We present a data set of images of the gravitationally lensed quasar Q2237+0305, that was obtained at the Apache Point Observatory (APO) between June 1995 and January 1998. Although the images were taken under variable, often poor seeing conditions and with coarse pixel sampling, photometry is possible for the two brighter quasar images A and B with the help of exact quasar image positions from HST observations. We obtain a light curve with 73 data points for each of the images A and B. There is evidence for a long (>~ 100 day) brightness peak in image A in 1996 with an amplitude of about 0.4 to 0.5 mag (relative to 1995), which indicates that microlensing has been taking place in the lensing galaxy. Image B does not vary much over the course of the observation period. The long, smooth variation of the light curve is similar to the results from the OGLE monitoring of the system (Wozniak et al. 2000a).

If the halo of the lensing galaxy 0957+561 is made of massive compact objects (MACHOs), they must affect the lightcurves of the quasar images Q0957+561 A and B differently. We search for this microlensing effect in the double quasar by comparing monitoring data for the two images A and B - obtained with the 3.5m Apache Point Observatory from 1995 to 1998 - with intensive numerical simulations. This way we test whether the halo of the lensing galaxy can be made of MACHOs of various masses. We can exclude a halo entirely made out of MACHOs with masses between 10e(-6) Msun and 10e(-2) Msun for quasar sizes of less than 3*10e(14)cm (h = 0.6), hereby extending previous limits upwards by one order of magnitude.

In order to use a gravitational lens to measure the Hubble constant accurately, it is necessary to derive a reliable model of the lens surface potential. If the analysis is restricted to the locations and magnifications of point images, the derived Hubble constant depends upon the class of mass models used to fit the data. However, when there is extended emission from an Einstein ring, it may be possible to derive a potential from the observed surface brightness in a model-independent manner. This procedure is illustrated with reference to B1608+656. The multi-band images are de-reddened, de-convolved and de-contaminated so that the luminous matter and the surface brightness contours in the Einstein ring are both faithfully mapped. This intensity distribution can then be used to reconstruct the potential. Progress in implementing this program is reported. The observed incidence of multiple-imaged galaxies on the Hubble Deep Fields is an order of magnitude smaller than naively predicted on the basis of radio lens surveys, like CLASS, but consistent with the rate computed using surface photometry of candidate lens galaxies assuming standard mass to light ratios. In order to resolve this paradox, it is suggested that most galaxy lenses are located in compact groups.

We present new results from a continuing Keck program to study gravitational lens systems. We have obtained redshifts for three lens systems, SBS 0909+532, HST 1411+5211, and CLASS B2319+051. For all of these systems, either the source or lens redshift (or both) has been previously unidentified. We find (z_l, z_s) = (0.830, 1.377) for SBS 0909+532; (z_l, z_s) = (0.465, 2.811) for HST 1411+5211, although the source redshift is still tentative; and (z_l1, z_l2) = (0.624, 0.588) for the two lensing galaxies in CLASS B2319+051. The background radio source in B2319+051 has not been detected optically; its redshift is, therefore, still unknown. We find that the spectral features of the central lensing galaxy in all three systems are typical of an early-type galaxy. The observed image splittings in SBS 0909+532 and HST 1411+5211 imply that the masses within the Einstein ring radii of the lensing galaxies are 1.4 x 10^{11} and 2.0 x 10^{11} h^{-1} M_sun, respectively. The resulting B band mass-to-light ratio for HST 1411+5211 is 41.3 +/- 1.2 h (M/L)_sun, a factor of 5 times higher than the average early-type lensing galaxy. This large mass-to-light is almost certainly the result of the additional mass contribution from the cluster CL 3C295 at z = 0.46. For the lensing galaxy in SBS 0909+532, we measure (M/L)_B = 4^{+11}_{-3} h (M/L)_sun where the large errors are the result of significant uncertainty in the galaxy luminosity. While we cannot measure directly the mass-to-light ratio of the lensing galaxy in B2319+051, we estimate that (M/L)_B is between 3-7 h (M/L)_sun.

In a number of multiply imaged quasar systems, a significant contribution to the lensing potential is provided by groups and clusters of galaxies associated with the primary lens. As part of an ongoing effort to gather observational data on these systems, we present spectroscopy and near-infrared and optical photometry of galaxies in the field of the quadruple lens system B 1422+231. The spectra show that the primary lens and five nearby galaxies belong to a compact group at z = 0.338. The median projected radius of this group is 35 h^{-1} kpc and its velocity dispersion is 550 km/s. A straightforward application of the virial theorem yields a group mass of 1.4 x 10^{13} h^{-1} M(sun), which provides sufficient external shear to produce the observed image configuration. This data rules out a class of models and improves the system's prospects for a measurement of the Hubble constant.

The quadruple system PG 1115+080 is the second gravitational lens with a reported measurement of the Hubble constant. In addition to the primary lens, three nearby galaxies are believed to contribute significantly to the lensing potential. In this paper we report accurate redshifts for all four galaxies and show that they belong to a single group at z_d = 0.311. This group has very similar properties to Hickson's compact groups of galaxies found at lower redshifts. We briefly discuss implications for the existing lens models and derive H_0 = 52 +/- 14 km/s/Mpc.

Photometric monitoring of the gravitational lens system 0957+561A,B in the g and r bands with the Apache Point Observatory (APO) 3.5 m telescope during 1996 shows a sharp g band event in the trailing (B) image light curve at the precise time predicted from the observation of an event during 1995 in the leading (A) image with a delay of 415 days. This success confirms the "short delay," and the lack of any feature at a delay near 540 days rejects the "long delay" for this system, resolving a long-standing controversy. A series of statistical analyses of our light curve data yield a best fit delay of 417 +/- 3 days (95% confidence interval). Recent improvements in the modeling of the lens system (consisting of a galaxy and cluster) allow us to derive a value of the global (at z = 0.36) value of Hubble's constant H_0 using Refsdal's method, a simple and direct distance determination based on securely understood physics and geometry. The result is H_0 = 63 +/- 12 km/s/Mpc (for Omega = 1) where this 95% confidence interval is dominated by remaining lens model uncertainties.

The potential of gravitational lenses for providing direct, physical measurements of the Hubble constant, free from systematic errors associated with the traditional distance ladder, has long been recognized. However, it is only recently that there has been a convincing measurement of a time delay sufficiently accurate to carry out this program. By itself, an accurate time delay measurement does not produce an equivalently definite Hubble constant and the errors associated with models of the primary lens, propagation through the potential fluctuations produced by the large-scale structure and the global geometry of the universe must also be taken into account. The prospects for measuring several more time delays and the feasibility of making the corresponding estimates of the Hubble constant with total error smaller than ten percent are critically assessed.

We report the detection of a very red source coincident with the gravitational lens 1938+66.6 (Patnaik et al. 1992) in K' (2.12 micron), H (1.6 micron), J (1.25 micron), and Thuan-Gunn r (0.65 micron) bands. 1938+66.6 has previously been detected as a partial radio ring indicating lensing. We find K'=17.1 +- 0.1 and r = 23.9 +- 0.2, making it a very red source with (r-K')=6.8 +- 0.25. We also observed in Thuan-Gunn g band (0.49 micron) and found g>24.5 at the 90% confidence level. We interpret our observations as a reddened gravitational lens on the basis of its optical-IR color and positional coincidence with the radio source.

CCD photometry of the gravitational lens system 0957+561A,B in the g and r bands was obtained on alternate nights, weather permitting, from December 1994 through May 1995 using the Double Imaging Spectrograph (DIS) on the Apache Point Observatory (APO) 3.5-meter telescope. The remote observing and fast instrument change capabilities of this facility allowed accumulation of light curves sampled frequently and consistently. The Honeycutt ensemble photometry algorithm was applied to the data set and yielded typical relative photometric errors of approximately 0.01 magnitudes. Image A exhibited a sharp drop of about 0.1 magnitudes in late December 1994; no other strong features were recorded in either image. This event displays none of the expected generic features of a microlensing-induced flux variation and is likely to be intrinsic to the quasar; if so, it should also be seen in the B image with the lensing differential time delay. We give the expected 1996 image B light curves based on two values of the time delay and brightness ratio which have been proposed and debated in the literature. Continued monitoring of the system in the first half of 1996 should easily detect the image B event and thus resolve the time-delay controversy.

We present a quantitative analysis of the effect of microlensing caused by random motion of individual stars in the galaxy which is lensing a background quasar. We calculate a large number of magnification patterns for positions of the stars slightly offset from one frame to the next, and in this way obtain light curves for fixed quasar and galaxy positions, only due to the change in the relative star positions. These light curves are analyzed to identify micro- lensing events, which are classified according to height, duration, and slope. We find that microlensing events produced by random motion of stars are shorter, steeper, and more frequent than bulk motion events. This difference is caused by the fact that in the case of random motion, caustics can move with arbitrarily high velocity, producing very short events. An accompanying video illustrates these results. For three different values of the surface mass density kappa, it shows time sequences of 1000 magnification patterns for slowly changing lens positions, together with the positions and velocity vectors of the microlensing stars. Short sequences of the video are available as MPEG movie by anonymous ftp at ftp://astro.princeton.edu/jkw/microlensing/moving_stars .