A dense clump of stars a few thousand light-years away has been harbouring a surprise in its core. Rather than one relatively chunky black hole, astronomers have found that the globular cluster NGC 6397 is wrapped around a cluster of smaller, stellar-mass ones.

This could not only help us better understand the formation of larger black holes, it suggests that globular clusters could be of great interest to gravitational wave astronomy as the black holes inevitably draw closer together towards collision.

Globular clusters of stars are often considered "fossils" of the early Universe. They're very dense, spherical clusters of roughly 100,000 to 1 million very old stars, some - like NGC 6397 - nearly as old as the Universe itself. In any globular cluster, all its stars formed at the same time, from the same cloud of gas. The Milky Way has around 150 known globular clusters.

These objects are excellent tools for studying, for example, the history of the Universe, or the dark matter content of the galaxies they orbit. Recently, however, astronomers have been looking at them more closely as the potential homes of an elusive class of objects - intermediate-mass black holes.

As the name suggests, these middleweights sit between stellar-mass and supermassive black holes, the latter of which are typically found at the centres of galaxies.

While the boundaries between intermediate-mass black holes and supermassive black holes are currently not very well defined, intermediate-mass black holes are generally considered to be larger than a typical collapsed star (up to a hundred solar masses) but not supermassive (between a million and a billion times more mass than a typical stellar black hole).

Firm evidence for the existence of intermediate-mass black holes is, however, scarce and largely inconclusive. Theory and modelling suggest that they could be found in globular clusters, the gravitational core around which the stars rally, like larger galaxies around supermassive black holes.

The properties of NGC 6397, about 7,800 light-years away, suggested that there might be one of these middleweights at its centre.

Since we can't see black holes (because they don't give off any detectable radiation), astronomers took a closer look at the orbits of stars in the cluster, based on years of Hubble data, to see if they indicated an intermediate-mass black hole.

"We found very strong evidence for an invisible mass in the dense core of the globular cluster," said astronomer Eduardo Vitral of the Paris Institute of Astrophysics in France, "but we were surprised to find that this extra mass is not 'point-like' (that would be expected for a solitary massive black hole) but extended to a few percent of the size of the cluster."

This is consistent with a type of drag known as dynamical friction, in which objects in the cluster exchange momentum, sending denser, more massive objects towards the core, and less massive objects towards the outskirts.

Dead stars such as white dwarfs, neutron stars and black holes are denser than main-sequence stars, so they move inwards, sending the lighter stars out.

"We used the theory of stellar evolution to conclude that most of the extra mass we found was in the form of black holes," said astronomer Gary Mamon of the Paris Institute of Astrophysics.

It is also consistent with two recent papers, which found that, instead of intermediate-mass black holes, populations of stellar-mass black holes could inhabit the central regions of globular clusters. Now those findings have been validated.

"Ours is the first study to provide both the mass and the extent of what appears to be a collection of mostly black holes in the center of a core-collapsed globular cluster," Vitral said.

This is useful information both in study for stellar-mass black holes and the hunt for intermediate-mass black holes. Now that we have observational evidence that this can happen, astronomers can refine their searches to rule out globular clusters that behave the same way.

There are implications for other black hole research, too.

Because the objects will continue to sink towards the centre of the cluster, the team believes that eventually, they will start to spiral into each other and merge. Eventually - a very, very long time from now - this could result in an intermediate-mass black hole.

More immediately, this ongoing process suggests that the cores of such clusters could be very important for gravitational wave astronomy. Because they're so closely packed, the processes should be accelerated, which means we could look to these regions both to study pre-merger conditions, and to try to pre-empt the gravitational wave events that will occur when the black holes merge.

The research has been published in Astronomy & Astrophysics.