months or less in both radio quiet and radio loud quasars. Combining this with similarities in other indicators, such as radio spectral index and the presence of VLBI-scale components, leads to the suggestion that the physics of radio emission in the inner regions of all quasars is essentially the same, involving a compact, partially opaque core together with a beamed jet.

The redshift distribution of flat-spectrum radio sources with 5 GHz flux densities S>5 mJy is a key component in using current radio lens surveys to probe the cosmological model. We have constructed the first flat-spectrum radio sample in the flux density range 3-20 mJy. Our new sample has 33 sources; we have determined the redshifts of 14 of these (42% complete). The low mean redshift, <z>~0.75, of our faintest sample needs to be confirmed by further observations to improve the sample completeness. We also increased the redshift completeness of several surveys of brighter flat-spectrum sources. While the mean redshift, <z>~1.1 of flat-spectrum samples fainter than 1 Jy is nearly constant, the fraction of the sources identifiable as quasars steadily drops from ~80% to ~10% as the flux density of the sources decreases.

We investigate the evolution of mass-selected early-type field galaxies using a sample of 28 gravitational lenses spanning the redshift range 0 < z < 1. Based on the redshift-dependent intercept of the fundamental plane in the rest frame B band, we measure an evolution rate of d log (M/L)_B / dz = -0.56 +/- 0.04 (all errors are 1 sigma unless noted) if we directly compare to the local intercept measured from the Coma cluster. Re-fitting the local intercept helps minimize potential systematic errors, and yields an evolution rate of d log (M/L)_B / dz = -0.54 +/- 0.09. An evolution analysis of properly-corrected aperture mass-to-light ratios (defined by the lensed image separations) is closely related to the Faber-Jackson relation. In rest frame B band we find an evolution rate of d log (M/L)_B / dz = -0.41 +/- 0.21, a present-day characteristic magnitude of M_{*0} = -19.70 + 5 log h +/- 0.29 (assuming a characteristic velocity dispersion of sigma_{DM*} = 225 km/s), and a Faber-Jackson slope of gamma_{FJ} = 3.29 +/- 0.58. The measured evolution rates favor old stellar populations (mean formation redshift z_f > 1.8 at 2 sigma confidence for a Salpeter initial mass function and a flat Omega_m =0.3 cosmology) among early-type field galaxies, and argue against significant episodes of star formation at z < 1.

We report on a program to obtain HST observations of galaxy-mass gravitational lens systems at optical and infrared wavelengths. Here we discuss the properties of 10 two-image gravitational lens systems (Q0142-100=UM673, B0218+357, SBS0909+532, BRI0952-0115, LBQS1009-0252, Q1017-207=J03.13, B1030+074, HE1104-1805, Q1208+1011, and PKS1830-211). We grouped these 10 systems because they have limited lens model constraints and often show poor contrast between the images and the lens galaxy. Of the 10 lens galaxies, 7 are probably early-type galaxies, 2 are probably late-type galaxies (B0218+357 and PKS1830-211), and one was not detected (Q1208+1011). We detect the host galaxies of the z_s=4.50 lensed quasar in BRI0952-0115, the z_s=2.32 lensed quasar in HE1104-1805, and the unlensed z=1.63 quasar near LBQS1009-0252. We fit a set of four standard lens models to each lens that had sufficient constraints to compare isothermal dark matter and constant mass-to-light lens models, and to explore the effects of local tidal shears.

HST V and I-band observations show that the gravitational lens B1359+154 consists of six images of a single z_s=3.235 radio source and its star-forming host galaxy, produced by a compact group of galaxies at z_l = 1. VLBA observations at 1.7 GHz strongly support this conclusion, showing six compact cores with similar low-frequency radio spectra. B1359+154 is the first example of galaxy-scale gravitational lensing in which more than four images are observed of the same background source. The configuration is due to the unique lensing mass distribution: three primary lens galaxies lying on the vertices of a triangle separated by 0.7 arcsec (4/h kpc), inside the 1.7 arcsec diameter Einstein ring defined by the radio images. The gravitational potential has additional extrema within this triangle, creating a pair of central images that supplement the ``standard'' four-image geometry of the outer components. Simple mass models consisting of three lens galaxies constrained by HST and VLBA astrometry naturally reproduce the observed image positions but must be finely-tuned to fit the flux densities.

We report on two surveys of radio-weak AGN to look for radio variability. We find significant variability with an RMS of 10-20% on a timescale of months in radio-quiet and radio-intermediate quasars. This exceeds the variability of radio cores in radio-loud quasars (excluding blazars), which vary only on a few percent level. The variability in radio-quiet quasars confirms that the radio emission in these sources is indeed related to the AGN. The most extremely variable source is the radio-intermediate quasar III Zw 2 which was recently found to contain a relativistic jet. In addition we find large amplitude variabilities (up to 300% peak-to-peak) in a sample of nearby low-luminosity AGN, Liners and dwarf-Seyferts, on a timescale of 1.5 years. The variability could be related to the activity of nuclear jets responding to changing accretion rates. Simultaneous radio/optical/X-ray monitoring also for radio-weak AGN, and not just for blazars, is therefore a potentially powerful tool to study the link between jets and accretion flows.

We performed an automated comparison of the FIRST radio survey with the APM optical catalog to find radio lobes with optical counterparts. Based on an initial survey covering ~3000 square degrees, we selected a sample of 33 lens candidates for VLA confirmation. VLA and optical observations of these candidates yielded two lens systems, one a new discovery (J0816+5003), and one of which was previously known (J1549+3047). Two other candidates have radio lobes with galaxies superposed, but lack evidence of multiple imaging. One of our targets (J0958+2947) is a projected close pair of quasars (8'' separation at redshifts 2.064 and 2.744). Our search method is highly efficient, with >5% of our observing targets being lensed, compared to the usual success rate of <1%. Using the whole FIRST survey, we expect to find 5--10 lenses in short order using this approach, and the sample could increase to hundreds of lensed lobes in the Northern sky, using deeper optical surveys and planned upgrades to the VLA. Such a sample would be a powerful probe of galaxy structure and evolution.

We report the discovery of a new double-image quasar that was found during a search for gravitational lenses in the southern sky. Radio source PMN J1838-3427 is composed of two flat-spectrum components with separation 1", flux density ratio 14:1 and matching spectral indices, in VLA and VLBA images. Ground-based BRI images show the optical counterpart (total I=18.6) is also double with the same separation and position angle as the radio components. An HST/WFPC2 image reveals the lens galaxy. The optical flux ratio (27:1) is higher than the radio value probably due to differential extinction of the components by the lens galaxy. An optical spectrum of the bright component contains quasar emission lines at z=2.78 and several absorption features, including prominent Ly-alpha absorption. The lens galaxy redshift could not be measured but is estimated to be z=0.36 +/- 0.08. The image configuration is consistent with the simplest plausible models for the lens potential. The flat radio spectrum and observed variability of PMN J1838-3427 suggest the time delay between flux variations of the components is measurable, and could thus provide an independent measurement of H_0.

We report the discovery of a new two-image gravitational lens candidate. The system RX J0921+4529 contains two z_s=1.66 quasars separated by 6."93 with an H band magnitude difference of $\Delta m=1.39$. The HST NIC2 H band images reveal an H=18.2 spiral galaxy between the quasar images, which is probably a member of a z_l=0.32 X-ray cluster centered on the field. We detect an extended source near the fainter quasar image but not in the brighter image. If this extended source is the host galaxy of the fainter quasar, then the system is a binary quasar rather than a gravitational lens. VLA observations at 3.6 cm reveal emission from the lens galaxy at the flux level of 1 mJy and a marginal detection of the brighter quasar.

Most gravitational lenses are early-type galaxies in relatively low density environments -- a ``field'' rather than a ``cluster'' population. We show that field early-type galaxies with 0 < z < 1, as represented by the lens galaxies, lie on the same fundamental plane as those in rich clusters at similar redshifts. We then use the fundamental plane to measure the combined evolutionary and K-corrections for early-type galaxies in the V, I and H bands. Only for passively evolving stellar populations formed at z > 2 (H_0=65 km/s Mpc, Omega_0=0.3, Lambda_0=0.7) can the lens galaxies be matched to the local fundamental plane. The high formation epoch and the lack of significant differences between the field and cluster populations contradict many current models of the formation history of early-type galaxies. Lens galaxy colors and the fundamental plane provide good photometric redshift estimates with an empirical accuracy of -0.03 +/- 0.11 for the 17 lenses with known redshifts. A mass model dominated by dark matter is more consistent with the data than either an isotropic or radially anisotropic constant M/L mass model, and a radially anisotropic model is better than an isotropic model.

Infrared images of the Q 0957+561 gravitational lens obtained with the Hubble Space Telescope show two large (\sim 5 arcsec) lensed images of the z_s=1.41 quasar host galaxy. Parts of the host galaxy are doubly-imaged like the quasar, while other parts are quadruply-imaged. The distortions of the host galaxy offer the best probe yet of the global structure of the lensing potential, which is essential for determining the Hubble constant from the measured time delay. The distortions are inconsistent with the predictions of previously published lens models, which invalidates those models and their implications for H_0. New models show that the distortions finally break the long-standing degeneracy between the shape of the lens galaxy and the tidal shear contributed by the cluster containing the lens galaxy. The shape of the lens galaxy's mass distribution must be remarkably similar to the shape of its luminosity distribution, and most models that produce reasonable values for the Hubble constant roughly match the observed ellipticity gradient and isophote twist of the lens galaxy. Also, the cluster must be non-spherical and produce a relatively small tidal shear. Although there are still degeneracies in the lens models that lead to a 25% uncertainty in the derived value of the Hubble constant, there are also strong prospects for new observations to further improve the constraints and reduce the uncertainties.

We present H-band observations of gravitationally lensed QSO host galaxies obtained with NICMOS on HST as part of the CfA-Arizona-Gravitational-Lens-Survey (CASTLES). The detections are greatly facilitated by the lensing magnification in these systems; we find that most hosts of radio-quiet QSOs (RQQ) at z~2 are of modest luminosity (L<L_*). They are 2-5 times fainter than the hosts of radio-loud QSOs at the same epoch. Compared to low redshifts, RQQ hosts at z~2 also support higher nuclear luminosities at given stellar host mass. This suggests that the supermassive black holes at their centers grew faster at early epochs than the stellar body of their surrounding host galaxies.

Most gravitational lens galaxies are early-type galaxies in relatively low density environments. We show that they lie on the same fundamental plane as early-type galaxies in both local and distant rich clusters. Their surface brightness evolution requires a typical star formation epoch of z=2-3, almost indistinguishable from that of rich cluster galaxies at comparable redshifts. The restricted galaxy type range of the lenses means that photometric redshifts work well even with only 1-3 filter photometry. We make preliminary measurements of the mass and luminosity functions of the lens galaxies, and find they are consistent with the standard model used for deriving cosmological limits using lens statistics. As expected for a mass-weighted sample, they are more massive and more luminous than the overall early-type galaxy population.

We use observations from the CASTLES survey of gravitational lenses to study extinction in 23 lens galaxies with $0 < z_l < 1$. The median differential extinction between lensed images is $\Delta E(\bv) = 0.05$ mag, and the directly measured extinctions agree with the amount needed to explain the differences between the statistics of radio and (optical) quasar lens surveys. We also measure the first extinction laws outside the local universe, including an $R_V=7.2$ curve for a molecular cloud at $z_l=0.68$ and an $R_V=1.5$ curve for the dust in a redshift $z_l=0.96$ elliptical galaxy.

We describe a series of new applications of gravitational lenses as astrophysical and cosmological tools. Such applications are becoming possible thanks to advances in the quality and quantity of observations. CASTLES (CfA-Arizona-Space-Telescope-LEns-Survey) {cfa-www.harvard.edu/castles} is an ongoing project that exploits the sensitivity and resolution of the Hubble Space Telescope (HST) at optical and infrared wavelengths to study the sample of over 50 known gravitational lenses. The observational goal of CASTLES is a uniform sample of multi-band images of all known galaxy-mass lens systems, to derive precise photometry and astrometry for the lens galaxies, all the known images, and any source or lens components that might have escaped detection. With these measurements we are investigating: (1) the properties of dust and of dark matter in lens galaxies out to z\sim1; (2) the dark matter in lens galaxies and in their environments; (3) the evolution of lens galaxies; and (4) the cosmological model, for instance by refining constraints on the Hubble constant H_0.

The lens candidate RXJ 0921+4529 consists of two z_s=1.66 quasar separated by 6."93 with an H band magnitude difference of \Delta m=1.39. The lens appears to be a z_l=0.31 X-ray cluster, including a m_H=18.5 late-type galaxy lying between the quasar images. We detect an extended source overlapping the faint quasar but not the bright quasar. If this extended source is the host galaxy of the fainter quasar, then the system is a quasar binary rather than a gravitational lens.

We have found three gravitational lenses (two are new) by observing 34 likely FIRST radio lobes with APM galaxy counterparts. We expect to find $\sim30$ such lenses in over the next few years, which will significantly improve lensing constraints on galaxy structure and cosmology.

We present a suite of observations of the recently identified ultraluminous BAL quasar APM 08279+5255, taken both in the infra-red with the NICMOS high resolution camera on board the Hubble Space Telescope, and at 3.5cm with the Very Large Array. With an inferred luminosity of ~5x10^15 Solar luminosities, APM 08279+5255 is apparently the most luminous system known. Extant ground-based images show that APM 08279+5255 is not point-like, but is instead separated into two components, indicative of gravitational lensing. The much higher resolution images presented here also reveal two point sources, A and B, of almost equal brightness (f_B/f_A=0.782 +/- 0.010), separated by 0."378 +/- 0."001, as well as a third, previously unknown, fainter image, C, seen between the brighter images. While the nature of C is not fully determined, several lines of evidence point to it being a third gravitationally lensed image of the quasar, rather than being the lensing galaxy. Simple models which recover the relative image configuration and brightnesses are presented. While proving to be substantially amplified, APM 08279+5255 possesses an intrinsic bolometric luminosity of ~10^14 to 10^15 Solar luminosities and remains amongst the most luminous objects known.

We examine models of the mass distribution for the first known case of gravitational lensing. Several new sets of constraints are used, based on recent observations. We remodel the VLBI observations of the radio jets in the two images of Q0957+561, showing that the previously derived positions and uncertainties were incorrect. We use as additional constraints the candidate lensed pairs of galaxies identified recently with the Hubble Space Telescope. We explore a wider range of lens models than before, and find that the Hubble constant is not tightly constrained once elliptical lens models are considered. We also discuss the systematic uncertainties caused by the cluster containing the lens galaxy. We conclude that additional observations of the candidate lensed galaxies as well as direct measurements of the cluster mass profile are needed before an accurate value of the Hubble constant can be derived from this lens.

Deep HST/NICMOS H (F160W) band observations of the z=1.96 quasar pair Q 1634+267A,B reveal no signs of a lens galaxy to a 1 sigma threshold of approximately 22.5 mag. The minimum luminosity for a normal lens galaxy would be a 6L_* galaxy at z > 0.5, which is 650 times greater than our detection threshold. Our observation constrains the infrared mass-to-light ratio of any putative, early-type, lens galaxy to (M/L)_H > 690h_65 (1200h_65) for Omega_0=0.1 (1.0) and H_0=65h_65 km/s/Mpc. We would expect to detect a galaxy somewhere in the field because of the very strong Mg II absorption lines at z=1.1262 in the Q 1634+267 A spectrum, but the HST H-band, I-band (F785LP) and V-band (F555W) images require that any associated galaxy be very under-luminous less than 0.1 L^*_H (1.0 L^*_I) if it lies within less than 40 h^{-1} (100 h^{-1}) kpc from Q 1634+267 A,B. While the large image separation (3.86 arcsec) and the lack of a lens galaxy strongly favor interpreting Q 1634+267A,B as a binary quasar system, the spectral similarity remains a puzzle. We estimate that at most 0.06% of randomly selected quasar pairs would have spectra as similar to each other as the spectra of Q 1634+267 A and B. Moreover, spectral similarities observed for the 14 quasar pairs are significantly greater than would be expected for an equivalent sample of randomly selected field quasars. Depending on how strictly we define similarity, we estimate that only 0.01--3% of randomly drawn samples of 14 quasar pairs would have as many similar pairs as the observational sample.

We describe the goals of the CASTLES (CfA-Arizona-Space-Telescope-LEns-Survey) project including a sample of NICMOS images of gravitational lenses and a brief list of the preliminary findings.

We determine 37 differential extinctions in 23 gravitational lens galaxies over the range 0 < z_l < 1. Only 7 of the 23 systems have spectral differences consistent with no differential extinction. The median differential extinction for the optically-selected (radio-selected) subsample is E(B-V)=0.04 (0.06) mag. The extinction is patchy and shows no correlation with impact parameter. The median total extinction of the bluest images is E(B-V)=0.08 mag, although the total extinction distribution is dominated by the uncertainties in the intrinsic colors of quasars. The directly measured extinction distributions are consistent with the mean extinction estimated by comparing the statistics of quasar and radio lens surveys, thereby confirming the need for extinction corrections when using the statistics of lensed quasars to estimate the cosmological model. A disjoint subsample of two face-on, radio-selected spiral lenses shows both high differential and total extinctions, but standard dust-to-gas ratios combined with the observed molecular gas column densities overpredict the amount of extinction by factors of 2-5. For several systems we can estimate the extinction law, ranging from R_V=1.5+/-0.2 for a z_l=0.96 elliptical, to R_V=7.2+/-0.1 for a z_l=0.68 spiral. For the four radio lenses where we can construct non-parametric extinction curves we find no evidence for gray dust over the IR-UV wavelength range. The dust can be used to estimate lens redshifts with reasonable accuracy, although we sometimes find two degenerate redshift solutions.

We show that most gravitational lenses lie on the passively evolving fundamental plane for early-type galaxies. For burst star formation models (1 Gyr of star formation, then quiescence) in low Omega_0 cosmologies, the stellar populations of the lens galaxies must have formed at z_f > 2. Typical lens galaxies contain modest amounts of patchy extinction, with a median differential extinction for the optical (radio) selected lenses of E(B-V) = 0.04 (0.07) mag. The dust can be used to determine both extinction laws and lens redshifts. For example, the z_l=0.96 elliptical lens in MG0414+0534 has an R_V=1.7 +/- 0.1 mean extinction law. Arc and ring images of the quasar and AGN source host galaxies are commonly seen in NICMOS H band observations. The hosts are typically blue, L < L_* galaxies.

We have obtained and modeled new NICMOS images of the lens system MG1131+0456, which show that its lens galaxy is an H=18.6 mag, transparent, early-type galaxy at a redshift of about z_l = 0.85; it has a major axis effective radius R_e=0.68+/-0.05 arcsec, projected axis ratio b/a=0.77+/-0.02, and major axis PA=60+/-2 degrees. The lens is the brightest member of a group of seven galaxies with similar R-I and I-H colors, and the two closest group members produce sufficient tidal perturbations to explain the ring morphology. The host galaxy of the MG1131+0456 source is a z_s > 2 ERO (``extremely red object'') which is lensed into optical and infrared rings of dramatically different morphologies. These differences imply a strongly wavelength-dependent source morphology that could be explained by embedding the host in a larger, dusty disk. At 1.6 micron (H), the ring is spectacularly luminous, with a total observed flux of H=17.4 mag and a de-magnified flux of 19.3 mag, corresponding to a 1-2L_* galaxy at the probable source redshift of z_s > 2. Thus, it is primarily the stellar emission of the radio source host galaxy that produces the overall colors of two of the reddest radio lenses, MG1131+0456 and B~1938+666, aided by the suppression of optical AGN emission by dust in the source galaxy. The dusty lens hypothesis -- that many massive early-type galaxies with 0.2 < z_l < 1.0 have large, uniform dust opacities -- is ruled out.

Hubble Space Telescope observations of the gravitational lens PG 1115+080 in the infrared show the known z =0.310 lens galaxy and reveal the z = 1.722 quasar host galaxy. The main lens galaxy G is a nearly circular (ellipticity < 0.07) elliptical galaxy with a de Vaucouleurs profile and an effective radius of R_e = 0.59 +/- 0.06 arcsec (1.7 +/- 0.2 h^{-1} kpc for Omega = 1 and h = H_0/100 km/s/Mpc). G is part of a group of galaxies that is a required component of all successful lens models. The new quasar and lens positions (3 milliarcsecond errors) yield constraints for these models that are statistically degenerate, but several conclusions are firmly established. (1) The principal lens galaxy is an elliptical galaxy with normal structural properties, lying close to the fundamental plane for its redshift. (2) The potential of the main lens galaxy is nearly round, even when not constrained by the small ellipticity of the light of this galaxy. (3) All models involving two mass distributions place the group component near the luminosity-weighted centroid of the brightest nearby group members. (4) All models predict a time delay ratio r_{ABC} = 1.3. (5) Our lens models predict H_0 = 44 +/- 4 km/s/Mpc if the lens galaxy contains dark matter and has a flat rotation curve, and H_0 = 65 +/- 5 km/s/Mpc if it has a constant mass-to-light ratio. (6) Any dark halo of the main lens galaxy must be truncated near 1.5 arcsec (4 h^{-1} kpc) before the inferred Ho rises above 60 km/s/Mpc. (7) The quasar host galaxy is lensed into an Einstein ring connecting the four quasar images, whose shape is reproduced by the models. Improved NICMOS imaging of the ring could be used to break the degeneracy of the lens models.

We present radio and optical observations of MG0248+0641, which contains a kiloparsec-scale ring-like structure in one of its radio lobes. The radio observations show a typical core-double morphology: a central core between two lobes, each of which has a hotspot. The western radio lobe appears as a nearly continuous ring, with linear polarization electric field vectors which are oriented in a radial direction from the ring center. We consider several different interpretations for the nature of this ring, including gravitational lensing of a normal jet by a foreground galaxy. Even though simple lensing models can describe the ring morphology reasonably well, the high linear polarization seen around the ring cannot be easily explained. The chance interposition of a galactic supernova remnant, nova, planetary nebula, or H II region, has been ruled out. The highly polarized ring of MG0248+0641 is much like the prominent ring seen in 3C219, and the multiple ones in 3C310 and Hercules A, suggesting that similar physical processes are producing shell structures in these radio galaxies. The ring in MG0248+0641 may be caused by the formation of ``bubbles'', as a result of instabilities in the energy flow down the western radio jet. It may also be possible that the required instabilities are triggered by the infall of gas, via tidal interaction of the central source with a nearby galaxy. This scenario may be indicated by our marginal detection of an optical source close to the western hotspot.

We have produced deep radio maps of the double quasar 0957+561 from multiple-epoch VLA observations. To achieve high sensitivity to extended structure we have re-reduced the best available 1.6 GHz observations and have combined 5 GHz data from multiple array configurations. Regions of faint emission approximately 15 arcsec north and south of the radio source G are probably lobes associated with the lensing galaxy. An arc 5 arcsec to the east of G may be a stretched image of emission in the background quasar's environment. 1.4 arcsec southwest of G we detect a source that we interpret as an image of emission from the quasar's western lobe, which could provide a constraint on the slope of the gravitational potential in the central region of the lens. We explore the consequences of these new constraints with simple lens models of the system.

We report the discovery of a new gravitationally lensed radio source. Radio maps of MG0751+2716 show four lensed images, which, at higher resolution, are resolved into long arcs of emission. A group of galaxies is present in optical images, including the principal lensing galaxy, with a much brighter galaxy just a few arcseconds away. We have measured the redshift of this brighter galaxy. No optical counterpart to the background source has been detected. Lens models that can readily reproduce the lensed image positions all require a substantial shear component. However, neither the very elongated lens nor the bright nearby galaxy are correctly positioned to explain the shear. Lens models which associate the mass with the light of galaxies in the group can produce an acceptable fit, but only with an extreme mass-to-light ratio in one of the minor group members.

Quadruple gravitational lens systems offer the possibility of measuring time delays for image pairs, microlensing effects, and extinction in distant galaxies. Observations of these systems may be used to obtain estimates of H_o and to study the various mass components of lens galaxies at high redshifts. With HST, we have observed the reddest known gravitational lens system, MG 0414+0534. We used WFPC2/PC1 to obtain deep, high-resolution images with two filters, F675W and F814W. We present a detailed analysis of all of the components, as well as macrolens models. Our main results are: (1) confirmation that MG 0414+0534 is inescapably a gravitational lens system; (2) discovery of a blue arc connecting the 3 brightest images of the QSO central core; (3) accurate positions and apparent brightnesses for all 4 known images of the QSO central core and for the lens galaxy G; (4) a good representation of the brightness distribution of G by elliptical isophotes with a De Vaucouleurs profile, characteristic of an elliptical galaxy; (5) models that consist of simple elliptical potentials and account qualitatively, not quantitatively, for the HST image positions, arc morphology and radio flux ratios for the images of the QSO central core; (6) a possible new test to distinguish between reddening in the host galaxy of the QSO and in the lens galaxy, based on future accurate measurements of spatial variations in the color of the arc; and (7) the suggestion that microlensing is a plausible cause for the differences between the radio and optical flux ratios for the brightest images, A1 and A2. Further observations and measurements such as of the redshift of the lens galaxy, can be used fruitfully to study microlensing for this system.

We report observations of the four-image gravitational lens system Q2237+0305 with the VLA at 20 cm and 3.6 cm. The quasar was detected at both frequencies (\approx 0.7 mJy) with a flat spectrum. All four lensed images are clearly resolved at 3.6 cm, and the agreement of the radio and optical image positions is excellent. No radio emission is detected from the lensing galaxy, and any fifth lensed quasar image must be fainter than \sim 20% of the A image flux density. Since the optical quasar images are variable and susceptible to extinction, radio flux ratios provide the best measurement of the macrolensing magnification ratios. The radio B/A and C/A image flux ratios are consistent with the observed range of optical variations, but the D/A ratio is consistently higher in the radio than in the optical. The radio ratios are consistent with magnification ratios predicted by lens models, and weaken alternative interpretations for Q2237+0305. More accurate radio ratios can distinguish between the models, as well as improve our understanding of both microlensing and extinction in this system.

We have measured a velocity dispersion for the foreground galaxy in this gravitationally lensed system. Our dispersion confirms the prediction from lens models, provided that the source is distant enough ($z_{_S}>0.2$). Current interpretations of lensing statistics depend sensitively on how the optical and mass dispersions are related. For $z_{_S}>0.5$, our observations favor $\sigma_{mass}/\sigma_{opt}\simeq1$, but our uncertainties prevent us from clearly distinguishing between dynamical models. We could not obtain a unique mass profile from lensing constraints, but models with a constant mass-to-light ratio are possible. We also found unusual rotation in the lensing galaxy, and confirmed its redshift.

We present hybrid maps of the A and B images of 0957+561 from each of four sessions of 6 cm VLBI observations that span the six-year interval 1987-1993. The inner- and outer-jets are clearly detected, and confirm the structures reported previously. There is no evidence of change in the separation between the core and inner-jet components, so the prospect of measuring the time delay using differential proper motions is not promising. The flux density in the core of each image peaked between 1989 and 1992. From the variation in these flux densities, we obtain a time-delay estimate of $\sim$1 yr.

We show that the high redshift IRAS source F10214 is highly magnified by the gravitational field of an intervening elliptical galaxy, accounting for its many anomalous properties. Detailed radio and near-IR images identify the IRAS source with a symmetric arc, centered on a red object, or lensing galaxy. To explain the observed structures, the center of the source must much more highly magnified than its outer regions. Lensing predicts a small counterimage to the arc, which we find adjacent to the lensing galaxy. A red component in the observed spectrum suggests a lens redshift of unity, and the lens model yields a mass estimate of $M(r<3kpc)\approx10^{11}{M_solar}$, consistent with an ordinary elliptical galaxy. We present new high-resolution optical images which show a thin arc of emission, implying an intrinsically small source ($<0.5kpc$) which is highly magnified ($\sim20\times$). Since the optical is strongly polarized with a Seyfert~II spectrum, we propose that the optical arc is magnified image of the inner region of an obscured AGN. The obscuring ``torus'' will be similarly magnified, naturally accounting for the large IR flux. We show that finding objects like F10214+4724 in redshift surveys is probable, given the level of magnification bias expected for compact luminous IRAS sources. Such cases represent the obscured AGN counterparts to the lensed QSO population and, because of their extended sizes, are useful in determining the mass distribution in the lensing galaxies.