Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Gravitational Lenses: The Current Sample, Recent Results, and Continuing Searches JACQUELINE N. HEWITT Massachusetts Institute of Technology Gravitational tensing is one of the topics in astrophysics that was quite extensively discussed over many decades in the theoretical literature before it was actually observed. We are now at the tenth anniversary of the discovery of the first gravitational lens in 1979, and it is interesting to note how the held has developed over the past decade. After an initial slow rate of discovery of gravitational lens systems (about one per year), the last few years have seen an explosion in the number of reported cases. Toe variety the types of systems has also increased markedly. Attention was drawn to the first few cases because quasars at the same redshift, with similar optical spectra, were observed with angular separations of only a few arc seconds. Recent verified and proposed gravitational lenses include the giant luminous arcs, their accompanying more common small blue "arclets," the radio rings, a field of twin galaxies, statistical tensing, and microlensing. In the last decade, most observational effort has been devoted to searching for new candidate lens systems and carefully measuring their properties, both to test whether they are indeed tensed and to provide constrains for modeling. Theoretical efforts have been extensive, and have included modeling of the Mown lens systems and more general theoreti- cal calculations aimed at understanding gravitational potentials with some simplifying properties. lPhe case of an elliptically symmetric potential is the most complicated potential that can be said to be thoroughly understood. Blandford and Kochanek (1987) and Kochanek and Blandford (1987) have examined the solutions for such a potential in considerable detail, and have simulated the statistical properties of an ensemble of elliptical lenses. As pointed by Narayan and Grossman (1989), the different solutions for the elliptical lens provide a useful framework for categorizing the known gravitational lens systems. If a source falls on the optical axis of the lens, 192
. HIGH-ENER~ ~TROP~ICS 193 four bright images surrounding the center of the lens are formed. If the quadrupole moment of the lens is sufficiently small or the extent of the source is sufficiently large, the four images merge to form a ring surround- ing the center of the lens. For an extended source that is moved off the awns of the lens, the ring breaks up into arcs; if the quadrupole moment of the lens is large, one long arc dominates. A smaller source moved off the awns breaks up into four small unages, and two of the images merge (giving three images) and disappear (leaving one image) as the source moves farther from the optical axis. Therefore, a convenient classification of the lenses is into rings, arcs, multiples, and doubles, where the progres- sion is from sources close (compared to their extent) to the optical axis to far from the optical axis. Table 1 lists the known candidate systems. I have attempted to include all candidate systems that have appeared in the refereed literature, including some for which the evidence for tensing is not very strong. The field is changing rapidly, and it is somewhat a matter of judgment which systems are candidates, so my list may differ slightly from other published lists. I have been generous in attributing tensing characteristics, and some rather speculative systems are included in this list. In addition to the individual lens systems described above, two other signatures of gravitational tensing have recently been discovered, statistical tensing (Webster e' aL 1988) and microlensing (Irwin et al. 1989~. The energy and ingenuity of observers and theorists are beginning to be brought together in this new astrophysics laboratory, and some of the long discussed promise of gravitational tensing is being realized. We are beginning to gather clues about the distn~ution and nature of dark matter, both inside and outside galaxies, and there are real prospects for measuring the values of cosmological parameters and learning about the structure of quasars. Available space limits me to a discussion of only a few topics. MICROLENSING: MEASURING THE MASS FtJNCTION OF A GALAXY AND THE SIZE OF A QUASAR? Microlensing has been discussed extens~vetr in the literature, and was predicted to occur when a small lens (for example, a star) passes through the line of sight from the observer to the source and causes an apparent brightening of the source. The gravitational lens 2237+0305 is the system most likely to show microlensing ejects: the quasar images surround the central region of the tensing galaxies where the surface density of stars is high; He low redshift of the galaxy causes the characteristic angular tensing region of each microlens to be large; and the low redshift of the lens causes the apparent relative velocities of the observer, microlens, and source to be large (Kayser and Refsdal 1989~. Microlensing in principle can do a lot of astrophysics by allowing us to measure the number density
194 AMERICAN AND SOVIET PERSPECTIVES TABLE 1 P=pmal and verified gravitational lens systems, grouped according to morphology. For B1900~14, 2345+007, 1635+267, 1146+111, 0023+171, and 0249-186 it Is general!,, accepted that the existing evidence is sufficient to conclusively demonstrate that they are gravitational lenses. Image Flux lope of Discovery Separation1 Ratio: Lens Reference RINGS MG1131+0456 2.1" ~ 1 ? Hewitt et at 19~ MG1654+1346 2.1" ~ 1 Galaxy Langston et at 1989 ARCS Abell 370 ~ 50"2 Cluster Soucail et at 1987 Cl 2244 OF 22"2 Cluster Lynds and Petrosian 1989 Abell 963 30" Cluster Lave~y and Henry 1988 Cl 0500-24 52~2 Cluster Giraud 19~ Abell 2218 Arclets - Cluster Pello-Descayre et aL 1988 MULTIPLES 111S+080 0.5'' ~ 1 Galaxy Weymann et al 1980 2016~112 3.4" ~ 1 2 galaxies Lawrence et at 1984 2237~0305 1.8'' 1 Galaxy Huchra et al 1985 B1900~14 ? ~ 1.5 ? Pa~mski 1986 3C324 3" 1.7 Galaxy Le Fivre et at 1987 H14131117 0.8'' 1.1 Galaxy? Magain et al 1988 DOUBLES 0957~561 6.1'' 1.3 Galaxy, cluster Walsh et al 1979 2345~007 7.3" ~ 4 ? Weedman et aL 1982 1635~267 3.8" 4.4 Galaxy?? Djorgovski and Spinrad 1984 1146+111 157" ~ 1 ? Turner et aL 1986 0023+171 4.8'' 3 Galas Hewitt e' aL 1987 UM673 2.2" 7.6 Galaxy Surdej et at 1987 0249-186 2.0 - 2.6" 1.0-1.1 Cosmic string??? Cowie and Hu 1987 UM425 6.5" 70 Galaxy? Meylan and Djorgovs~ 1989 1 when there are more than two images, the image separation and flux ratio of the brightest pair are tabulated. Malice the radius of curvature of the dominant arm
HIGH-ENERGY ASTROPHYSICS 9.000 1 8.500 _ _ _ - D _ 18.000 - _ ~ A ~_ ~ B ·c 17.500 - ~_ ~_ 1 7.000 - - , 6.500 ~ TC 195 i I D IC Is I A LT D TIC -B 1 A 16.000 -1,, 1 1 1 1,, 1 1, 1,, i,,, 1,,,, 1,,,, 1,,,, 1,,,, 1,,,,,, i I i, I I , " " " 1 i; i , i 1 , ' 1 " " 0 200 400 600 800 1 OOC , 20C 1 400 Day ( 1 = 1 Janucry 198~) FIGURE 1 Light cunres of the quasar images of 2237+0305, constructed Mom the data of Schneider a at (1988; day 286), Wee (1988; day 998), and Irwin et al (1989; days 1325 and 1354~. Component A for day 286 has been onset 10 days to the right for clarity. The error bats represent the estimated errors in the absolute flux scales. Of compact objects inside galaxies, and to measure the size of the emitting region of quasars. For the geometry of the 2237+0305 system, the expected characteristic timescale for microlensing is l / M 6000 kmisec At= 8,7 M V years where V is the apparent transverse velocity. An apparent transverse velocity as large as 6000 km/see is expected for relative source, lens, and observer velocities of several hundred km/see; therefore, we may expect to find variations in the brightness of the images of 2237+0305 on times scales of months to years. "High amplification events" occur when a compact source crosses a caustic in the source plane. From the rise time of these events, the sue of the source can be measured. The overall time scale of the variations gives the mass of the tensing objects. Figure 1 shows a plot of a light curve of 2237+0305 constructed from the data of Irwin et aL (1989), Schneider et at (1988), and Wee (1988),
196 AMERICAN AND SOVIET PERSPECTwES including their estimated errors. A direct comparison of these data must be viewed with caution since the measurements were earned out with different instruments; furthermore, the measurements of the first two dates are of the Thuan and Gunn r magnitudes, and those of the last two dates are of the Mould R magnitudes. In any case, the evidence for microlensing in 2237~0305 is in the change in the magnitude of component A relative to the other quasar components. For example, the last two measurements show a change in the magnitude difference between components A and B. relative to the second measurement, of 0.38 and 0.26, well above the estimated error ~ the relative magnitudes (0.02 for the second measurement, and 0.05 for the last two measurements). Since the tune delay between the images is expected to be of order a day, the variation is probably due to microlensing rather than intrinsic variations in the quasar. The light curve is not well enough sampled to determine the mass of the microlens, nor whether we have witnessed a "high amplification event" in which the quasar passes behind a lens caustic. However, reasonable assumptions give a range in the estimated microlens mass of 0.0001M~, < M < 0.1M`3, but larger masses are of course consistent with the data (Invin et at 1989~. Models predict a large tensing optical depth, so continued, frequent monitoring of the quasar images of 2237+0305 is important. DO COSMIC STRINGS EXIST? If cosmic strings exist, they may be observable through their tensing effects. T ensed images caused by cosmic loops that have radii smaller than the image separation are likely to be difficult to distinguish from tensed images caused by centrally condensed mass distributions such as galaxies and clusters. Lensed images caused by strings with radii of curvature much larger than the image separation may be easier to distinguish because of the following properties unique to straight string lenses: (1) the images are not magnified; (2) the parity of the images is the same; (3) if there is a sufficient surface density of background sources, many pairs of images will stretch along the string; and (4) if the string is moving relativisticall~r, there is a small systematic velocity shift between the images on either side of the string (see Hogan 1987 and references therein). Cows and Hu (1987) discovered an unusual field of "twin galaxies" in which there are four pairs of gal~es with angular separations between 2.0" and 2.6", magnitude differences of 0.15 or less, and velocity differences (for the three pairs in which they have been measured) consistent with zero. More recent optical work (Hu and Cowie, private communications shows that one pair has significantly different colors, but has also discovered four more pairs with similar magnitudes and colors. Comparison with control fields shows that the number of pairs is much larger than would be expected by chance.
HIGH-ENERGY ASTROPHYSICS 197 In collaboration with R. Perley and E lbrner, I have acquired a A20 cm VISA image of the field. A contour plot of the radio emission (henceforth referred to as 0249-184) is shown in Figure 2, superimposed on an optical image (ldudly provided by E. Hu) for comparison One may use the radio data to address the question of tensing by a cosmic string in two ways. First, is the emission peculiar in some way that would be explained by tensing? The answer to this question is no. The coincidence of the 0249-184 and the optical galaxy implies the two are physically associated. The radio emission has Fanaroff and Riley Type I (FRT, Fanaroff and Riley 1974) morphology which is common among radio galaxies. The radio power implied by the redshift of the optical galaxy falls within the range commonly seen in FRI sources (Shaver et at 1982), though it is brighter than that usually seen in optically selected elliptical galaxies JIummel e! al. 1983~. If a string fell in front of the jet, one would expect to see a distortion of the type sketched in Figure 3; none is seen. Second, do the properties of 0249-184 exclude the possibility that there is a cosmic string in the field? Our judgment in this case is that the answer is probably, but not definitely. At galaxy A2, the radio emission is at the position of optical emission that is "known" to be doubly imaged. The lack of corresponding doubling of the radio emission is Cadence against the tensing interpretation. However, given the uncertainly in the registration of the optical and radio images (about 0.6") and their finite resolution, it is possible that the radio emission falls just outside the doubly-imaged region. In summary, the case for tensing has been weakened by the radio data and by the discovery of color differences in one galaxy pair; however, the evidence may not rule it out, and the field has peculiar properties that would be well explained by string tensing. In any case, gravitational tensing brings a phenomenon predicted by theories of the earn universe under observational scrutinity. MASS-TO-LIGHT RATIO MEASUREMENTS IN GALAXIES Gravitational tensing is one of the few phenomena in astrophysics in which the system under study is not necessarily luminous, and is therefore well suited to studying dark matter. Gravitational tensing has been used to provide an independent measure of the mass-to-light ratio in two systems in which the mass distribution is reasonably well constrained by surrounding images and in which the redshifts of both the source and the lens are known. The first system is 2237+0305, already discussed in the context of microlensing above. The foreground galaxy has a small redshift (z = .0394), and its surface brightness can be used directly in lens models. Four images of the background quasar, surrounding the central region (radius < 0.5 kpc) of the galaxy, constrain models of the mass distn~ution in that region Schneider e' al (1988) calculated models based on the observed
198 E C L I N T26 00 I o N05 -18 25 4s S0 S~7 10 _ 15 ~1~;' 1 1 ~' ~ AMERICAN AND SOVIET PERSPECTIVES 1 0 1 1 1 - -- 1 O ;~ - .,_~ V ,-~;r, . . . 02 49 24. B 23. S 23.0 22.S 22.B 21.S 21.0 20.S RI GHT ASCENSI ON FIGURE 2 Contour plot of a )20 cm radio map (solid contoured of 0249-184 superimposed on an optical I band image (dotted contours) of the field. The radio contours are 20%0, 30~o' 40~o, 50~O, 60%, 70%, 80~o, and 9O~o of the peak brightness of 370 ~Jy/beam; the resolution is apprmimateh~r 2". No attempt has been made to calibrate the optical image; the contours are linear in the CCD data numbers. light distn~ution, ~ a 12~' x 12)' region centered on the bulge of the galaxy, and a constant mass-to-light ratio. The best model has a blue mass-to- light ratio of 9.4h (Ho = lOOh lun/sec/Mpc) with 20% variations causing significant differences between the data and the model. MG1654~1345 is the second Einstein ring to be discovered, and consists of the ring image of the radio lobe of a z = 1.74 quasar tensed by a z = 0.254 galaxy (Langston et al 1989~. The strength of the lens in this case can be estimated i rom the angular radius of the observed ring, and the enclosed mass calculated from the measured redshifts, assuming q0 = 1/2, is 9.5 ~ 1.9 x 10~° h Me. From the measured brightness of the galaxy, the blue mass-to-light ratio with the central 5 l~c of the galaxy is 19 ~ 4h. The above values of the mass-to-light ratio are completely independent of measurements made using dynamical techniques. In addition to using gravitational lens systems as astrophysics laborato- ries, work continues on searches for new systems. I know of six deliberate searches at radio and optical wavelengths for which results have been pub- lished. Gravitational tensing is a relatively rare event; therefore, all these searches require some sort of filter to select objects which we believe are
HIGH-ENERGY ASTROP~ICS 199 · . . am. Ernst ~ r ········ ~c_~,JI_e~,- ············- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . · · · · · . · . · · ma- . .,_,% . lo, _ . . . . . . . . . . . . . FIGURE 3 Schematic diagram of tensing of a linear structure, such as a radio jet, by a cosmic string. Ibe angle between the string and the let is 9, and the part of the sly that is imaged twice is represented by the shaded region. A point source in the shaded region has two images with angular separation /~. The cosmic stung causes an apparent break ~ the jet with separation s and overlap 1. more likelier to be tensed. More than half the searches use the morphology of the object as the criterion in selecting lens candidates. The others use the fact that gravitationally tensed images, because of their magnification, will appear brighter than they otherwise would. This property indicates that objects of high absolute luminosity (calculated from the redshift and the apparent l~'nlinosity, not corrected for the magnification of any lens) are good lens candidates. Once the lens candidates have been selected, further tests must be camed out to test whether images are tensed. Possi- ble tests include: (1) Do the images have the same spectra (and the same
200 AMERICAN AND S0~7ET PERSPECTIVES redshift)? (2) Do the images have the same polarization properties? (3) If there is any structure in the images, is it consistent with gravitational tensing? (4) Do We images have the same light curve, offset by the the delay between the images? (5) If there are other distant objects behind the lens, are they also imaged? (6) Are there more than two images of the source? (7) Is there a lens visible? (8) Is the source redshift larger than the lens redshift? It is a question of judgment at what point one decides that a system is a gravitational lens, and opinion on this issue vanes somewhat. In many cases, optical spectra showing emission lines at the same wavelengths with approximately the same relative intensities have been the primary evidence for gravitational tensing. Most of the searches rely on optical spectroscopy as a means of verification. However, there is at least one clear counterexample. In the double quasar 1145-a71, the mo components have very similar optical spectra, but one is a radio source and the other is not (Djorgovski et al. 1987~. The following is a brief summary of the gravitational lens searches. (1) VLA Sun ey This search makes use of a large program of VLA snap shots (Hewitt et aL 1989; Lawrence et at 1986i) of radio sources from the MIT-Green Bank (MG) single-dish survey (Bennett et at 1986). The ob- senations are at 5 GHz, giving a resolution of approximately 0.4". Sources are selected as lens candidates on the basis of their radio morphology. The goal of the project is the detection of multiply-imaged quasars, since these are likelier to be verifiable with optical and very long baseline ~B) radio observations. Therefore, sources with more than one unresolved comply Dent are given the highest prionty. Four gravitational lens systems have resulted from the suney: 2016+112, MG1131~0456, MG1654+1346, and MG0414+0534 (Hewitt et aL 1989). Leo are Einstein ring images of radio lobes; the other Go are multiple images of compact stellar objects. A fifth source, 0023~171, shows Go stellar objects with similar spectra, but the radio morphology is not easily explained through gravitational tensing, and its interpretation remains uncertain JIewitt et at 1987~. Four of these objects have been detected on VLBI baselines. (2) High Luminosity Quasars I Surdej et at (1988) selected 111 quasars from the Veron et Veron (1985) catalog by the following criteria: ma < 18.5, Mv ~-29.0, and declination < 20°. These quasars were imaged with the Z2m ESO/MPI telescope and a COD camera, often under conditions of good seeing. I~enty-five candidates appear "interesting" in that they display multiple structure or are near a faint galaxy, and the lens systems UM673 and H1413+117 have been discovered. The evidence for tensing is good in both cases; in addition to the spectroscopic evidence, the tensing galaxy of UM673 has been detected, and H1413+117 shows four tensed images.
HIGH-ENERGY ASTROPHYSICS 201 (3) High Luminosity Quasars II Djorgovski and Meylan (1989) have selected high luminosity quasars with large redshifts from the Hewitt and Burbidge (1987) catalog and are carrying out CCD observations. So far one probable gravitational lens has been discovered: UM425 with four components, of which two show very similar spectra (Meylan and Djorgovski 1989~. Spectra of the other two components have not yet been measured. (4) Pairs of Quasars I Webster and collaborators are using the Automated Plate Measuring machine in Cambridge to scan direct and objective prism plates (Webster e! at 1988~. Quasar candidates are selected on the basis of their spectra, and the morphology of the quasar candidates is known from the scanning of the direct plates. Note that quasar candidates are not limited to just stellar objects; multiply-imaged quasars and quasar-gala~y associations are included in the sample. This survey has resulted In the detection of statistical tensing described above, and one multiply imaged quasar (Hewett et aL, preprint). (5) Pairs of Quasars II Weedman and Djorgovsld (1988) examined seven grens plates for close pairs with similar spectra. Eight candidate tensed pairs of images, with separations ranging from 4" to 9.S" were found through visual inspection. COD images and spectra for eight of the pairs were obtained, and it was found that none is tensed. From the area of the sly surveyed and quasar counts, Weedman and Djorgovski estimate that 200 to 300 quasars were examined, and from the qualibr of their plates estimate the limits on the frequency of tensing. They find their results are consistent with the theoretical results of Turner et at (1984), but are perhaps surprising if the known wide separation lens candidates really are gravitational lenses. (6) Pairs of Blue Objects Reboul et at (1987) have selected pairs of blue objects from several catalogs. The 62 candidate lens systems consist of 46 pairs of blue objects separated by less then 9", and 16 pairs in which one of the objects Is blue. Fifteen of the systems have been investigated spectroscopially, but none is tensed. In addition to the lens searches described above, there are a number that are in preluninarg stages. There are at least three other searches in grens and objective prism plates (see Webster and Hewett 1989 and references therein). Gorenstein, Elby, Rogers, and myself are examining archived Mark III VLBI data for tensed compact radio sources. The existing data can be reprocessed to extend the search in interferometric delay and rate so that regions in the sly of typically several arc seconds on a side are examined. The advantage of this technique over the VLA suIvey technique is that a major component of confusion, classical double radio sources, are resolved out. Hogan (1987) has proposed a search for tensing by cosmic strings in CCD data collected for other purposes. The expected signature
202 AMERICAN AND SOVIET PERSPECTIVES of multiple pairs of galaxies stretching along the string is well suited to an automatic search. Burke, Turner, and Gott have observed a sample of high luminosity quasars at the VLA looking for radio structure that might be a product of gravitational tensing. 13 son, Fort, Mellier, Turner, and collaborators are surveying clusters of galaxies for evidence of luminous arcs. In summary, the searches for gravitational lenses are proving to be successful, and more lenses continue to be discovered serendipitous Many searches are under way, and new instruments that will routines increasely the resolution of astronomical imaging (such as the Hubble Space Telescope and the Very Long Baseline Array), and automated data analysis techniques may greatly increase their yield. A wide varieW of types of lenses are being discovered, many particularly well suited to a specific application, and we are beginning to see result of astrophysical interest. ACKNOWLEDGMENT This work was supported by grant AS1~6-18257 from the National Science Foundation. REFERENCES Bennett, C L, Lawrence, C R. Burke, B. F., Hewitt, J. N., and Mahoney, J. H. l9B6; Ap. J. Supp; 61; 1. Blandford, R D., and Kochanek, C. S. 1987; Ap. J.; 321; 658. Cowie, L L., and Hu, E. M. 1987; Ap. J.; 318; 133. Djorgovski, S., and Meylan, G. 1989, in Gravitational Lenses, Lecture Notes in Physics, Vol. 330 teds. 3. M. Moran, J. N. Hewitt, and K.-Y. Lob, Berlin: Springer-Verlag. Djorgovsld, S., Parley, R. Meylan, G., and McCarthy, P. 1987; Ap. J. Lett.; 321; L17. Djorgovski, S., and Spinrad, H. 1984; Ap. J. Lett; 282; L1. Fanaroff, B. id, and Riley, J. M. 1974; M.N.R.^S.; 167; 31p. Giraud, E. 1988, Ap. J. Lett.; 334; L69. Hewitt, J. N., Burke, B. F., lbrner, E. L-, Schneider, D. P., I>wrence, C R. Langston, G. I., and Brady, J. P. 1989, in Gravitational Lenses, Lecture Notes in Physics, Vol. 330 feds. J. M. Moran, J. N. Hewitt, and K-Y. Lo), Berlin: Spnnger-Verlag. Hewitt, As, and Burbidge, G. 1987; Ap. J. Supp.; 63; 1. Hewitt, J. N., lbrner, E. L~, I~wrenct:, C R., Schneider, D. P., Gunn, J. E., Bennett, C L, Burke, B. F., Mahoney, J. H., Langston, G. I., Schmidt, M., Oke, J. B., and Hoessel, J. G. 1987; Ap. J.; 321; 706. Hewitt, J. N., lDrner, E. L, Schneider, D. P., Burke, B. F., Langston, G. I., and I=wrenoe, C R 1988; Nature; 333; 537. Hogan, C J., 1987, in 13th Texas Symposium on Relativistic Astrophysics (ea. M. P. Ulmer), Singapore: World Scientific Publishing Company Huchra, J., Gorenstein, M., Kent, S., Shapiro, I., Smith, G., Homne, E., and Perter, R. 1985; ~ J.; 90; 691. Hummel, E., Kotanyi, ~ G., and Ekers, R. D. 1983; Astron. Astroph.; 127; 205. Irwin, M. J., Webster, R. L, Hewett, P. C., Comgan, R 1:, and JedrzeJewski, R. I. 19B9, ~ J., in press. Kayser, R. and Refsdal, S. 1989; Stature; 338; 745. Kochanek, C S., and Blandford, R. D. 1987; Ap. J.; 321; 676.
HIGH-ENERGY ASTROPHYSICS 2~)3 Narayan, R., and Grossman S. 1989, in Gravitational Lenses, Lecture Notes in Physics, Vol. 330 feds. J. M. Moran, J. N. Hewitt, and K-Y. Lo), Berlin: Spnnger-Verlag. Langston, G. I., Schneider, D. P., Conner, S., Canlli, C L, Lehar, J., Burke, B. F., liner, E. L*, Gunn, J. E., Hewitt, J. N., and Schmidt, M. 1989; ~ J.; 97; 1283. Lavery, R J., and Henry, J. P. 1988; Ap. J. Lett.; 329; 1~1. Lawrence, C R. Bennett, C. L-, Hewitt, J. N., Langston, G. I., Klotz, S. E., and Burke, B. F. 19K; Ap. J. Supp.; 61; 105. Lawrence, C. R., Schneider, D. P., Schmidt, M., Bennett, ~ L, Hewitt, J. N., Burke, B. F., lbrner, E. Lo, and Gunn, J. E. 1984; Science; 223; 46. Le Fe - :, O., Hammer, F., Nottale, L., and Mathez, G. 1987; Nature; 326; 268. Lynds, R., and Petrosian, V., 1989; Ap. J.; 336; 1. Magain, P., Surdej, J., Swings, J.-P., Borgeest, U., Kayser, R., Kuhr, H., Refidal, S., and Remy, M. 1988; Nature; 334; 325. Meylan, G., and Djorgovski, S. 1989; Ap. J. Lett.; 338; L1. Pac~mski, B. 1986; Ap. J.; 308; L43. Pello-Descayre, R. Soucail, G., Sanahuja, B., Mathez, G. and Ojero, E. 1988; Astron. Astroph.; 190; L11. Reboul, H., Vanderriest, C., Pnngant, A. M., and Cayrel, R. 1987; Astron. Astroph.; 177; 337. Schneider, D. P., lbrner, E. L, Gunn, J. E., Hewitt, J. N., Schmidt, M., and I~wrence, C. R 1988; ~ J.; 95; 1619. Shaver, P. A, Danziger, I. J., Ekers, R. D., Fosbu~y, R ~ E., Goss, W. M., Malin, D., Mo~o~wood, ~ F. M., and Wall, J. V. 1982, in Extragalactic Radio Sources (ea. D. S. Heeschen and C M. ~de), Dordrecht: D. Reidel Publishing Company. Soucail, G., Fort, B., Mellier, Y., Picat, J. P. 1987; Astron. Astroph.; 172; L14. Surdej, J., Swings, J.-P., Magain, P., Borgeest, U., Kayser, R. Refsdal, S., Cou~voisier, C J.-L, Kellermann, K I., and Kuhr, H. 1988 in Proceedings of a Worlcshop on Optical Sunreys for Quasam (eds. P. S. Osmer, A C Porter, R F. Green, and C B. Foltz), San Francisco: Ast~nomical S~ociety of the Pacifi~ Surdej, J., Magain, P., Swings, J.-P., Borgeest, U., Courvoisier, 11 J.-L, Kayser, R. Kellennann, K I., Kuhr, H., and Refsdal, S. 1987; Nature; 329; 695. Ih~ner, ~ L-, Ostriker, J. P., and Gott, J. R. III. 1984; Ap. J.; 284; 1. I~ner, E. L, Schneider, D. P., Burice, B. F., Hewitt, J. N., Langston, G. I., Gunn, J. ~, Lawrenoe, C R., and Schmidt, M. 1986; Nature; 321; 14Z Veron-Cet~, M.-P., and Veron, P. 1985, ESO Sci. Rep. No. 4. Walsh, D., Ca~swell, R. F., and We3rmann, R. J. 1979; Nature; 279; 381. Webster, R. L, and Hewett, P. ~ 1989, in Gravitational kenses, Lecture Notes in Physics, Vol 330 (ed~ J. M. Moran, J. N. Hewitt, and K.-Y. Lo), Berlin: Spnnger-Verlag. Webster, R L, Hewett, P. C., Harding, M. E., and Wegner, G. ~1988; Nature; 336; 358. Webster, R L, Hewett, P. C, and I~win, M. 3. 1988; A J.; 95; 19. Weedman, D. W., and Djorgovski, S. 1988, in Proceedings of a Workshop on Optical Surve~ for Q~m (eds. P. S. Osmer, A C Porter, R. F. Green, and ~ B. Folts3, San Francisoo: Astronomical Some~ of the Pacific Weedman, D. W., Weymann, R. J., Green, R. F., and Heckmann, 1: M. 1982; Ap. J. Lett.; ~5; LS. Weymann, R J., Lath~sr, D., Angel, J. R P., Green, R. F., Liebert, J. W., Turnshek, D. ~, lbrnshek, D. E., and ~son, J. ~ 1980; Nature; 285; 641. Yee, H. K C 1988; A~ J.; 95; 1331.