Release 621 of 1,056
Hubble Discovers Black Holes in Unexpected Places
Release date: Sep 17, 2002 12:00 PM (EDT)
Release type: Space Science Update

Medium-size black holes actually do exist, according to the latest findings from NASA's Hubble Space Telescope, but scientists had to look in some unexpected places to find them. The previously undiscovered black holes provide an important link that sheds light on the way in which black holes grow. Even more odd, these new black holes were found in the cores of glittering, "beehive" swarms of stars called globular star clusters, which orbit our Milky Way and other galaxies. The black hole in globular cluster M15 [left] is 4,000 times more massive than our Sun. G1 [right], a much larger globular cluster, harbors a heftier black hole, about 20,000 times more massive than our Sun.

Release ID: STScI-2002-18
Release images (10)
Black Holes Unexpectedly Discovered in Globular Clusters M15 and G1
Release videos (2)
Black Hole in a Globular Cluster

Image Credits for M15: NASA and The Hubble Heritage Team (STScI/AURA)

Image Credits for G1: NASA and Michael Rich (UCLA)

Science Credits: NASA, Roeland Van Der Marel and Joris Gerssen (Space Telescope Science Institute), Puragra Guhathakurta and Ruth Peterson (University of California Observatories/Lick Observatory), Carlton Pryor (Rutgers University), Michael Rich (UCLA), Karl Gebhardt (University of Texas), and Luis Ho (Carnegie Institution of Washington)

The Full Story
Release date: Sep 17, 2002
Hubble Discovers Black Holes in Unexpected Places

Medium-size black holes actually do exist, according to the latest findings from NASA's Hubble Space Telescope, but scientists had to look in some unexpected places to find them.

The previously undiscovered black holes provide an important link that sheds light on the way black holes grow. Even more odd, these new black holes were found in the cores of glittering, "beehive" swarms of stars - called globular star clusters - that orbit our Milky Way and other galaxies.

The new findings promise a better understanding of how galaxies and globular clusters first formed billions of years ago. Globular star clusters contain the oldest stars in the universe. If globulars have black holes now, then globulars most likely had black holes when they originally formed. The new results indicate that the very sedate, elderly environments of globular clusters house these exotic objects, quite unlike the violent cores of some galaxies.

"These findings may be telling us something very deep about the formation of star clusters and black holes in the early universe," says Roeland Van Der Marel of the Space Telescope Science Institute in Baltimore. "Black holes are even more common in the universe than previously thought."

"Not only will we learn about the formation of the black holes, but these new data from Hubble help us connect globular clusters to galaxies, providing information on one of the most important unsolved problems in astronomy today: how galaxy structure forms in the universe," adds Michael Rich of the University of California, Los Angeles (UCLA).

These intermediate-mass black holes may provide a link between stellar-mass and supermassive black holes. This link is important because it may hold the clue to how supermassive black holes form in galaxies. This is reinforced by the uncanny fact found by these investigations that a black hole's mass is proportional to the mass of the stellar environment it inhabits. Supermassive black holes found by Hubble in the centers of galaxies represent about 0.5 percent of the galaxies' mass. Amazingly, the black holes now found in star clusters, which are 10,000 times less massive than a galaxy, also obey this trend. It appears that there is some yet-to-be-discovered underlying process that ties a black hole to its host in a fundamental way. Nature is providing a big clue as to how these systems and their black holes form.

"The intermediate-mass black holes that have now been found with Hubble may be the building blocks of the supermassive black holes that dwell in the centers of most galaxies," says Karl Gebhardt of the University of Texas at Austin.

Van Der Marel led a team that uncovered a black hole in the center of the globular star cluster M15, 32,000 light-years away in the constellation Pegasus. His collaborator Joris Gerssen, also of the Space Telescope Science Institute, pinned down the black hole's mass at 4,000 times that of our Sun.

In a separate observing program, a team led by Rich, and including Gebhardt and Luis Ho of the Carnegie Institution of Washington, found a 20,000-solar-mass black hole in the giant globular cluster G1, located 70 times farther - 2.2 million light-years away - in the neighboring Andromeda galaxy. By contrast, stellar-mass black holes are only a few times the mass of our Sun, and galactic-center black holes can be millions or billions of times more massive than our Sun.

"G1 has a total mass of 10 million suns, making it about the most massive globular cluster known." says Rich. "It also has a very bright core, so I thought it would be a good place to search for a massive black hole."

A black hole is an infinitely small and dense region where space is so tightly warped by gravity that not even light can escape. For many years, astronomers have known two types - "supermassive" black holes at the centers of large galaxies and the so-called "stellar-mass" black holes that result when a star about 10 times the Sun's mass ends its life in a supernova explosion. Both types have been detected and measured.

"There are two main theories of black hole formation," says Gebhardt. "You could either make the black hole all at once, when the galaxy is forming, by dumping a lot of material in the middle, or you could start with a seed black hole that subsequently grows over time. The observational evidence now points to the idea that you start out with a small seed black hole." The fact that globular clusters have these small black holes implies that they are excellent candidates to act as the seeds for the supermassive black holes that lurk in the centers of nearly all galaxies.

"The Hubble results add new credibility to the latter scenario," says Van Der Marel. "Black holes similar to the ones now found in globular clusters may have been the building blocks that formed supermassive black holes."

Previously, X-ray observations from the ROSAT Observatory and NASA's Chandra Observatory have identified ultra-bright X-ray sources that could also be interpreted as intermediate-mass black holes in star-forming galaxies. However, alternative interpretations for these X-ray sources continue to exist. By contrast, Hubble's measurements are based on the velocities of stars whirling around in the dense cores of globular clusters, which yield a direct measurement of the black hole masses.

The M15 globular star cluster is close enough that individual star speeds can be measured. By contrast, the G1 observations rely on measurements of the collective properties of many stars. In either case, a black hole can be identified by using a common Hubble black-hole-hunting technique, which searches for a rise in velocities toward the cluster center. Stars close to the black-hole "whirlpool" orbit at a faster rate, in keeping with fundamental laws of orbital motion around a massive central body, as described by Johannes Kepler four centuries ago.

Black holes cannot be seen directly. Some emit X-rays, or show other telltale evidence of their presence, when they capture nearby material. However, the dark objects in the centers of G1 and M15 are quiet. Nonetheless, they are presumed to be black holes because of their small size and large mass. An alternative explanation would be to assume that the centers of these clusters harbor a swarm of neutron stars or other exotic objects that sank to the cluster's center. However, theoretical studies do not predict swarms that are massive enough to account for the Hubble observations.

Astronomers have searched for black holes in globular clusters for nearly 30 years. The roadblock has been the fact that ground-based telescopes cannot easily resolve the stars closest to the suspected black hole. As far back as the 1970s, hunting for globular-cluster black holes was recognized as a task suited for Hubble Space Telescope's exquisite resolution, which is needed for looking close to a black hole. The researchers say that the quest is now over.

To further understand these issues, it is now extremely important to search for black holes in other star clusters as well. Some globular clusters are so close to us that, if they had black holes, we would be able to probe closer to these monsters than we have ever been able to before.

The members of the G1 research team are Michael Rich, Karl Gebhardt and Luis Ho. The members of the M15 research team are Roeland Van Der Marel, Joris Gerssen, Karl Gebhardt, Puragra Guhathakurta, Ruth Peterson and Carlton Pryor.


Don Savage
NASA Headquarters, Washington, DC
(Phone: 202-358-1547; E-mail:

Nancy Neal
Goddard Space Flight Center, Greenbelt, MD
(Phone: 301-286-0039; E-mail:

Ray Villard
Space Telescope Science Institute, Baltimore, MD
(Phone: 410-338-4514; E-mail:

Rebecca Johnson
The University of Texas, Austin, TX
(Phone: 512-475-6763; E-mail:

Harlan Lebo
UCLA, Los Angeles, CA
(Phone: 310-206-0510; E-mail: