Annu. Rev. Astron. Astrophys. 1981. 19: 373-410 Copyright © 1981 by Annual Reviews. All rights reserved

6.2 Superluminal Sources

About a decade ago, observations of 3C 273 and 3C 279 made over a single transcontinental baseline showed that the structure of both sources had clearly changed during only a four-month period. The limited data then available could be most simply described by a simple double source whose components had separated with apparent velocities well in excess of the speed of light (Cohen et al. 1971, Whitney et al. 1971). During the first few years following the discovery of these apparent superluminal motions, interpretations in terms of fixed components that vary in flux density appeared attractive (Cohen et al. 1971, Dent 1972a, b, Kellermann et al. 1974), although somewhat too contrived to be consistent with the observed variations in the total flux density and the apparent absence of any systematic contractions. At the time the superluminal motion appeared surprising, although it might have been predicted as a direct consequence of the observed rapid flux density variations.

Since then considerably more data have been accumulated on the sources 3C 120, 3C 273, 3C 279, and 3C 345 (Wittels et al. 1976b, Cohen et al. 1977, Kellermann & Shaffer 1977, Seielstad et al. 1979, Cohen et al. 1979, Cotton et al. 1979, Pauliny-Toth et al. 1981). The cumulative evidence for component motions with apparent velocities up to 45 c is impressive although, with a few important exceptions, largely circumstantial. In the case of 3C 120 and 3C 279 it is necessary to postulate multiple events in order to interpret the apparent changes in the separation with time in terms of separating motion alone.

The model of multiple outbursts is consistent with the observed total flux density variations as well as the VLBI data (Seielstad 1974, Pauliny-Toth et al. 1981). The possibility of apparent contractions is, however, not unequivocally excluded, and this detracts somewhat from the arguments against interpretations in terms of flux density variations of fixed components, which are based on the absence of such contractions.

The quasars 3C 273 and 3C 279 and the galaxy 3C 120 are located at low declinations and, with the limited number of (mostly east-west) baselines used in VLB interferometer observations, it is difficult to construct a reliable two-dimensional picture. Particularly in the case of 3C 273, the early evidence for superluminal motion rested primarily on a detailed analysis of the manner in which the fringe visibility changes with time, rather than from a direct comparison of the structure observed at different epochs (Schilizzi et al. 1975, Cohen et al. 1977, Seielstad et al. 1979, Cohen et al. 1979). Legg et al. (1977) were able to fit all of the data between 1970 and 1973 with a simple, fixed-component model and found no need for superluminal velocities during that period. But as shown in Figure 6, more recent hybrid maps of 3C 273, made over a three-year period, show clearly the motion of the "jet" component away from the core with a velocity ~ 10 c (Pearson et al. 1981).

The evidence for superluminal motion in the quasar 3C 345 came originally from observations made during the period 1970 to 1977 which have been interpreted in terms of a double source whose components appeared to separate at a constant angular rate of about 0.15 milliarcsec per year, corresponding to a linear velocity about seven times the speed of light (Cohen et al. 1976, Wittels et al. 1976a, b, Cohen et al. 1977). After 1977, however, the available data indicate a more complex structure, most simply described by a bright double core plus an elongated low surface brightness jet-like component (Readhead et al. 1979, Bååth et al. 1981, J. Spencer and K. Johnson, in preparation, Cohen et al. 1981). For a two-year period from early 1978 to early 1980, there was no change the separation of the core components. Although the jet-like component was too diffuse to enable specifying its position accurately, it appears to have moved away from the core with an apparent velocity of about 20 c (Schraml et al. 1981).

Figure 6. Contour maps showing the superluminal component separation in 3C 273, from Pearson et al. (1981).

Although the possibility of flux variations in fixed components giving the illusion of rapid component motion cannot be completely excluded, attempts to do so are very artificial in view of the great wealth of data suggesting moving components. The apparent motion may be summarized as follows (Kellermann & Shaffer 1977, Cohen et al. 1977, Seielstad et al. 1979):

(a) The apparent component separations measured at different wavelengths show small, but significant differences. Observations at widely different wavelengths show very different component sizes and separations with the largest and most distant components being found at the longer wavelengths. In tracing the structure out from the core, the elongated structure may curve through an angle as much as 90 degrees up to a few seconds of arc away from the core.

(b) Apparent changes in component separation up to a factor of 10 are observed with apparent linear velocities up to 45 c.

(c) The apparent component velocities and direction of motion remain constant for as long as individual components can he recognized.

(d) In two sources, 3C 120 and 3C 279, where there has been more than one outburst, the direction of apparent component motion remains constant, but the apparent velocity can differ.

(e) There are no observed contractions, only separations; however, this conclusion is based on the association of occasional sudden decreases in observed separation with the beginning of new outbursts.

(f) Extrapolation of the component separation backwards in time to zero spacing is usually coincident with a flux density outburst, but in some cases of multiple outbursts, such as 3C 120, the association, while self-consistent, is not unique.