|Annu. Rev. Astron. Astrophys. 1999. 37:
Copyright © 1999 by . All rights reserved
Expansions at up to ten or more times the speed of light have been observed in quasars for more than 20 years (Pearson & Zensus 1987, Zensus 1997). At first these superluminal motions provoked concern because they appeared to violate relativity, but they were soon interpreted as illusions due to relativistic aberration (Rees 1966). However, the ultimate physical interpretation had remained uncertain. In the extragalactic case the moving jets are observed as one-sided (because strong Doppler favoritism renders the approaching ejecta detectable) and it is not possible to know if superluminal motions represent the propagation of waves through a slowly moving jet, or if they reflect the actual bulk motion of the sources of radiation.
In the context of the microquasar analogy, one may ask if superluminal motions could be observed from sources known to be in our own Galaxy. Among the handful of black holes of stellar mass known so far, three transient X-ray sources have indeed been identified at radio waves as sporadic sources of superluminal jets. The first superluminal source to be discovered (Mirabel & Rodríguez 1994) was GRS 1915+105, a recurrent transient source of hard X-rays first found and studied with the satellite GRANAT (Castro-Tirado et al 1994, Finoguenov et al 1994). The discovery of superluminal motions in GRS 1915+105 stimulated a search for similar relativistic ejecta in other transient hard X-ray sources. Soon after, the same phenomenon was observed by two different groups (Tingay et al 1995, Hjellming & Rupen 1995) in GRO J1655-40, a hard X-ray nova found with the Compton Gamma Ray Observatory (Zhang et al 1994). A third superluminal source may be XTE J1748-288 (Hjellming et al 1998), a transient source with a hard X-ray spectrum recently found with XTE (Smith et al 1998).
GRS 1915+105 is at ~ 12 kpc from the Sun (Rodríguez et al 1995, Chaty et al 1996) on the opposite side of the galactic plane and cannot be studied in the optical. Given the large extinction by dust along the line of sight (Mirabel et al 1994, Chaty et al 1996), the precise nature of the binary has been elusive. Castro-Tirado et al (1996) proposed that GRS 1915+105 is a low mass binary, while Mirabel et al (1997) proposed that it is a long period binary with a companion star of transitional spectral type. From the nature of the line variability in the infrared, Eikenberry et al (1998b) propose that the emission lines in GRS 1915+105 arise in an accretion disk rather than in the circumstellar disk of an Oe/Be companion (Mirabel et al 1997). GRS 1915+105 has similarities in the X-rays and gamma-rays with GRO J1655-40 and other black hole binaries, and it is also likely to harbor a black hole (Greiner et al 1996). The X-ray luminosity of GRS 1915+105 (reaching 2 × 106 solar luminosities) far exceeds the Eddington limit (above which the radiation pressure will catastrophically blow out the external layers of the source) for a 3 solar mass object, which is 105 solar luminosities. Furthermore, it shows the typical hard X-ray tail beyond 100 keV seen in black hole binaries (Cordier et al 1993, Finoguenov et al 1994, Grove et al 1998). Finally, it is known that the absolute hard X-ray luminosities in black hole systems are systematically higher than in neutron star systems (Ballet et al 1994, Barret et al 1996), another result that points to a black hole in GRS 1915+105.
GRO J1655-40 is at a distance of 3.2 kpc and the apparent transverse motions of its ejecta in the sky are the largest yet observed (Tingay et al 1995, Hjellming & Rupen 1995) until now from an object beyond the solar system. It has a bright optical counterpart and consists of a star of 1.7-3.3 solar masses orbiting around a collapsed object of 4-7 solar masses (Orosz & Bailyn 1997, Phillips et al 1999). The compact object is certainly a black hole, since its mass is beyond the theoretical maximum mass limit of ~ 3 solar masses for neutron stars (Kalogera & Baym 1996).
King (1998) proposes that the superluminal sources are black hole binaries with the secondary in the Hertzsprung-Russell gap, which provides super-Eddington accretion into the black hole. In the Galaxy there would have been 103 systems of this class with a lifetime for the jet phase of 107 years, which is the duration of the spin-down phase of the black hole.