I have mixed feelings about giving the opening talk at this symposium. On the one hand I am delighted to welcome you to this conference on the impact of VLBI on astrophysics and geophysics. On the other hand I am very sad that Ed Purcell is bothered by a bad back and cannot give the introduction as originally scheduled. Ed is one of the truly great of the world's physicists and one of the nicest persons ever born. My substituting for Ed is somewhat like Bill Buckner pinch hitting for Lou Gehrig. (Those who do not understand this reference to American baseball might have a better appreciation after viewing tonight's Red Sox game.)

There are now at least 23 radio sources in which apparent superluminal motion has been found. The effect is not limited to the “classic” examples like 3C 345, but is found in objects with a wide range of characteristic properties. We review the observations of the phenomenon and present an up-to-date list of the known superluminal sources.

We discuss VLBI observations of the bright core in the large, triple radio source 4C 34.37, associated with a quasar at z = 0.206. We have discovered superluminal motion in this quasar, which is the largest quasar known. We comment on the implications of these findings, and subsequently examine the projected size distributions of lobe and core dominated quasars in general. The results of these investigations lead us to conclude that neither superluminal motion, nor core prominence can simply be translated into an orientation of the overall radio source.

The well known superluminal source 3C345 is associated with a 16th magnitude quasar at z=0.595. The optical continuum is highly polarized and shows violent variations. At x-ray frequencies it is a weak source; while at radio frequencies it is bright and variable. The radio structure consists of a 200 kpc halo, a 20 kpc jet, and a very bright pc scale region which shows superluminal motion. Recent MkII VLBI monitoring (Biretta, Moore, Cohen 1986) at 2, 5, 11, and 22 GHz has provided accurate spectra, sizes, expansion rates, velocities, and trajectories for the superluminal knots and the “core.” Here we consider physical conditions in the emission regions and models for the kinematics.

Until recently, the compact structure in 3C 395 could be described as follows (Simon et al. 1987):

1. (1) A bright, unresolved, flat spectrum core to the northwest. The core brightened by about 0.4 Jy between 1983.3 and 1985.4, from 0.6 Jy to 1.0 Jy.

2. (2) A moderately resolved component 15.8+/−0.2 mas to the southwest along P.A.I 18o. The position of this component relative to the core has changed by less than 0.2 mas in 6 years and the flux density of this component has remained roughly constant.

3. (3) A component which was moving rapidly away from the core towards component (2) with a proper motion of 0.64+/−0.1 mas yr−1. For 3C 395 (z=0.635), this implies (v/c)∼20c for Ho =100 km/s/Mpc and qo =0.5.

The μ–z diagram (Figure 1) plots the observed internal proper motion μ versus redshift z for 32 extragalactic radio sources associated with active galactic nuclei. The observed points fall below an upper bound which decreases with redshift; there is a statistically significant anticorrelation between redshift and internal proper motion.

The standard model for apparent superluminal motion is that of relativistic motion in a jet pointed nearly along the line of sight. Variants of this model give a natural explanation for the extremely asymmetric morphology observed in the radio cores of these sources. Despite this, examination of the large scale structure in superluminal radio sources strongly suggests that large scale symmetrical structure with two large radio lobes placed on either side of the compact emission is a common phenomenon among these objects. When a comparison is done between the large scale structure typically seen in double quasars and that seen in superluminal sources, it is found that the de-projected physical sizes estimated for superluminal quasars are similar to the size scales observed in the double quasars. Large scale radio structure is probably as common among apparent superluminal sources as it is in radio quasars.

We present a new high dynamic range map of the quasar 3C 273, made from observations with a VLBI network of 12 telescopes. This new map at 18 cm wavelength has one of the highest dynamic ranges yet achieved with VLBI, and it shows the ‘jet’ extending to at least 180 milliarcsec, or 330 pc from the nucleus of the quasar. Strong limits can be placed on the brightness of any ‘counter-jet’ on kiloparsec-scales, as no emission is visible on the opposite side of the ‘core’. Combining with other VLBI, VLA and MERLIN maps shows that the jet is visible and continuous over a very large range of scales, from 1 pc to 40 kpc.

The jet in the superluminal source 3C 120 has been observed on all scales from under a parsec to about 100 kpc. The jet brightness decays rather smoothly with core distance, indicating that the basic characteristics of the jet are established on parsec or smaller scales and evolve simply with distance.

3C 120 is a nearby radio galaxy (z = 0.033) that shows superluminal motions in its radio jet within one parsec of the nucleus. The apparent velocities of the moving features are approximately 1.5 mas/yr. This object is of particular interest because it is clearly identified with an N-type galaxy, unlike all other superluminal radio sources which are only seen in quasars and BL Lac objects. Because of the close proximity of 3C 120, we have been able to make a series of VLBI and VLA images at stepped resolutions, and thereby map the radio jet continuously from the radio nucleus (0.5 h−1 parsec) into the diffuse 400 h−1 kpc structure.

Hybrid maps of 3C 273 and DA 193 at 2.3 and 8.4 GHz have been produced with VLBI data acquired during a Crustal Dynamics campaign. By comparing to maps at previous epochs, superluminal components of 3C 273 are clearly detected. DA 193 is relatively compact. Structure corrections for the VLBI delay used in astrometry have been estimated from these maps. Some corrections are significant when compared with the precision of models used in astrometry and geodesy.

The author and his collaborators have observed NRAO 140 twice at X-ray energies and numerous times with multifrequency VLBI. The VLBI observations reveal a knotty jet structure with superluminal motion of the innermost two knots relative to the core. The VLBI core decreased by about a factor of 2 in flux density between 1980 and late 1984. The X-ray flux also declined by about the same factor during this period. Monitoring at 18 cm during periods of low-frequency variability has revealed pronounced changes in the relative brightnesses of the components of the source while the total flux density has varied by ≲ 10%.

The galaxy M87 (3C274, NGC4486) contains one of the nearest jet sources in the northern hemisphere, and one of our best opportunities to study a jet in detail. We have made MkII VLBI observations of the nuclear jet at 1.66 GHz using 18 stations of the “World Array.” The array included Arecibo which greatly improved the north-south resolution. The data were global fringe fitted, and self-calibrated to remove both antenna and baseline amplitude and phase errors. The resulting map (Fig. 1) has a dynamic range of 2300:1 and reveals a very complex structure. At distances ≥50 mas from the core, the jet is resolved across its width and appears limb-brightened. The brightness centroid oscillates from side to side by an amount roughly equal to the jet's width; this may be caused by brightness variations at the edges of a limb-brightened jet. Within 50 mas of the core the opening angle (at the FWHM points) is 6.5°±0.3°; this is larger than that of the 20″ jet, and suggests there is recollimation between the pc and kpc scales. The jet is aligned within 1° of the 20″ jet. There is no evidence for a counter-jet; the jet/counter-jet brightness ratio is R≥200.

The radiosource 4C39.25 (0923+392) is an 18 mag quasar with a redshift of 0.699. Throughout the seventies its compact structure at centimetric wavelengths consisted of 2 components separated by an angle of about 2 milliarcseconds (mas). In the eighties the structure has been found to consist of 3 components (Marcaide et al., 1985; Shaffer et al., 1987). The angular separation between the “a” and “c” components (see Fig. 1, which shows our most recent map of this source made at a wavelength of 2.8 cm) has remained constant at about 2 mas, while component “b” is separating from component “c” at a rate which we estimate from 2.8 cm observations as 0.18± 0.01 mas/yr, and which translates into an apparent linear separation velocity of 4.0 ± 0.2 c (HO=100 km/s/Mpc, qO=0.5, c is the speed of light).

VLBI observations at 18 and 6 cm with European baselines of lengths up to 15 and 23 million wavelengths respectively have been used to search for structure in the quasar 4C39.25 on scales of 10 to 100 milliarcsec. No significant structure on these scales brighter than 5 percent of the peak brightness has been detected. The core itself, however appears to be extended on a scale of 2 to 4 milliarcsec and variable in flux density.

During the 1970's, the compact radio structure of 4C39.25 consisted of a simple double source, whose separation did not change. A new component, which probably originated in mid-1979, is now seen to be moving superluminally between the original, still stationary, components. It is not clear which component of the source is the core, nor how the moving component will interact with the stationary component that it is now approaching on a collision course.