Theoretical ideas aside, is there any observational evidence for black holes? Can we prove that these strange invisible objects really do exist?
One way in which we might think of detecting a black hole would be to observe it transit (pass in front of) a star. Unfortunately, such an event would be extremely hard to see; the 12,000-km planet Venus is barely noticeable when transiting the Sun, so a 10-km-wide object moving across the image of a faraway star would be completely invisible with either current equipment or any equipment available in the foreseeable future.
Actually, this observation is not even as clear-cut as just suggested. Even if we were close enough to the star to resolve the disk of the transiting black hole, the observable effect would not be a black dot superimposed on a bright background. The background starlight would be deflected as it passed the black hole on its way to Earth, as indicated in Figure 22.17. The effect is the same as the bending of distant starlight around the edge of the Sun, a phenomenon that has been repeatedly measured during solar eclipses throughout the last several decades (see More Precisely 22-2). With a black hole, much larger deflections would occur. As a result, the image of a black hole in front of a bright companion star would show not a neat, well-defined black dot but rather a fuzzy image virtually impossible to observe, even from nearby.
Figure 22.17 The gravitational bending of light around the edges of a small, massive black hole makes it impossible to observe the hole as a black dot superimposed against the bright background of its stellar companion.
A much better way to find black holes is to look for their effects on other objects. Our galaxy harbors many binary-star systems in which only one object can be seen. Recall from our study of binary-star systems in Section 17.9 that we need to observe the motion of only one star to infer the existence of an unseen companion and measure some of its properties. In the majority of cases, the invisible companion is simply small and dim, nothing more than an M star hidden in the glare of an O or B partner, or perhaps shrouded by dust or other debris, making it invisible to even the best available equipment. In either case, the invisible object is not a black hole.
A few close binary systems, however, have peculiarities suggesting that one of their members may be a black hole. Some of the most interesting observations, made during the 1970s and 1980s by Earth-orbiting satellites, revealed binary systems in which the invisible member emits large amounts of X-rays. The mass of the emitting object is measured as several solar masses, so we know it is not simply a small, dim star, and radiation pressure from the binary members makes circumstellar debris an unlikely explanation for its invisibility.
One particular binary system drawing much attention lies in the constellation Cygnus. Figure 22.18(a) shows the area of the sky in Cygnus where astronomers have reasonably good evidence for a black hole. The rectangle outlines the celestial system of interest, some 2000 pc from Earth. The black-hole candidate is an X-ray source called Cygnus X-1, discovered by the Uhuru satellite in the early 1970s. The main observational features of this binary system are as follows:
(Sec. 17.9)
Figure 22.18 (a) The brightest star in this photograph is a member of a binary system whose unseen companion, called Cygnus X-1, is thought to be a good candidate for a black hole. (b) X-rays emitted by the Cygnus X-1 source were analyzed by changing them into electronic signals that were then viewed on a video screen, from which this picture was taken. (The field of view here is outlined by the rectangle in the previous figure.)
These general properties suggest that the invisible X-ray-emitting companion could be a black hole. The X-ray-emitting region is likely an accretion disk formed as matter drawn from the visible star spirals down onto the unseen component. The rapid variability of the X-ray emission indicates that the unseen component must be compact—a neutron star or a black hole. The mass limit of the dark component argues for the latter, as astronomers believe that neutron stars' masses cannot exceed about 3 solar masses. Figure 22.19 is an artist's conception of this intriguing object. As shown, most of the gas drawn from the visible star ends up in a donut-shaped accretion disk of matter. As the gas flows toward the black hole it becomes superheated and emits the X-rays we observe, just before they are trapped forever below the event horizon.
Figure 22.19 Artist's conception of a binary system containing a large, bright, visible star and an invisible, X-ray-emitting black hole. This drawing is based on data obtained from detailed study of Cygnus X-1.
A few other black-hole candidates are known. For example, the third X-ray source ever discovered in the Large Magellanic Cloud—called LMC X-3—is an invisible object that, like Cygnus X-1, orbits a bright companion star. LMC X-3's visible companion seems to be distorted into the shape of an egg by the unseen object's intense gravitational pull. Reasoning similar to that applied to Cygnus X-1 leads to the conclusion that the compact object LMC X-3 has a mass nearly 10 times that of the Sun, making it too massive to be anything but a black hole. The X-ray binary system A0620-00 has been found to contain an invisible compact object of mass 3.8 times the mass of the Sun.
In total, there are perhaps half a dozen known objects that may turn out to be black holes, although Cygnus X-1, LMC X-3, and A0620-00 have the strongest claims. But how sure are we that these objects are black holes? Cygnus X-1 and the other suspected black holes in binary systems all have masses relatively close to the dividing line separating neutron stars from black holes. Given the present uncertainties in both observations and theory, might they conceivably be merely dim, dense neutron stars and not black holes at all?
Most astronomers do not regard this as a likely possibility, but it highlights a problem. There is presently no clear observational test that can unambiguously distinguish a 10—solar mass black hole from, say, a 10—solar mass neutron star (if one could somehow exist). Both objects would affect a companion star's orbit in the same way; both would tear mass from its surface, and both would form an accretion disk around themselves that would emit intense X-rays (although some researchers believe that the accretion disks may differ sufficiently in their detailed properties that the nature of the central object may be identifiable from observations). The radiation from the surface of the neutron star itself, which would distinguish it from a black hole, might in some cases be so weak that it would be impossible to detect against the emission from the disk. We rule out the neutron star on purely theoretical grounds (on the basis of our understanding of neutron star masses), but we have no hard observational evidence to support our theory.
The argument really proceeds by elimination. Loosely stated, it goes: "Object X is compact and very massive. We don't know of anything else that can be that small and that massive. Therefore, object 
is a black hole." We will see other instances later in this book in which astronomers use similar reasoning to infer the existence of very massive black holes in the hearts of galaxies. Still, some astronomers are troubled that the black-hole category has in some ways become a catch-all for things that have no other reasonable explanation.
So have stellar black holes really been discovered? The answer is probably yes. Skepticism is healthy in science, but only the most stubborn astronomers (and some do exist!) would take serious issue with the reasoning that supports the case for black holes. Can we guarantee that future modifications to the theory of compact objects will not invalidate our arguments? No, but similar statements could be made in many areas of astronomy—indeed, about any theory in any area of science. We conclude that, strange as they are, black holes have been detected in our galaxy. Perhaps some day, future generations of space travelers will visit Cygnus X-1 or LMC X-3 and (carefully!) test these conclusions firsthand. Until then, we will have to continue to rely on improving theoretical models and observational techniques to guide our discussions of these mysterious objects.
